This the multi-page printable view of this section. Click here to print.
This section of the Kubernetes documentation contains pages that show how to do individual tasks. A task page shows how to do a single thing, typically by giving a short sequence of steps.
If you would like to write a task page, see Creating a Documentation Pull Request.
The Kubernetes command-line tool, kubectl, allows
you to run commands against Kubernetes clusters.
You can use kubectl to deploy applications, inspect and manage cluster resources,
and view logs. For more information including a complete list of kubectl operations, see the
kubectl reference documentation.
kubectl is installable on a variety of Linux platforms, macOS and Windows. Find your preferred operating system below.
kind lets you run Kubernetes on
your local computer. This tool requires that you have
Docker installed and configured.
The kind Quick Start page shows you what you need to do to get up and running with kind.
Like kind, minikube is a tool that lets you run Kubernetes
locally. minikube runs a single-node Kubernetes cluster on your personal
computer (including Windows, macOS and Linux PCs) so that you can try out
Kubernetes, or for daily development work.
You can follow the official Get Started! guide if your focus is on getting the tool installed.
View minikube Get Started! Guide
Once you have minikube working, you can use it to
run a sample application.
You can use the kubeadm tool to create and manage Kubernetes clusters. It performs the actions necessary to get a minimum viable, secure cluster up and running in a user friendly way.
Installing kubeadm shows you how to install kubeadm. Once installed, you can use it to create a cluster.
You must use a kubectl version that is within one minor version difference of your cluster. For example, a v1.22 client can communicate with v1.21, v1.22, and v1.23 control planes. Using the latest version of kubectl helps avoid unforeseen issues.
The following methods exist for installing kubectl on Linux:
Download the latest release with the command:
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl"
To download a specific version, replace the $(curl -L -s https://dl.k8s.io/release/stable.txt) portion of the command with the specific version.
For example, to download version v1.22.0 on Linux, type:
curl -LO https://dl.k8s.io/release/v1.22.0/bin/linux/amd64/kubectl
Validate the binary (optional)
Download the kubectl checksum file:
curl -LO "https://dl.k8s.io/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl.sha256"
Validate the kubectl binary against the checksum file:
echo "$(<kubectl.sha256) kubectl" | sha256sum --check
If valid, the output is:
kubectl: OK
If the check fails, sha256 exits with nonzero status and prints output similar to:
kubectl: FAILED
sha256sum: WARNING: 1 computed checksum did NOT match
Install kubectl
sudo install -o root -g root -m 0755 kubectl /usr/local/bin/kubectl
If you do not have root access on the target system, you can still install kubectl to the ~/.local/bin directory:
chmod +x kubectl
mkdir -p ~/.local/bin/kubectl
mv ./kubectl ~/.local/bin/kubectl
# and then add ~/.local/bin/kubectl to $PATH
Test to ensure the version you installed is up-to-date:
kubectl version --client
Update the apt package index and install packages needed to use the Kubernetes apt repository:
sudo apt-get update
sudo apt-get install -y apt-transport-https ca-certificates curl
Download the Google Cloud public signing key:
sudo curl -fsSLo /usr/share/keyrings/kubernetes-archive-keyring.gpg https://packages.cloud.google.com/apt/doc/apt-key.gpg
Add the Kubernetes apt repository:
echo "deb [signed-by=/usr/share/keyrings/kubernetes-archive-keyring.gpg] https://apt.kubernetes.io/ kubernetes-xenial main" | sudo tee /etc/apt/sources.list.d/kubernetes.list
Update apt package index with the new repository and install kubectl:
sudo apt-get update
sudo apt-get install -y kubectl
cat <<EOF > /etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=https://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64
enabled=1
gpgcheck=1
repo_gpgcheck=1
gpgkey=https://packages.cloud.google.com/yum/doc/yum-key.gpg https://packages.cloud.google.com/yum/doc/rpm-package-key.gpg
EOF
yum install -y kubectl
If you are on Ubuntu or another Linux distribution that support snap package manager, kubectl is available as a snap application.
snap install kubectl --classic
kubectl version --client
If you are on Linux and using Homebrew package manager, kubectl is available for installation.
brew install kubectl
kubectl version --client
In order for kubectl to find and access a Kubernetes cluster, it needs a
kubeconfig file,
which is created automatically when you create a cluster using
kube-up.sh
or successfully deploy a Minikube cluster.
By default, kubectl configuration is located at ~/.kube/config.
Check that kubectl is properly configured by getting the cluster state:
kubectl cluster-info
If you see a URL response, kubectl is correctly configured to access your cluster.
If you see a message similar to the following, kubectl is not configured correctly or is not able to connect to a Kubernetes cluster.
The connection to the server <server-name:port> was refused - did you specify the right host or port?
For example, if you are intending to run a Kubernetes cluster on your laptop (locally), you will need a tool like Minikube to be installed first and then re-run the commands stated above.
If kubectl cluster-info returns the url response but you can't access your cluster, to check whether it is configured properly, use:
kubectl cluster-info dump
kubectl provides autocompletion support for Bash and Zsh, which can save you a lot of typing.
Below are the procedures to set up autocompletion for Bash and Zsh.
The kubectl completion script for Bash can be generated with the command kubectl completion bash. Sourcing the completion script in your shell enables kubectl autocompletion.
However, the completion script depends on bash-completion, which means that you have to install this software first (you can test if you have bash-completion already installed by running type _init_completion).
bash-completion is provided by many package managers (see here). You can install it with apt-get install bash-completion or yum install bash-completion, etc.
The above commands create /usr/share/bash-completion/bash_completion, which is the main script of bash-completion. Depending on your package manager, you have to manually source this file in your ~/.bashrc file.
To find out, reload your shell and run type _init_completion. If the command succeeds, you're already set, otherwise add the following to your ~/.bashrc file:
source /usr/share/bash-completion/bash_completion
Reload your shell and verify that bash-completion is correctly installed by typing type _init_completion.
You now need to ensure that the kubectl completion script gets sourced in all your shell sessions. There are two ways in which you can do this:
Source the completion script in your ~/.bashrc file:
echo 'source <(kubectl completion bash)' >>~/.bashrc
Add the completion script to the /etc/bash_completion.d directory:
kubectl completion bash >/etc/bash_completion.d/kubectl
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.bashrc
echo 'complete -F __start_kubectl k' >>~/.bashrc
/etc/bash_completion.d.
Both approaches are equivalent. After reloading your shell, kubectl autocompletion should be working.
The kubectl completion script for Zsh can be generated with the command kubectl completion zsh. Sourcing the completion script in your shell enables kubectl autocompletion.
To do so in all your shell sessions, add the following to your ~/.zshrc file:
source <(kubectl completion zsh)
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.zshrc
echo 'complete -F __start_kubectl k' >>~/.zshrc
After reloading your shell, kubectl autocompletion should be working.
If you get an error like complete:13: command not found: compdef, then add the following to the beginning of your ~/.zshrc file:
autoload -Uz compinit
compinit
kubectl convert pluginA plugin for Kubernetes command-line tool kubectl, which allows you to convert manifests between different API
versions. This can be particularly helpful to migrate manifests to a non-deprecated api version with newer Kubernetes release.
For more info, visit migrate to non deprecated apis
Download the latest release with the command:
curl -LO https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl-convert
Validate the binary (optional)
Download the kubectl-convert checksum file:
curl -LO "https://dl.k8s.io/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl-convert.sha256"
Validate the kubectl-convert binary against the checksum file:
echo "$(<kubectl-convert.sha256) kubectl-convert" | sha256sum --check
If valid, the output is:
kubectl-convert: OK
If the check fails, sha256 exits with nonzero status and prints output similar to:
kubectl-convert: FAILED
sha256sum: WARNING: 1 computed checksum did NOT match
Install kubectl-convert
sudo install -o root -g root -m 0755 kubectl-convert /usr/local/bin/kubectl-convert
Verify plugin is successfully installed
kubectl convert --help
If you do not see an error, it means the plugin is successfully installed.
You must use a kubectl version that is within one minor version difference of your cluster. For example, a v1.22 client can communicate with v1.21, v1.22, and v1.23 control planes. Using the latest version of kubectl helps avoid unforeseen issues.
The following methods exist for installing kubectl on macOS:
Download the latest release:
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl"
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/arm64/kubectl"
To download a specific version, replace the $(curl -L -s https://dl.k8s.io/release/stable.txt) portion of the command with the specific version.
For example, to download version v1.22.0 on Intel macOS, type:
curl -LO "https://dl.k8s.io/release/v1.22.0/bin/darwin/amd64/kubectl"
And for macOS on Apple Silicon, type:
curl -LO "https://dl.k8s.io/release/v1.22.0/bin/darwin/arm64/kubectl"
Validate the binary (optional)
Download the kubectl checksum file:
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl.sha256"
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/arm64/kubectl.sha256"
Validate the kubectl binary against the checksum file:
echo "$(<kubectl.sha256) kubectl" | shasum -a 256 --check
If valid, the output is:
kubectl: OK
If the check fails, shasum exits with nonzero status and prints output similar to:
kubectl: FAILED
shasum: WARNING: 1 computed checksum did NOT match
Make the kubectl binary executable.
chmod +x ./kubectl
Move the kubectl binary to a file location on your system PATH.
sudo mv ./kubectl /usr/local/bin/kubectl
sudo chown root: /usr/local/bin/kubectl
/usr/local/bin is in your PATH environment variable.
Test to ensure the version you installed is up-to-date:
kubectl version --client
If you are on macOS and using Homebrew package manager, you can install kubectl with Homebrew.
Run the installation command:
brew install kubectl
or
brew install kubernetes-cli
Test to ensure the version you installed is up-to-date:
kubectl version --client
If you are on macOS and using Macports package manager, you can install kubectl with Macports.
Run the installation command:
sudo port selfupdate
sudo port install kubectl
Test to ensure the version you installed is up-to-date:
kubectl version --client
In order for kubectl to find and access a Kubernetes cluster, it needs a
kubeconfig file,
which is created automatically when you create a cluster using
kube-up.sh
or successfully deploy a Minikube cluster.
By default, kubectl configuration is located at ~/.kube/config.
Check that kubectl is properly configured by getting the cluster state:
kubectl cluster-info
If you see a URL response, kubectl is correctly configured to access your cluster.
If you see a message similar to the following, kubectl is not configured correctly or is not able to connect to a Kubernetes cluster.
The connection to the server <server-name:port> was refused - did you specify the right host or port?
For example, if you are intending to run a Kubernetes cluster on your laptop (locally), you will need a tool like Minikube to be installed first and then re-run the commands stated above.
If kubectl cluster-info returns the url response but you can't access your cluster, to check whether it is configured properly, use:
kubectl cluster-info dump
kubectl provides autocompletion support for Bash and Zsh, which can save you a lot of typing.
Below are the procedures to set up autocompletion for Bash and Zsh.
The kubectl completion script for Bash can be generated with kubectl completion bash. Sourcing this script in your shell enables kubectl completion.
However, the kubectl completion script depends on bash-completion which you thus have to previously install.
The instructions here assume you use Bash 4.1+. You can check your Bash's version by running:
echo $BASH_VERSION
If it is too old, you can install/upgrade it using Homebrew:
brew install bash
Reload your shell and verify that the desired version is being used:
echo $BASH_VERSION $SHELL
Homebrew usually installs it at /usr/local/bin/bash.
You can test if you have bash-completion v2 already installed with type _init_completion. If not, you can install it with Homebrew:
brew install bash-completion@2
As stated in the output of this command, add the following to your ~/.bash_profile file:
export BASH_COMPLETION_COMPAT_DIR="/usr/local/etc/bash_completion.d"
[[ -r "/usr/local/etc/profile.d/bash_completion.sh" ]] && . "/usr/local/etc/profile.d/bash_completion.sh"
Reload your shell and verify that bash-completion v2 is correctly installed with type _init_completion.
You now have to ensure that the kubectl completion script gets sourced in all your shell sessions. There are multiple ways to achieve this:
Source the completion script in your ~/.bash_profile file:
echo 'source <(kubectl completion bash)' >>~/.bash_profile
Add the completion script to the /usr/local/etc/bash_completion.d directory:
kubectl completion bash >/usr/local/etc/bash_completion.d/kubectl
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.bash_profile
echo 'complete -F __start_kubectl k' >>~/.bash_profile
If you installed kubectl with Homebrew (as explained here), then the kubectl completion script should already be in /usr/local/etc/bash_completion.d/kubectl. In that case, you don't need to do anything.
BASH_COMPLETION_COMPAT_DIR directory, that's why the latter two methods work.
In any case, after reloading your shell, kubectl completion should be working.
The kubectl completion script for Zsh can be generated with the command kubectl completion zsh. Sourcing the completion script in your shell enables kubectl autocompletion.
To do so in all your shell sessions, add the following to your ~/.zshrc file:
source <(kubectl completion zsh)
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.zshrc
echo 'complete -F __start_kubectl k' >>~/.zshrc
After reloading your shell, kubectl autocompletion should be working.
If you get an error like complete:13: command not found: compdef, then add the following to the beginning of your ~/.zshrc file:
autoload -Uz compinit
compinit
kubectl convert pluginA plugin for Kubernetes command-line tool kubectl, which allows you to convert manifests between different API
versions. This can be particularly helpful to migrate manifests to a non-deprecated api version with newer Kubernetes release.
For more info, visit migrate to non deprecated apis
Download the latest release with the command:
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl-convert"
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/arm64/kubectl-convert"
Validate the binary (optional)
Download the kubectl-convert checksum file:
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/amd64/kubectl-convert.sha256"
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/darwin/arm64/kubectl-convert.sha256"
Validate the kubectl-convert binary against the checksum file:
echo "$(<kubectl-convert.sha256) kubectl-convert" | shasum -a 256 --check
If valid, the output is:
kubectl-convert: OK
If the check fails, shasum exits with nonzero status and prints output similar to:
kubectl-convert: FAILED
shasum: WARNING: 1 computed checksum did NOT match
Make kubectl-convert binary executable
chmod +x ./kubectl-convert
Move the kubectl-convert binary to a file location on your system PATH.
sudo mv ./kubectl-convert /usr/local/bin/kubectl-convert
sudo chown root: /usr/local/bin/kubectl-convert
/usr/local/bin is in your PATH environment variable.
Verify plugin is successfully installed
kubectl convert --help
If you do not see an error, it means the plugin is successfully installed.
You must use a kubectl version that is within one minor version difference of your cluster. For example, a v1.22 client can communicate with v1.21, v1.22, and v1.23 control planes. Using the latest version of kubectl helps avoid unforeseen issues.
The following methods exist for installing kubectl on Windows:
Download the latest release v1.22.0.
Or if you have curl installed, use this command:
curl -LO "https://dl.k8s.io/release/v1.22.0/bin/windows/amd64/kubectl.exe"
Validate the binary (optional)
Download the kubectl checksum file:
curl -LO "https://dl.k8s.io/v1.22.0/bin/windows/amd64/kubectl.exe.sha256"
Validate the kubectl binary against the checksum file:
Using Command Prompt to manually compare CertUtil's output to the checksum file downloaded:
CertUtil -hashfile kubectl.exe SHA256
type kubectl.exe.sha256
Using PowerShell to automate the verification using the -eq operator to get a True or False result:
$($(CertUtil -hashfile .\kubectl.exe SHA256)[1] -replace " ", "") -eq $(type .\kubectl.exe.sha256)
Add the binary in to your PATH.
Test to ensure the version of kubectl is the same as downloaded:
kubectl version --client
kubectl to PATH.
If you have installed Docker Desktop before, you may need to place your PATH entry before the one added by the Docker Desktop installer or remove the Docker Desktop's kubectl.
To install kubectl on Windows you can use either Chocolatey package manager or Scoop command-line installer.
choco install kubernetes-cli
scoop install kubectl
Test to ensure the version you installed is up-to-date:
kubectl version --client
Navigate to your home directory:
# If you're using cmd.exe, run: cd %USERPROFILE%
cd ~
Create the .kube directory:
mkdir .kube
Change to the .kube directory you just created:
cd .kube
Configure kubectl to use a remote Kubernetes cluster:
New-Item config -type file
In order for kubectl to find and access a Kubernetes cluster, it needs a
kubeconfig file,
which is created automatically when you create a cluster using
kube-up.sh
or successfully deploy a Minikube cluster.
By default, kubectl configuration is located at ~/.kube/config.
Check that kubectl is properly configured by getting the cluster state:
kubectl cluster-info
If you see a URL response, kubectl is correctly configured to access your cluster.
If you see a message similar to the following, kubectl is not configured correctly or is not able to connect to a Kubernetes cluster.
The connection to the server <server-name:port> was refused - did you specify the right host or port?
For example, if you are intending to run a Kubernetes cluster on your laptop (locally), you will need a tool like Minikube to be installed first and then re-run the commands stated above.
If kubectl cluster-info returns the url response but you can't access your cluster, to check whether it is configured properly, use:
kubectl cluster-info dump
kubectl provides autocompletion support for Bash and Zsh, which can save you a lot of typing.
Below are the procedures to set up autocompletion for Zsh, if you are running that on Windows.
The kubectl completion script for Zsh can be generated with the command kubectl completion zsh. Sourcing the completion script in your shell enables kubectl autocompletion.
To do so in all your shell sessions, add the following to your ~/.zshrc file:
source <(kubectl completion zsh)
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.zshrc
echo 'complete -F __start_kubectl k' >>~/.zshrc
After reloading your shell, kubectl autocompletion should be working.
If you get an error like complete:13: command not found: compdef, then add the following to the beginning of your ~/.zshrc file:
autoload -Uz compinit
compinit
kubectl convert pluginA plugin for Kubernetes command-line tool kubectl, which allows you to convert manifests between different API
versions. This can be particularly helpful to migrate manifests to a non-deprecated api version with newer Kubernetes release.
For more info, visit migrate to non deprecated apis
Download the latest release with the command:
curl -LO "https://dl.k8s.io/release/v1.22.0/bin/windows/amd64/kubectl-convert.exe"
Validate the binary (optional)
Download the kubectl-convert checksum file:
curl -LO "https://dl.k8s.io/v1.22.0/bin/windows/amd64/kubectl-convert.exe.sha256"
Validate the kubectl-convert binary against the checksum file:
Using Command Prompt to manually compare CertUtil's output to the checksum file downloaded:
CertUtil -hashfile kubectl-convert.exe SHA256
type kubectl-convert.exe.sha256
Using PowerShell to automate the verification using the -eq operator to get a True or False result:
$($(CertUtil -hashfile .\kubectl-convert.exe SHA256)[1] -replace " ", "") -eq $(type .\kubectl-convert.exe.sha256)
Add the binary in to your PATH.
Verify plugin is successfully installed
kubectl convert --help
If you do not see an error, it means the plugin is successfully installed.
The kubectl completion script for Bash can be generated with the command kubectl completion bash. Sourcing the completion script in your shell enables kubectl autocompletion.
However, the completion script depends on bash-completion, which means that you have to install this software first (you can test if you have bash-completion already installed by running type _init_completion).
bash-completion is provided by many package managers (see here). You can install it with apt-get install bash-completion or yum install bash-completion, etc.
The above commands create /usr/share/bash-completion/bash_completion, which is the main script of bash-completion. Depending on your package manager, you have to manually source this file in your ~/.bashrc file.
To find out, reload your shell and run type _init_completion. If the command succeeds, you're already set, otherwise add the following to your ~/.bashrc file:
source /usr/share/bash-completion/bash_completion
Reload your shell and verify that bash-completion is correctly installed by typing type _init_completion.
You now need to ensure that the kubectl completion script gets sourced in all your shell sessions. There are two ways in which you can do this:
Source the completion script in your ~/.bashrc file:
echo 'source <(kubectl completion bash)' >>~/.bashrc
Add the completion script to the /etc/bash_completion.d directory:
kubectl completion bash >/etc/bash_completion.d/kubectl
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.bashrc
echo 'complete -F __start_kubectl k' >>~/.bashrc
/etc/bash_completion.d.
Both approaches are equivalent. After reloading your shell, kubectl autocompletion should be working.
The kubectl completion script for Bash can be generated with kubectl completion bash. Sourcing this script in your shell enables kubectl completion.
However, the kubectl completion script depends on bash-completion which you thus have to previously install.
The instructions here assume you use Bash 4.1+. You can check your Bash's version by running:
echo $BASH_VERSION
If it is too old, you can install/upgrade it using Homebrew:
brew install bash
Reload your shell and verify that the desired version is being used:
echo $BASH_VERSION $SHELL
Homebrew usually installs it at /usr/local/bin/bash.
You can test if you have bash-completion v2 already installed with type _init_completion. If not, you can install it with Homebrew:
brew install bash-completion@2
As stated in the output of this command, add the following to your ~/.bash_profile file:
export BASH_COMPLETION_COMPAT_DIR="/usr/local/etc/bash_completion.d"
[[ -r "/usr/local/etc/profile.d/bash_completion.sh" ]] && . "/usr/local/etc/profile.d/bash_completion.sh"
Reload your shell and verify that bash-completion v2 is correctly installed with type _init_completion.
You now have to ensure that the kubectl completion script gets sourced in all your shell sessions. There are multiple ways to achieve this:
Source the completion script in your ~/.bash_profile file:
echo 'source <(kubectl completion bash)' >>~/.bash_profile
Add the completion script to the /usr/local/etc/bash_completion.d directory:
kubectl completion bash >/usr/local/etc/bash_completion.d/kubectl
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.bash_profile
echo 'complete -F __start_kubectl k' >>~/.bash_profile
If you installed kubectl with Homebrew (as explained here), then the kubectl completion script should already be in /usr/local/etc/bash_completion.d/kubectl. In that case, you don't need to do anything.
BASH_COMPLETION_COMPAT_DIR directory, that's why the latter two methods work.
In any case, after reloading your shell, kubectl completion should be working.
A plugin for Kubernetes command-line tool kubectl, which allows you to convert manifests between different API
versions. This can be particularly helpful to migrate manifests to a non-deprecated api version with newer Kubernetes release.
For more info, visit migrate to non deprecated apis
In order for kubectl to find and access a Kubernetes cluster, it needs a
kubeconfig file,
which is created automatically when you create a cluster using
kube-up.sh
or successfully deploy a Minikube cluster.
By default, kubectl configuration is located at ~/.kube/config.
Check that kubectl is properly configured by getting the cluster state:
kubectl cluster-info
If you see a URL response, kubectl is correctly configured to access your cluster.
If you see a message similar to the following, kubectl is not configured correctly or is not able to connect to a Kubernetes cluster.
The connection to the server <server-name:port> was refused - did you specify the right host or port?
For example, if you are intending to run a Kubernetes cluster on your laptop (locally), you will need a tool like Minikube to be installed first and then re-run the commands stated above.
If kubectl cluster-info returns the url response but you can't access your cluster, to check whether it is configured properly, use:
kubectl cluster-info dump
The kubectl completion script for Zsh can be generated with the command kubectl completion zsh. Sourcing the completion script in your shell enables kubectl autocompletion.
To do so in all your shell sessions, add the following to your ~/.zshrc file:
source <(kubectl completion zsh)
If you have an alias for kubectl, you can extend shell completion to work with that alias:
echo 'alias k=kubectl' >>~/.zshrc
echo 'complete -F __start_kubectl k' >>~/.zshrc
After reloading your shell, kubectl autocompletion should be working.
If you get an error like complete:13: command not found: compdef, then add the following to the beginning of your ~/.zshrc file:
autoload -Uz compinit
compinit
Kubernetes v1.15 [stable]
Client certificates generated by kubeadm expire after 1 year. This page explains how to manage certificate renewals with kubeadm.
You should be familiar with PKI certificates and requirements in Kubernetes.
By default, kubeadm generates all the certificates needed for a cluster to run. You can override this behavior by providing your own certificates.
To do so, you must place them in whatever directory is specified by the
--cert-dir flag or the certificatesDir field of kubeadm's ClusterConfiguration.
By default this is /etc/kubernetes/pki.
If a given certificate and private key pair exists before running kubeadm init,
kubeadm does not overwrite them. This means you can, for example, copy an existing
CA into /etc/kubernetes/pki/ca.crt and /etc/kubernetes/pki/ca.key,
and kubeadm will use this CA for signing the rest of the certificates.
It is also possible to provide only the ca.crt file and not the
ca.key file (this is only available for the root CA file, not other cert pairs).
If all other certificates and kubeconfig files are in place, kubeadm recognizes
this condition and activates the "External CA" mode. kubeadm will proceed without the
CA key on disk.
Instead, run the controller-manager standalone with --controllers=csrsigner and
point to the CA certificate and key.
PKI certificates and requirements includes guidance on setting up a cluster to use an external CA.
You can use the check-expiration subcommand to check when certificates expire:
kubeadm certs check-expiration
The output is similar to this:
CERTIFICATE EXPIRES RESIDUAL TIME CERTIFICATE AUTHORITY EXTERNALLY MANAGED
admin.conf Dec 30, 2020 23:36 UTC 364d no
apiserver Dec 30, 2020 23:36 UTC 364d ca no
apiserver-etcd-client Dec 30, 2020 23:36 UTC 364d etcd-ca no
apiserver-kubelet-client Dec 30, 2020 23:36 UTC 364d ca no
controller-manager.conf Dec 30, 2020 23:36 UTC 364d no
etcd-healthcheck-client Dec 30, 2020 23:36 UTC 364d etcd-ca no
etcd-peer Dec 30, 2020 23:36 UTC 364d etcd-ca no
etcd-server Dec 30, 2020 23:36 UTC 364d etcd-ca no
front-proxy-client Dec 30, 2020 23:36 UTC 364d front-proxy-ca no
scheduler.conf Dec 30, 2020 23:36 UTC 364d no
CERTIFICATE AUTHORITY EXPIRES RESIDUAL TIME EXTERNALLY MANAGED
ca Dec 28, 2029 23:36 UTC 9y no
etcd-ca Dec 28, 2029 23:36 UTC 9y no
front-proxy-ca Dec 28, 2029 23:36 UTC 9y no
The command shows expiration/residual time for the client certificates in the /etc/kubernetes/pki folder and for the client certificate embedded in the KUBECONFIG files used by kubeadm (admin.conf, controller-manager.conf and scheduler.conf).
Additionally, kubeadm informs the user if the certificate is externally managed; in this case, the user should take care of managing certificate renewal manually/using other tools.
kubeadm cannot manage certificates signed by an external CA.
kubelet.conf is not included in the list above because kubeadm configures kubelet
for automatic certificate renewal
with rotatable certificates under /var/lib/kubelet/pki.
To repair an expired kubelet client certificate see
Kubelet client certificate rotation fails.
On nodes created with kubeadm init, prior to kubeadm version 1.17, there is a
bug where you manually have to modify the contents of kubelet.conf. After kubeadm init finishes, you should update kubelet.conf to point to the
rotated kubelet client certificates, by replacing client-certificate-data and client-key-data with:
client-certificate: /var/lib/kubelet/pki/kubelet-client-current.pem
client-key: /var/lib/kubelet/pki/kubelet-client-current.pem
kubeadm renews all the certificates during control plane upgrade.
This feature is designed for addressing the simplest use cases; if you don't have specific requirements on certificate renewal and perform Kubernetes version upgrades regularly (less than 1 year in between each upgrade), kubeadm will take care of keeping your cluster up to date and reasonably secure.
If you have more complex requirements for certificate renewal, you can opt out from the default behavior by passing --certificate-renewal=false to kubeadm upgrade apply or to kubeadm upgrade node.
--certificate-renewal is false for the kubeadm upgrade node
command. In that case, you should explicitly set --certificate-renewal=true.
You can renew your certificates manually at any time with the kubeadm certs renew command.
This command performs the renewal using CA (or front-proxy-CA) certificate and key stored in /etc/kubernetes/pki.
After running the command you should restart the control plane Pods. This is required since
dynamic certificate reload is currently not supported for all components and certificates.
Static Pods are managed by the local kubelet
and not by the API Server, thus kubectl cannot be used to delete and restart them.
To restart a static Pod you can temporarily remove its manifest file from /etc/kubernetes/manifests/
and wait for 20 seconds (see the fileCheckFrequency value in KubeletConfiguration struct.
The kubelet will terminate the Pod if it's no longer in the manifest directory.
You can then move the file back and after another fileCheckFrequency period, the kubelet will recreate
the Pod and the certificate renewal for the component can complete.
certs renew uses the existing certificates as the authoritative source for attributes (Common Name, Organization, SAN, etc.) instead of the kubeadm-config ConfigMap. It is strongly recommended to keep them both in sync.
kubeadm certs renew provides the following options:
The Kubernetes certificates normally reach their expiration date after one year.
--csr-only can be used to renew certificates with an external CA by generating certificate signing requests (without actually renewing certificates in place); see next paragraph for more information.
It's also possible to renew a single certificate instead of all.
This section provides more details about how to execute manual certificate renewal using the Kubernetes certificates API.
The Kubernetes Certificate Authority does not work out of the box. You can configure an external signer such as cert-manager, or you can use the built-in signer.
The built-in signer is part of kube-controller-manager.
To activate the built-in signer, you must pass the --cluster-signing-cert-file and --cluster-signing-key-file flags.
If you're creating a new cluster, you can use a kubeadm configuration file:
apiVersion: kubeadm.k8s.io/v1beta3
kind: ClusterConfiguration
controllerManager:
extraArgs:
cluster-signing-cert-file: /etc/kubernetes/pki/ca.crt
cluster-signing-key-file: /etc/kubernetes/pki/ca.key
See Create CertificateSigningRequest for creating CSRs with the Kubernetes API.
This section provide more details about how to execute manual certificate renewal using an external CA.
To better integrate with external CAs, kubeadm can also produce certificate signing requests (CSRs). A CSR represents a request to a CA for a signed certificate for a client. In kubeadm terms, any certificate that would normally be signed by an on-disk CA can be produced as a CSR instead. A CA, however, cannot be produced as a CSR.
You can create certificate signing requests with kubeadm certs renew --csr-only.
Both the CSR and the accompanying private key are given in the output.
You can pass in a directory with --csr-dir to output the CSRs to the specified location.
If --csr-dir is not specified, the default certificate directory (/etc/kubernetes/pki) is used.
Certificates can be renewed with kubeadm certs renew --csr-only.
As with kubeadm init, an output directory can be specified with the --csr-dir flag.
A CSR contains a certificate's name, domains, and IPs, but it does not specify usages. It is the responsibility of the CA to specify the correct cert usages when issuing a certificate.
openssl this is done with the
openssl ca command.cfssl you specify
usages in the config file.After a certificate is signed using your preferred method, the certificate and the private key must be copied to the PKI directory (by default /etc/kubernetes/pki).
Kubeadm does not support rotation or replacement of CA certificates out of the box.
For more information about manual rotation or replacement of CA, see manual rotation of CA certificates.
By default the kubelet serving certificate deployed by kubeadm is self-signed. This means a connection from external services like the metrics-server to a kubelet cannot be secured with TLS.
To configure the kubelets in a new kubeadm cluster to obtain properly signed serving
certificates you must pass the following minimal configuration to kubeadm init:
apiVersion: kubeadm.k8s.io/v1beta3
kind: ClusterConfiguration
---
apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration
serverTLSBootstrap: true
If you have already created the cluster you must adapt it by doing the following:
kubelet-config-1.22 ConfigMap in the kube-system namespace.
In that ConfigMap, the kubelet key has a
KubeletConfiguration
document as its value. Edit the KubeletConfiguration document to set serverTLSBootstrap: true.serverTLSBootstrap: true field in /var/lib/kubelet/config.yaml
and restart the kubelet with systemctl restart kubeletThe field serverTLSBootstrap: true will enable the bootstrap of kubelet serving
certificates by requesting them from the certificates.k8s.io API. One known limitation
is that the CSRs (Certificate Signing Requests) for these certificates cannot be automatically
approved by the default signer in the kube-controller-manager -
kubernetes.io/kubelet-serving.
This will require action from the user or a third party controller.
These CSRs can be viewed using:
kubectl get csr
NAME AGE SIGNERNAME REQUESTOR CONDITION
csr-9wvgt 112s kubernetes.io/kubelet-serving system:node:worker-1 Pending
csr-lz97v 1m58s kubernetes.io/kubelet-serving system:node:control-plane-1 Pending
To approve them you can do the following:
kubectl certificate approve <CSR-name>
By default, these serving certificate will expire after one year. Kubeadm sets the
KubeletConfiguration field rotateCertificates to true, which means that close
to expiration a new set of CSRs for the serving certificates will be created and must
be approved to complete the rotation. To understand more see
Certificate Rotation.
If you are looking for a solution for automatic approval of these CSRs it is recommended that you contact your cloud provider and ask if they have a CSR signer that verifies the node identity with an out of band mechanism.
Third party custom controllers can be used:
Such a controller is not a secure mechanism unless it not only verifies the CommonName in the CSR but also verifies the requested IPs and domain names. This would prevent a malicious actor that has access to a kubelet client certificate to create CSRs requesting serving certificates for any IP or domain name.
This page explains how to configure the kubelet cgroup driver to match the container runtime cgroup driver for kubeadm clusters.
You should be familiar with the Kubernetes container runtime requirements.
The Container runtimes page
explains that the systemd driver is recommended for kubeadm based setups instead
of the cgroupfs driver, because kubeadm manages the kubelet as a systemd service.
The page also provides details on how to setup a number of different container runtimes with the
systemd driver by default.
kubeadm allows you to pass a KubeletConfiguration structure during kubeadm init.
This KubeletConfiguration can include the cgroupDriver field which controls the cgroup
driver of the kubelet.
cgroupDriver field under KubeletConfiguration,
kubeadm will default it to systemd.
A minimal example of configuring the field explicitly:
# kubeadm-config.yaml
kind: ClusterConfiguration
apiVersion: kubeadm.k8s.io/v1beta3
kubernetesVersion: v1.21.0
---
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
cgroupDriver: systemd
Such a configuration file can then be passed to the kubeadm command:
kubeadm init --config kubeadm-config.yaml
Kubeadm uses the same KubeletConfiguration for all nodes in the cluster.
The KubeletConfiguration is stored in a ConfigMap
object under the kube-system namespace.
Executing the sub commands init, join and upgrade would result in kubeadm
writing the KubeletConfiguration as a file under /var/lib/kubelet/config.yaml
and passing it to the local node kubelet.
cgroupfs driverAs this guide explains using the cgroupfs driver with kubeadm is not recommended.
To continue using cgroupfs and to prevent kubeadm upgrade from modifying the
KubeletConfiguration cgroup driver on existing setups, you must be explicit
about its value. This applies to a case where you do not wish future versions
of kubeadm to apply the systemd driver by default.
See the below section on "Modify the kubelet ConfigMap" for details on how to be explicit about the value.
If you wish to configure a container runtime to use the cgroupfs driver,
you must refer to the documentation of the container runtime of your choice.
systemd driverTo change the cgroup driver of an existing kubeadm cluster to systemd in-place,
a similar procedure to a kubelet upgrade is required. This must include both
steps outlined below.
systemd driver. This requires executing only the first step below
before joining the new nodes and ensuring the workloads can safely move to the new
nodes before deleting the old nodes.
Find the kubelet ConfigMap name using kubectl get cm -n kube-system | grep kubelet-config.
Call kubectl edit cm kubelet-config-x.yy -n kube-system (replace x.yy with
the Kubernetes version).
Either modify the existing cgroupDriver value or add a new field that looks like this:
cgroupDriver: systemd
This field must be present under the kubelet: section of the ConfigMap.
For each node in the cluster:
kubectl drain <node-name> --ignore-daemonsetssystemctl stop kubeletsystemdcgroupDriver: systemd in /var/lib/kubelet/config.yamlsystemctl start kubeletkubectl uncordon <node-name>Execute these steps on nodes one at a time to ensure workloads have sufficient time to schedule on different nodes.
Once the process is complete ensure that all nodes and workloads are healthy.
This page explains how to upgrade a Kubernetes cluster created with kubeadm from version
1.21.x to version 1.22.x, and from version
1.22.x to 1.22.y (where y > x). Skipping MINOR versions
when upgrading is unsupported.
To see information about upgrading clusters created using older versions of kubeadm, please refer to following pages instead:
The upgrade workflow at high level is the following:
kubeadm upgrade does not touch your workloads, only components internal to Kubernetes, but backups are always a best practice.Find the latest patch release for Kubernetes 1.22 using the OS package manager:
apt update
apt-cache madison kubeadm
# find the latest 1.22 version in the list
# it should look like 1.22.x-00, where x is the latest patch
yum list --showduplicates kubeadm --disableexcludes=kubernetes
# find the latest 1.22 version in the list
# it should look like 1.22.x-0, where x is the latest patch
The upgrade procedure on control plane nodes should be executed one node at a time.
Pick a control plane node that you wish to upgrade first. It must have the /etc/kubernetes/admin.conf file.
For the first control plane node
# replace x in 1.22.x-00 with the latest patch version
apt-mark unhold kubeadm && \
apt-get update && apt-get install -y kubeadm=1.22.x-00 && \
apt-mark hold kubeadm
-
# since apt-get version 1.1 you can also use the following method
apt-get update && \
apt-get install -y --allow-change-held-packages kubeadm=1.22.x-00
# replace x in 1.22.x-0 with the latest patch version
yum install -y kubeadm-1.22.x-0 --disableexcludes=kubernetes
Verify that the download works and has the expected version:
kubeadm version
Verify the upgrade plan:
kubeadm upgrade plan
This command checks that your cluster can be upgraded, and fetches the versions you can upgrade to. It also shows a table with the component config version states.
kubeadm upgrade also automatically renews the certificates that it manages on this node.
To opt-out of certificate renewal the flag --certificate-renewal=false can be used.
For more information see the certificate management guide.
kubeadm upgrade plan shows any component configs that require manual upgrade, users must provide
a config file with replacement configs to kubeadm upgrade apply via the --config command line flag.
Failing to do so will cause kubeadm upgrade apply to exit with an error and not perform an upgrade.
Choose a version to upgrade to, and run the appropriate command. For example:
# replace x with the patch version you picked for this upgrade
sudo kubeadm upgrade apply v1.22.x
Once the command finishes you should see:
[upgrade/successful] SUCCESS! Your cluster was upgraded to "v1.22.x". Enjoy!
[upgrade/kubelet] Now that your control plane is upgraded, please proceed with upgrading your kubelets if you haven't already done so.
Manually upgrade your CNI provider plugin.
Your Container Network Interface (CNI) provider may have its own upgrade instructions to follow. Check the addons page to find your CNI provider and see whether additional upgrade steps are required.
This step is not required on additional control plane nodes if the CNI provider runs as a DaemonSet.
For the other control plane nodes
Same as the first control plane node but use:
sudo kubeadm upgrade node
instead of:
sudo kubeadm upgrade apply
Also calling kubeadm upgrade plan and upgrading the CNI provider plugin is no longer needed.
Prepare the node for maintenance by marking it unschedulable and evicting the workloads:
# replace <node-to-drain> with the name of your node you are draining
kubectl drain <node-to-drain> --ignore-daemonsets
# replace x in 1.22.x-00 with the latest patch version
apt-mark unhold kubelet kubectl && \
apt-get update && apt-get install -y kubelet=1.22.x-00 kubectl=1.22.x-00 && \
apt-mark hold kubelet kubectl
-
# since apt-get version 1.1 you can also use the following method
apt-get update && \
apt-get install -y --allow-change-held-packages kubelet=1.22.x-00 kubectl=1.22.x-00
# replace x in 1.22.x-0 with the latest patch version
yum install -y kubelet-1.22.x-0 kubectl-1.22.x-0 --disableexcludes=kubernetes
Restart the kubelet:
sudo systemctl daemon-reload
sudo systemctl restart kubelet
Bring the node back online by marking it schedulable:
# replace <node-to-drain> with the name of your node
kubectl uncordon <node-to-drain>
The upgrade procedure on worker nodes should be executed one node at a time or few nodes at a time, without compromising the minimum required capacity for running your workloads.
# replace x in 1.22.x-00 with the latest patch version
apt-mark unhold kubeadm && \
apt-get update && apt-get install -y kubeadm=1.22.x-00 && \
apt-mark hold kubeadm
-
# since apt-get version 1.1 you can also use the following method
apt-get update && \
apt-get install -y --allow-change-held-packages kubeadm=1.22.x-00
# replace x in 1.22.x-0 with the latest patch version
yum install -y kubeadm-1.22.x-0 --disableexcludes=kubernetes
For worker nodes this upgrades the local kubelet configuration:
sudo kubeadm upgrade node
Prepare the node for maintenance by marking it unschedulable and evicting the workloads:
# replace <node-to-drain> with the name of your node you are draining
kubectl drain <node-to-drain> --ignore-daemonsets
# replace x in 1.22.x-00 with the latest patch version
apt-mark unhold kubelet kubectl && \
apt-get update && apt-get install -y kubelet=1.22.x-00 kubectl=1.22.x-00 && \
apt-mark hold kubelet kubectl
-
# since apt-get version 1.1 you can also use the following method
apt-get update && \
apt-get install -y --allow-change-held-packages kubelet=1.22.x-00 kubectl=1.22.x-00
# replace x in 1.22.x-0 with the latest patch version
yum install -y kubelet-1.22.x-0 kubectl-1.22.x-0 --disableexcludes=kubernetes
Restart the kubelet:
sudo systemctl daemon-reload
sudo systemctl restart kubelet
Bring the node back online by marking it schedulable:
# replace <node-to-drain> with the name of your node
kubectl uncordon <node-to-drain>
After the kubelet is upgraded on all nodes verify that all nodes are available again by running the following command from anywhere kubectl can access the cluster:
kubectl get nodes
The STATUS column should show Ready for all your nodes, and the version number should be updated.
If kubeadm upgrade fails and does not roll back, for example because of an unexpected shutdown during execution, you can run kubeadm upgrade again.
This command is idempotent and eventually makes sure that the actual state is the desired state you declare.
To recover from a bad state, you can also run kubeadm upgrade apply --force without changing the version that your cluster is running.
During upgrade kubeadm writes the following backup folders under /etc/kubernetes/tmp:
kubeadm-backup-etcd-<date>-<time>kubeadm-backup-manifests-<date>-<time>kubeadm-backup-etcd contains a backup of the local etcd member data for this control plane Node.
In case of an etcd upgrade failure and if the automatic rollback does not work, the contents of this folder
can be manually restored in /var/lib/etcd. In case external etcd is used this backup folder will be empty.
kubeadm-backup-manifests contains a backup of the static Pod manifest files for this control plane Node.
In case of a upgrade failure and if the automatic rollback does not work, the contents of this folder can be
manually restored in /etc/kubernetes/manifests. If for some reason there is no difference between a pre-upgrade
and post-upgrade manifest file for a certain component, a backup file for it will not be written.
kubeadm upgrade apply does the following:
Ready stateCoreDNS and kube-proxy manifests and makes sure that all necessary RBAC rules are created.kubeadm upgrade node does the following on additional control plane nodes:
ClusterConfiguration from the cluster.kubeadm upgrade node does the following on worker nodes:
ClusterConfiguration from the cluster.Kubernetes v1.18 [beta]
You can use Kubernetes to run a mixture of Linux and Windows nodes, so you can mix Pods that run on Linux on with Pods that run on Windows. This page shows how to register Windows nodes to your cluster.
kubectl version.
Obtain a Windows Server 2019 license (or higher) in order to configure the Windows node that hosts Windows containers. If you are using VXLAN/Overlay networking you must have also have KB4489899 installed.
A Linux-based Kubernetes kubeadm cluster in which you have access to the control plane (see Creating a single control-plane cluster with kubeadm).
Once you have a Linux-based Kubernetes control-plane node you are ready to choose a networking solution. This guide illustrates using Flannel in VXLAN mode for simplicity.
Prepare Kubernetes control plane for Flannel
Some minor preparation is recommended on the Kubernetes control plane in our cluster. It is recommended to enable bridged IPv4 traffic to iptables chains when using Flannel. The following command must be run on all Linux nodes:
sudo sysctl net.bridge.bridge-nf-call-iptables=1
Download & configure Flannel for Linux
Download the most recent Flannel manifest:
wget https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
Modify the net-conf.json section of the flannel manifest in order to set the VNI to 4096 and the Port to 4789. It should look as follows:
net-conf.json: |
{
"Network": "10.244.0.0/16",
"Backend": {
"Type": "vxlan",
"VNI": 4096,
"Port": 4789
}
}
Type to "host-gw" and omit VNI and Port.
Apply the Flannel manifest and validate
Let's apply the Flannel configuration:
kubectl apply -f kube-flannel.yml
After a few minutes, you should see all the pods as running if the Flannel pod network was deployed.
kubectl get pods -n kube-system
The output should include the Linux flannel DaemonSet as running:
NAMESPACE NAME READY STATUS RESTARTS AGE
...
kube-system kube-flannel-ds-54954 1/1 Running 0 1m
Add Windows Flannel and kube-proxy DaemonSets
Now you can add Windows-compatible versions of Flannel and kube-proxy. In order to ensure that you get a compatible version of kube-proxy, you'll need to substitute the tag of the image. The following example shows usage for Kubernetes v1.22.0, but you should adjust the version for your own deployment.
curl -L https://github.com/kubernetes-sigs/sig-windows-tools/releases/latest/download/kube-proxy.yml | sed 's/VERSION/v1.22.0/g' | kubectl apply -f -
kubectl apply -f https://github.com/kubernetes-sigs/sig-windows-tools/releases/latest/download/flannel-overlay.yml
If you're using a different interface rather than Ethernet (i.e. "Ethernet0 2") on the Windows nodes, you have to modify the line:
wins cli process run --path /k/flannel/setup.exe --args "--mode=overlay --interface=Ethernet"
in the flannel-host-gw.yml or flannel-overlay.yml file and specify your interface accordingly.
# Example
curl -L https://github.com/kubernetes-sigs/sig-windows-tools/releases/latest/download/flannel-overlay.yml | sed 's/Ethernet/Ethernet0 2/g' | kubectl apply -f -
Install the Containers feature
Install-WindowsFeature -Name containers
Install Docker Instructions to do so are available at Install Docker Engine - Enterprise on Windows Servers.
curl.exe -LO https://raw.githubusercontent.com/kubernetes-sigs/sig-windows-tools/master/kubeadm/scripts/PrepareNode.ps1
.\PrepareNode.ps1 -KubernetesVersion v1.22.0
kubeadm to join the nodeUse the command that was given to you when you ran kubeadm init on a control plane host.
If you no longer have this command, or the token has expired, you can run kubeadm token create --print-join-command
(on a control plane host) to generate a new token and join command.
curl.exe -LO https://github.com/kubernetes-sigs/sig-windows-tools/releases/latest/download/Install-Containerd.ps1
.\Install-Containerd.ps1
To install a specific version of containerD specify the version with -ContainerDVersion.
# Example
.\Install-Containerd.ps1 -ContainerDVersion 1.4.1
If you're using a different interface rather than Ethernet (i.e. "Ethernet0 2") on the Windows nodes, specify the name with -netAdapterName.
# Example
.\Install-Containerd.ps1 -netAdapterName "Ethernet0 2"
curl.exe -LO https://raw.githubusercontent.com/kubernetes-sigs/sig-windows-tools/master/kubeadm/scripts/PrepareNode.ps1
.\PrepareNode.ps1 -KubernetesVersion v1.22.0 -ContainerRuntime containerD
kubeadm to join the nodeUse the command that was given to you when you ran kubeadm init on a control plane host.
If you no longer have this command, or the token has expired, you can run kubeadm token create --print-join-command
(on a control plane host) to generate a new token and join command.
--cri-socket "npipe:////./pipe/containerd-containerd" to the kubeadm call
You should now be able to view the Windows node in your cluster by running:
kubectl get nodes -o wide
If your new node is in the NotReady state it is likely because the flannel image is still downloading.
You can check the progress as before by checking on the flannel pods in the kube-system namespace:
kubectl -n kube-system get pods -l app=flannel
Once the flannel Pod is running, your node should enter the Ready state and then be available to handle workloads.
Kubernetes v1.18 [beta]
This page explains how to upgrade a Windows node created with kubeadm.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.17. To check the version, enterkubectl version.
From the Windows node, upgrade kubeadm:
# replace v1.22.0 with your desired version
curl.exe -Lo C:\k\kubeadm.exe https://dl.k8s.io//bin/windows/amd64/kubeadm.exe
From a machine with access to the Kubernetes API, prepare the node for maintenance by marking it unschedulable and evicting the workloads:
# replace <node-to-drain> with the name of your node you are draining
kubectl drain <node-to-drain> --ignore-daemonsets
You should see output similar to this:
node/ip-172-31-85-18 cordoned
node/ip-172-31-85-18 drained
From the Windows node, call the following command to sync new kubelet configuration:
kubeadm upgrade node
From the Windows node, upgrade and restart the kubelet:
stop-service kubelet
curl.exe -Lo C:\k\kubelet.exe https://dl.k8s.io//bin/windows/amd64/kubelet.exe
restart-service kubelet
From a machine with access to the Kubernetes API, bring the node back online by marking it schedulable:
# replace <node-to-drain> with the name of your node
kubectl uncordon <node-to-drain>
From a machine with access to the Kubernetes API, run the following, again replacing v1.22.0 with your desired version:
curl -L https://github.com/kubernetes-sigs/sig-windows-tools/releases/latest/download/kube-proxy.yml | sed 's/VERSION/v1.22.0/g' | kubectl apply -f -
This section presents information you need to know when migrating from dockershim to other container runtimes.
Since the announcement of dockershim deprecation in Kubernetes 1.20, there were questions on how this will affect various workloads and Kubernetes installations. You can find this blog post useful to understand the problem better: Dockershim Deprecation FAQ
It is recommended to migrate from dockershim to alternative container runtimes. Check out container runtimes section to know your options. Make sure to report issues you encountered with the migration. So the issue can be fixed in a timely manner and your cluster would be ready for dockershim removal.
This page outlines steps to find out what container runtime the nodes in your cluster use.
Depending on the way you run your cluster, the container runtime for the nodes may
have been pre-configured or you need to configure it. If you're using a managed
Kubernetes service, there might be vendor-specific ways to check what container runtime is
configured for the nodes. The method described on this page should work whenever
the execution of kubectl is allowed.
Install and configure kubectl. See Install Tools section for details.
Use kubectl to fetch and show node information:
kubectl get nodes -o wide
The output is similar to the following. The column CONTAINER-RUNTIME outputs
the runtime and its version.
# For dockershim
NAME STATUS VERSION CONTAINER-RUNTIME
node-1 Ready v1.16.15 docker://19.3.1
node-2 Ready v1.16.15 docker://19.3.1
node-3 Ready v1.16.15 docker://19.3.1
# For containerd
NAME STATUS VERSION CONTAINER-RUNTIME
node-1 Ready v1.19.6 containerd://1.4.1
node-2 Ready v1.19.6 containerd://1.4.1
node-3 Ready v1.19.6 containerd://1.4.1
Find out more information about container runtimes on Container Runtimes page.
The dockershim component of Kubernetes allows to use Docker as a Kubernetes's
container runtime.
Kubernetes' built-in dockershim component was deprecated in release v1.20.
This page explains how your cluster could be using Docker as a container runtime,
provides details on the role that dockershim plays when in use, and shows steps
you can take to check whether any workloads could be affected by dockershim deprecation.
If you are using Docker for building your application containers, you can still run these containers on any container runtime. This use of Docker does not count as a dependency on Docker as a container runtime.
When alternative container runtime is used, executing Docker commands may either not work or yield unexpected output. This is how you can find whether you have a dependency on Docker:
docker ps),
restart the Docker service (commands such as systemctl restart docker.service),
or modify Docker-specific files such as /etc/docker/daemon.json./etc/docker/daemon.json). Those typically need to
be reconfigured for another container runtime.A container runtime is software that can execute the containers that make up a Kubernetes pod. Kubernetes is responsible for orchestration and scheduling of Pods; on each node, the kubelet uses the container runtime interface as an abstraction so that you can use any compatible container runtime.
In its earliest releases, Kubernetes offered compatibility with one container runtime: Docker.
Later in the Kubernetes project's history, cluster operators wanted to adopt additional container runtimes.
The CRI was designed to allow this kind of flexibility - and the kubelet began supporting CRI. However,
because Docker existed before the CRI specification was invented, the Kubernetes project created an
adapter component, dockershim. The dockershim adapter allows the kubelet to interact with Docker as
if Docker were a CRI compatible runtime.
You can read about it in Kubernetes Containerd integration goes GA blog post.
Switching to Containerd as a container runtime eliminates the middleman. All the same containers can be run by container runtimes like Containerd as before. But now, since containers schedule directly with the container runtime, they are not visible to Docker. So any Docker tooling or fancy UI you might have used before to check on these containers is no longer available.
You cannot get container information using docker ps or docker inspect
commands. As you cannot list containers, you cannot get logs, stop containers,
or execute something inside container using docker exec.
You can still pull images or build them using docker build command. But images
built or pulled by Docker would not be visible to container runtime and
Kubernetes. They needed to be pushed to some registry to allow them to be used
by Kubernetes.
With Kubernetes 1.20 dockershim was deprecated. From the Dockershim Deprecation FAQ you might already know that most apps do not have a direct dependency on runtime hosting containers. However, there are still a lot of telemetry and security agents that has a dependency on docker to collect containers metadata, logs and metrics. This document aggregates information on how to detect these dependencies and links on how to migrate these agents to use generic tools or alternative runtimes.
There are a few ways agents may run on Kubernetes cluster. Agents may run on nodes directly or as DaemonSets.
Historically, Kubernetes was built on top of Docker. Kubernetes is managing networking and scheduling, Docker was placing and operating containers on a node. So you can get scheduling-related metadata like a pod name from Kubernetes and containers state information from Docker. Over time more runtimes were created to manage containers. Also there are projects and Kubernetes features that generalize container status information extraction across many runtimes.
Some agents are tied specifically to the Docker tool. The agents may run
commands like docker ps
or docker top to list
containers and processes or docker logs
to subscribe on docker logs. With the deprecating of Docker as a container runtime,
these commands will not work any longer.
If a pod wants to make calls to the dockerd running on the node, the pod must either:
For example: on COS images, Docker exposes its Unix domain socket at
/var/run/docker.sock This means that the pod spec will include a
hostPath volume mount of /var/run/docker.sock.
Here's a sample shell script to find Pods that have a mount directly mapping the
Docker socket. This script outputs the namespace and name of the pod. You can
remove the grep /var/run/docker.sock to review other mounts.
kubectl get pods --all-namespaces \
-o=jsonpath='{range .items[*]}{"\n"}{.metadata.namespace}{":\t"}{.metadata.name}{":\t"}{range .spec.volumes[*]}{.hostPath.path}{", "}{end}{end}' \
| sort \
| grep '/var/run/docker.sock'
/var/run may be mounted instead of the full path (like in this
example).
The script above only detects the most common uses.
In case your cluster nodes are customized and install additional security and telemetry agents on the node, make sure to check with the vendor of the agent whether it has dependency on Docker.
We keep the work in progress version of migration instructions for various telemetry and security agent vendors in Google doc. Please contact the vendor to get up to date instructions for migrating from dockershim.
When using client certificate authentication, you can generate certificates
manually through easyrsa, openssl or cfssl.
easyrsa can manually generate certificates for your cluster.
Download, unpack, and initialize the patched version of easyrsa3.
curl -LO https://storage.googleapis.com/kubernetes-release/easy-rsa/easy-rsa.tar.gz
tar xzf easy-rsa.tar.gz
cd easy-rsa-master/easyrsa3
./easyrsa init-pki
Generate a new certificate authority (CA). --batch sets automatic mode;
--req-cn specifies the Common Name (CN) for the CA's new root certificate.
./easyrsa --batch "--req-cn=${MASTER_IP}@`date +%s`" build-ca nopass
Generate server certificate and key.
The argument --subject-alt-name sets the possible IPs and DNS names the API server will
be accessed with. The MASTER_CLUSTER_IP is usually the first IP from the service CIDR
that is specified as the --service-cluster-ip-range argument for both the API server and
the controller manager component. The argument --days is used to set the number of days
after which the certificate expires.
The sample below also assumes that you are using cluster.local as the default
DNS domain name.
./easyrsa --subject-alt-name="IP:${MASTER_IP},"\
"IP:${MASTER_CLUSTER_IP},"\
"DNS:kubernetes,"\
"DNS:kubernetes.default,"\
"DNS:kubernetes.default.svc,"\
"DNS:kubernetes.default.svc.cluster,"\
"DNS:kubernetes.default.svc.cluster.local" \
--days=10000 \
build-server-full server nopass
Copy pki/ca.crt, pki/issued/server.crt, and pki/private/server.key to your directory.
Fill in and add the following parameters into the API server start parameters:
--client-ca-file=/yourdirectory/ca.crt
--tls-cert-file=/yourdirectory/server.crt
--tls-private-key-file=/yourdirectory/server.key
openssl can manually generate certificates for your cluster.
Generate a ca.key with 2048bit:
openssl genrsa -out ca.key 2048
According to the ca.key generate a ca.crt (use -days to set the certificate effective time):
openssl req -x509 -new -nodes -key ca.key -subj "/CN=${MASTER_IP}" -days 10000 -out ca.crt
Generate a server.key with 2048bit:
openssl genrsa -out server.key 2048
Create a config file for generating a Certificate Signing Request (CSR).
Be sure to substitute the values marked with angle brackets (e.g. <MASTER_IP>)
with real values before saving this to a file (e.g. csr.conf).
Note that the value for MASTER_CLUSTER_IP is the service cluster IP for the
API server as described in previous subsection.
The sample below also assumes that you are using cluster.local as the default
DNS domain name.
[ req ]
default_bits = 2048
prompt = no
default_md = sha256
req_extensions = req_ext
distinguished_name = dn
[ dn ]
C = <country>
ST = <state>
L = <city>
O = <organization>
OU = <organization unit>
CN = <MASTER_IP>
[ req_ext ]
subjectAltName = @alt_names
[ alt_names ]
DNS.1 = kubernetes
DNS.2 = kubernetes.default
DNS.3 = kubernetes.default.svc
DNS.4 = kubernetes.default.svc.cluster
DNS.5 = kubernetes.default.svc.cluster.local
IP.1 = <MASTER_IP>
IP.2 = <MASTER_CLUSTER_IP>
[ v3_ext ]
authorityKeyIdentifier=keyid,issuer:always
basicConstraints=CA:FALSE
keyUsage=keyEncipherment,dataEncipherment
extendedKeyUsage=serverAuth,clientAuth
subjectAltName=@alt_names
Generate the certificate signing request based on the config file:
openssl req -new -key server.key -out server.csr -config csr.conf
Generate the server certificate using the ca.key, ca.crt and server.csr:
openssl x509 -req -in server.csr -CA ca.crt -CAkey ca.key \
-CAcreateserial -out server.crt -days 10000 \
-extensions v3_ext -extfile csr.conf
View the certificate signing request:
openssl req -noout -text -in ./server.csr
View the certificate:
openssl x509 -noout -text -in ./server.crt
Finally, add the same parameters into the API server start parameters.
cfssl is another tool for certificate generation.
Download, unpack and prepare the command line tools as shown below. Note that you may need to adapt the sample commands based on the hardware architecture and cfssl version you are using.
curl -L https://github.com/cloudflare/cfssl/releases/download/v1.5.0/cfssl_1.5.0_linux_amd64 -o cfssl
chmod +x cfssl
curl -L https://github.com/cloudflare/cfssl/releases/download/v1.5.0/cfssljson_1.5.0_linux_amd64 -o cfssljson
chmod +x cfssljson
curl -L https://github.com/cloudflare/cfssl/releases/download/v1.5.0/cfssl-certinfo_1.5.0_linux_amd64 -o cfssl-certinfo
chmod +x cfssl-certinfo
Create a directory to hold the artifacts and initialize cfssl:
mkdir cert
cd cert
../cfssl print-defaults config > config.json
../cfssl print-defaults csr > csr.json
Create a JSON config file for generating the CA file, for example, ca-config.json:
{
"signing": {
"default": {
"expiry": "8760h"
},
"profiles": {
"kubernetes": {
"usages": [
"signing",
"key encipherment",
"server auth",
"client auth"
],
"expiry": "8760h"
}
}
}
}
Create a JSON config file for CA certificate signing request (CSR), for example,
ca-csr.json. Be sure to replace the values marked with angle brackets with
real values you want to use.
{
"CN": "kubernetes",
"key": {
"algo": "rsa",
"size": 2048
},
"names":[{
"C": "<country>",
"ST": "<state>",
"L": "<city>",
"O": "<organization>",
"OU": "<organization unit>"
}]
}
Generate CA key (ca-key.pem) and certificate (ca.pem):
../cfssl gencert -initca ca-csr.json | ../cfssljson -bare ca
Create a JSON config file for generating keys and certificates for the API
server, for example, server-csr.json. Be sure to replace the values in angle brackets with
real values you want to use. The MASTER_CLUSTER_IP is the service cluster
IP for the API server as described in previous subsection.
The sample below also assumes that you are using cluster.local as the default
DNS domain name.
{
"CN": "kubernetes",
"hosts": [
"127.0.0.1",
"<MASTER_IP>",
"<MASTER_CLUSTER_IP>",
"kubernetes",
"kubernetes.default",
"kubernetes.default.svc",
"kubernetes.default.svc.cluster",
"kubernetes.default.svc.cluster.local"
],
"key": {
"algo": "rsa",
"size": 2048
},
"names": [{
"C": "<country>",
"ST": "<state>",
"L": "<city>",
"O": "<organization>",
"OU": "<organization unit>"
}]
}
Generate the key and certificate for the API server, which are by default
saved into file server-key.pem and server.pem respectively:
../cfssl gencert -ca=ca.pem -ca-key=ca-key.pem \
--config=ca-config.json -profile=kubernetes \
server-csr.json | ../cfssljson -bare server
A client node may refuse to recognize a self-signed CA certificate as valid. For a non-production deployment, or for a deployment that runs behind a company firewall, you can distribute a self-signed CA certificate to all clients and refresh the local list for valid certificates.
On each client, perform the following operations:
sudo cp ca.crt /usr/local/share/ca-certificates/kubernetes.crt
sudo update-ca-certificates
Updating certificates in /etc/ssl/certs...
1 added, 0 removed; done.
Running hooks in /etc/ca-certificates/update.d....
done.
You can use the certificates.k8s.io API to provision
x509 certificates to use for authentication as documented
here.
This page shows how to configure default memory requests and limits for a namespace. If a Container is created in a namespace that has a default memory limit, and the Container does not specify its own memory limit, then the Container is assigned the default memory limit. Kubernetes assigns a default memory request under certain conditions that are explained later in this topic.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Each node in your cluster must have at least 2 GiB of memory.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace default-mem-example
Here's the configuration file for a LimitRange object. The configuration specifies a default memory request and a default memory limit.
apiVersion: v1
kind: LimitRange
metadata:
name: mem-limit-range
spec:
limits:
- default:
memory: 512Mi
defaultRequest:
memory: 256Mi
type: Container
Create the LimitRange in the default-mem-example namespace:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-defaults.yaml --namespace=default-mem-example
Now if a Container is created in the default-mem-example namespace, and the Container does not specify its own values for memory request and memory limit, the Container is given a default memory request of 256 MiB and a default memory limit of 512 MiB.
Here's the configuration file for a Pod that has one Container. The Container does not specify a memory request and limit.
apiVersion: v1
kind: Pod
metadata:
name: default-mem-demo
spec:
containers:
- name: default-mem-demo-ctr
image: nginx
Create the Pod.
kubectl apply -f https://k8s.io/examples/admin/resource/memory-defaults-pod.yaml --namespace=default-mem-example
View detailed information about the Pod:
kubectl get pod default-mem-demo --output=yaml --namespace=default-mem-example
The output shows that the Pod's Container has a memory request of 256 MiB and a memory limit of 512 MiB. These are the default values specified by the LimitRange.
containers:
- image: nginx
imagePullPolicy: Always
name: default-mem-demo-ctr
resources:
limits:
memory: 512Mi
requests:
memory: 256Mi
Delete your Pod:
kubectl delete pod default-mem-demo --namespace=default-mem-example
Here's the configuration file for a Pod that has one Container. The Container specifies a memory limit, but not a request:
apiVersion: v1
kind: Pod
metadata:
name: default-mem-demo-2
spec:
containers:
- name: default-mem-demo-2-ctr
image: nginx
resources:
limits:
memory: "1Gi"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-defaults-pod-2.yaml --namespace=default-mem-example
View detailed information about the Pod:
kubectl get pod default-mem-demo-2 --output=yaml --namespace=default-mem-example
The output shows that the Container's memory request is set to match its memory limit. Notice that the Container was not assigned the default memory request value of 256Mi.
resources:
limits:
memory: 1Gi
requests:
memory: 1Gi
Here's the configuration file for a Pod that has one Container. The Container specifies a memory request, but not a limit:
apiVersion: v1
kind: Pod
metadata:
name: default-mem-demo-3
spec:
containers:
- name: default-mem-demo-3-ctr
image: nginx
resources:
requests:
memory: "128Mi"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-defaults-pod-3.yaml --namespace=default-mem-example
View the Pod's specification:
kubectl get pod default-mem-demo-3 --output=yaml --namespace=default-mem-example
The output shows that the Container's memory request is set to the value specified in the Container's configuration file. The Container's memory limit is set to 512Mi, which is the default memory limit for the namespace.
resources:
limits:
memory: 512Mi
requests:
memory: 128Mi
If your namespace has a resource quota, it is helpful to have a default value in place for memory limit. Here are two of the restrictions that a resource quota imposes on a namespace:
If a Container does not specify its own memory limit, it is given the default limit, and then it can be allowed to run in a namespace that is restricted by a quota.
Delete your namespace:
kubectl delete namespace default-mem-example
Configure Minimum and Maximum Memory Constraints for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
This page shows how to configure default CPU requests and limits for a namespace. A Kubernetes cluster can be divided into namespaces. If a Container is created in a namespace that has a default CPU limit, and the Container does not specify its own CPU limit, then the Container is assigned the default CPU limit. Kubernetes assigns a default CPU request under certain conditions that are explained later in this topic.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace default-cpu-example
Here's the configuration file for a LimitRange object. The configuration specifies a default CPU request and a default CPU limit.
apiVersion: v1
kind: LimitRange
metadata:
name: cpu-limit-range
spec:
limits:
- default:
cpu: 1
defaultRequest:
cpu: 0.5
type: Container
Create the LimitRange in the default-cpu-example namespace:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-defaults.yaml --namespace=default-cpu-example
Now if a Container is created in the default-cpu-example namespace, and the Container does not specify its own values for CPU request and CPU limit, the Container is given a default CPU request of 0.5 and a default CPU limit of 1.
Here's the configuration file for a Pod that has one Container. The Container does not specify a CPU request and limit.
apiVersion: v1
kind: Pod
metadata:
name: default-cpu-demo
spec:
containers:
- name: default-cpu-demo-ctr
image: nginx
Create the Pod.
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-defaults-pod.yaml --namespace=default-cpu-example
View the Pod's specification:
kubectl get pod default-cpu-demo --output=yaml --namespace=default-cpu-example
The output shows that the Pod's Container has a CPU request of 500 millicpus and a CPU limit of 1 cpu. These are the default values specified by the LimitRange.
containers:
- image: nginx
imagePullPolicy: Always
name: default-cpu-demo-ctr
resources:
limits:
cpu: "1"
requests:
cpu: 500m
Here's the configuration file for a Pod that has one Container. The Container specifies a CPU limit, but not a request:
apiVersion: v1
kind: Pod
metadata:
name: default-cpu-demo-2
spec:
containers:
- name: default-cpu-demo-2-ctr
image: nginx
resources:
limits:
cpu: "1"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-defaults-pod-2.yaml --namespace=default-cpu-example
View the Pod specification:
kubectl get pod default-cpu-demo-2 --output=yaml --namespace=default-cpu-example
The output shows that the Container's CPU request is set to match its CPU limit. Notice that the Container was not assigned the default CPU request value of 0.5 cpu.
resources:
limits:
cpu: "1"
requests:
cpu: "1"
Here's the configuration file for a Pod that has one Container. The Container specifies a CPU request, but not a limit:
apiVersion: v1
kind: Pod
metadata:
name: default-cpu-demo-3
spec:
containers:
- name: default-cpu-demo-3-ctr
image: nginx
resources:
requests:
cpu: "0.75"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-defaults-pod-3.yaml --namespace=default-cpu-example
View the Pod specification:
kubectl get pod default-cpu-demo-3 --output=yaml --namespace=default-cpu-example
The output shows that the Container's CPU request is set to the value specified in the Container's configuration file. The Container's CPU limit is set to 1 cpu, which is the default CPU limit for the namespace.
resources:
limits:
cpu: "1"
requests:
cpu: 750m
If your namespace has a resource quota, it is helpful to have a default value in place for CPU limit. Here are two of the restrictions that a resource quota imposes on a namespace:
If a Container does not specify its own CPU limit, it is given the default limit, and then it can be allowed to run in a namespace that is restricted by a quota.
Delete your namespace:
kubectl delete namespace default-cpu-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum Memory Constraints for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
This page shows how to set minimum and maximum values for memory used by Containers running in a namespace. You specify minimum and maximum memory values in a LimitRange object. If a Pod does not meet the constraints imposed by the LimitRange, it cannot be created in the namespace.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Each node in your cluster must have at least 1 GiB of memory.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace constraints-mem-example
Here's the configuration file for a LimitRange:
apiVersion: v1
kind: LimitRange
metadata:
name: mem-min-max-demo-lr
spec:
limits:
- max:
memory: 1Gi
min:
memory: 500Mi
type: Container
Create the LimitRange:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-constraints.yaml --namespace=constraints-mem-example
View detailed information about the LimitRange:
kubectl get limitrange mem-min-max-demo-lr --namespace=constraints-mem-example --output=yaml
The output shows the minimum and maximum memory constraints as expected. But notice that even though you didn't specify default values in the configuration file for the LimitRange, they were created automatically.
limits:
- default:
memory: 1Gi
defaultRequest:
memory: 1Gi
max:
memory: 1Gi
min:
memory: 500Mi
type: Container
Now whenever a Container is created in the constraints-mem-example namespace, Kubernetes performs these steps:
If the Container does not specify its own memory request and limit, assign the default memory request and limit to the Container.
Verify that the Container has a memory request that is greater than or equal to 500 MiB.
Verify that the Container has a memory limit that is less than or equal to 1 GiB.
Here's the configuration file for a Pod that has one Container. The Container manifest specifies a memory request of 600 MiB and a memory limit of 800 MiB. These satisfy the minimum and maximum memory constraints imposed by the LimitRange.
apiVersion: v1
kind: Pod
metadata:
name: constraints-mem-demo
spec:
containers:
- name: constraints-mem-demo-ctr
image: nginx
resources:
limits:
memory: "800Mi"
requests:
memory: "600Mi"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-constraints-pod.yaml --namespace=constraints-mem-example
Verify that the Pod's Container is running:
kubectl get pod constraints-mem-demo --namespace=constraints-mem-example
View detailed information about the Pod:
kubectl get pod constraints-mem-demo --output=yaml --namespace=constraints-mem-example
The output shows that the Container has a memory request of 600 MiB and a memory limit of 800 MiB. These satisfy the constraints imposed by the LimitRange.
resources:
limits:
memory: 800Mi
requests:
memory: 600Mi
Delete your Pod:
kubectl delete pod constraints-mem-demo --namespace=constraints-mem-example
Here's the configuration file for a Pod that has one Container. The Container specifies a memory request of 800 MiB and a memory limit of 1.5 GiB.
apiVersion: v1
kind: Pod
metadata:
name: constraints-mem-demo-2
spec:
containers:
- name: constraints-mem-demo-2-ctr
image: nginx
resources:
limits:
memory: "1.5Gi"
requests:
memory: "800Mi"
Attempt to create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-constraints-pod-2.yaml --namespace=constraints-mem-example
The output shows that the Pod does not get created, because the Container specifies a memory limit that is too large:
Error from server (Forbidden): error when creating "examples/admin/resource/memory-constraints-pod-2.yaml":
pods "constraints-mem-demo-2" is forbidden: maximum memory usage per Container is 1Gi, but limit is 1536Mi.
Here's the configuration file for a Pod that has one Container. The Container specifies a memory request of 100 MiB and a memory limit of 800 MiB.
apiVersion: v1
kind: Pod
metadata:
name: constraints-mem-demo-3
spec:
containers:
- name: constraints-mem-demo-3-ctr
image: nginx
resources:
limits:
memory: "800Mi"
requests:
memory: "100Mi"
Attempt to create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-constraints-pod-3.yaml --namespace=constraints-mem-example
The output shows that the Pod does not get created, because the Container specifies a memory request that is too small:
Error from server (Forbidden): error when creating "examples/admin/resource/memory-constraints-pod-3.yaml":
pods "constraints-mem-demo-3" is forbidden: minimum memory usage per Container is 500Mi, but request is 100Mi.
Here's the configuration file for a Pod that has one Container. The Container does not specify a memory request, and it does not specify a memory limit.
apiVersion: v1
kind: Pod
metadata:
name: constraints-mem-demo-4
spec:
containers:
- name: constraints-mem-demo-4-ctr
image: nginx
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/memory-constraints-pod-4.yaml --namespace=constraints-mem-example
View detailed information about the Pod:
kubectl get pod constraints-mem-demo-4 --namespace=constraints-mem-example --output=yaml
The output shows that the Pod's Container has a memory request of 1 GiB and a memory limit of 1 GiB. How did the Container get those values?
resources:
limits:
memory: 1Gi
requests:
memory: 1Gi
Because your Container did not specify its own memory request and limit, it was given the default memory request and limit from the LimitRange.
At this point, your Container might be running or it might not be running. Recall that a prerequisite for this task is that your Nodes have at least 1 GiB of memory. If each of your Nodes has only 1 GiB of memory, then there is not enough allocatable memory on any Node to accommodate a memory request of 1 GiB. If you happen to be using Nodes with 2 GiB of memory, then you probably have enough space to accommodate the 1 GiB request.
Delete your Pod:
kubectl delete pod constraints-mem-demo-4 --namespace=constraints-mem-example
The maximum and minimum memory constraints imposed on a namespace by a LimitRange are enforced only when a Pod is created or updated. If you change the LimitRange, it does not affect Pods that were created previously.
As a cluster administrator, you might want to impose restrictions on the amount of memory that Pods can use. For example:
Each Node in a cluster has 2 GB of memory. You do not want to accept any Pod that requests more than 2 GB of memory, because no Node in the cluster can support the request.
A cluster is shared by your production and development departments. You want to allow production workloads to consume up to 8 GB of memory, but you want development workloads to be limited to 512 MB. You create separate namespaces for production and development, and you apply memory constraints to each namespace.
Delete your namespace:
kubectl delete namespace constraints-mem-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
This page shows how to set minimum and maximum values for the CPU resources used by Containers and Pods in a namespace. You specify minimum and maximum CPU values in a LimitRange object. If a Pod does not meet the constraints imposed by the LimitRange, it cannot be created in the namespace.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Your cluster must have at least 1 CPU available for use to run the task examples.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace constraints-cpu-example
Here's the configuration file for a LimitRange:
apiVersion: v1
kind: LimitRange
metadata:
name: cpu-min-max-demo-lr
spec:
limits:
- max:
cpu: "800m"
min:
cpu: "200m"
type: Container
Create the LimitRange:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-constraints.yaml --namespace=constraints-cpu-example
View detailed information about the LimitRange:
kubectl get limitrange cpu-min-max-demo-lr --output=yaml --namespace=constraints-cpu-example
The output shows the minimum and maximum CPU constraints as expected. But notice that even though you didn't specify default values in the configuration file for the LimitRange, they were created automatically.
limits:
- default:
cpu: 800m
defaultRequest:
cpu: 800m
max:
cpu: 800m
min:
cpu: 200m
type: Container
Now whenever a Container is created in the constraints-cpu-example namespace, Kubernetes performs these steps:
If the Container does not specify its own CPU request and limit, assign the default CPU request and limit to the Container.
Verify that the Container specifies a CPU request that is greater than or equal to 200 millicpu.
Verify that the Container specifies a CPU limit that is less than or equal to 800 millicpu.
LimitRange object, you can specify limits on huge-pages
or GPUs as well. However, when both default and defaultRequest are specified
on these resources, the two values must be the same.
Here's the configuration file for a Pod that has one Container. The Container manifest specifies a CPU request of 500 millicpu and a CPU limit of 800 millicpu. These satisfy the minimum and maximum CPU constraints imposed by the LimitRange.
apiVersion: v1
kind: Pod
metadata:
name: constraints-cpu-demo
spec:
containers:
- name: constraints-cpu-demo-ctr
image: nginx
resources:
limits:
cpu: "800m"
requests:
cpu: "500m"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-constraints-pod.yaml --namespace=constraints-cpu-example
Verify that the Pod's Container is running:
kubectl get pod constraints-cpu-demo --namespace=constraints-cpu-example
View detailed information about the Pod:
kubectl get pod constraints-cpu-demo --output=yaml --namespace=constraints-cpu-example
The output shows that the Container has a CPU request of 500 millicpu and CPU limit of 800 millicpu. These satisfy the constraints imposed by the LimitRange.
resources:
limits:
cpu: 800m
requests:
cpu: 500m
kubectl delete pod constraints-cpu-demo --namespace=constraints-cpu-example
Here's the configuration file for a Pod that has one Container. The Container specifies a CPU request of 500 millicpu and a cpu limit of 1.5 cpu.
apiVersion: v1
kind: Pod
metadata:
name: constraints-cpu-demo-2
spec:
containers:
- name: constraints-cpu-demo-2-ctr
image: nginx
resources:
limits:
cpu: "1.5"
requests:
cpu: "500m"
Attempt to create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-constraints-pod-2.yaml --namespace=constraints-cpu-example
The output shows that the Pod does not get created, because the Container specifies a CPU limit that is too large:
Error from server (Forbidden): error when creating "examples/admin/resource/cpu-constraints-pod-2.yaml":
pods "constraints-cpu-demo-2" is forbidden: maximum cpu usage per Container is 800m, but limit is 1500m.
Here's the configuration file for a Pod that has one Container. The Container specifies a CPU request of 100 millicpu and a CPU limit of 800 millicpu.
apiVersion: v1
kind: Pod
metadata:
name: constraints-cpu-demo-3
spec:
containers:
- name: constraints-cpu-demo-3-ctr
image: nginx
resources:
limits:
cpu: "800m"
requests:
cpu: "100m"
Attempt to create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-constraints-pod-3.yaml --namespace=constraints-cpu-example
The output shows that the Pod does not get created, because the Container specifies a CPU request that is too small:
Error from server (Forbidden): error when creating "examples/admin/resource/cpu-constraints-pod-3.yaml":
pods "constraints-cpu-demo-3" is forbidden: minimum cpu usage per Container is 200m, but request is 100m.
Here's the configuration file for a Pod that has one Container. The Container does not specify a CPU request, and it does not specify a CPU limit.
apiVersion: v1
kind: Pod
metadata:
name: constraints-cpu-demo-4
spec:
containers:
- name: constraints-cpu-demo-4-ctr
image: vish/stress
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/cpu-constraints-pod-4.yaml --namespace=constraints-cpu-example
View detailed information about the Pod:
kubectl get pod constraints-cpu-demo-4 --namespace=constraints-cpu-example --output=yaml
The output shows that the Pod's Container has a CPU request of 800 millicpu and a CPU limit of 800 millicpu. How did the Container get those values?
resources:
limits:
cpu: 800m
requests:
cpu: 800m
Because your Container did not specify its own CPU request and limit, it was given the default CPU request and limit from the LimitRange.
At this point, your Container might be running or it might not be running. Recall that a prerequisite for this task is that your cluster must have at least 1 CPU available for use. If each of your Nodes has only 1 CPU, then there might not be enough allocatable CPU on any Node to accommodate a request of 800 millicpu. If you happen to be using Nodes with 2 CPU, then you probably have enough CPU to accommodate the 800 millicpu request.
Delete your Pod:
kubectl delete pod constraints-cpu-demo-4 --namespace=constraints-cpu-example
The maximum and minimum CPU constraints imposed on a namespace by a LimitRange are enforced only when a Pod is created or updated. If you change the LimitRange, it does not affect Pods that were created previously.
As a cluster administrator, you might want to impose restrictions on the CPU resources that Pods can use. For example:
Each Node in a cluster has 2 CPU. You do not want to accept any Pod that requests more than 2 CPU, because no Node in the cluster can support the request.
A cluster is shared by your production and development departments. You want to allow production workloads to consume up to 3 CPU, but you want development workloads to be limited to 1 CPU. You create separate namespaces for production and development, and you apply CPU constraints to each namespace.
Delete your namespace:
kubectl delete namespace constraints-cpu-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum Memory Constraints for a Namespace
This page shows how to set quotas for the total amount memory and CPU that can be used by all Containers running in a namespace. You specify quotas in a ResourceQuota object.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Each node in your cluster must have at least 1 GiB of memory.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace quota-mem-cpu-example
Here is the configuration file for a ResourceQuota object:
apiVersion: v1
kind: ResourceQuota
metadata:
name: mem-cpu-demo
spec:
hard:
requests.cpu: "1"
requests.memory: 1Gi
limits.cpu: "2"
limits.memory: 2Gi
Create the ResourceQuota:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-mem-cpu.yaml --namespace=quota-mem-cpu-example
View detailed information about the ResourceQuota:
kubectl get resourcequota mem-cpu-demo --namespace=quota-mem-cpu-example --output=yaml
The ResourceQuota places these requirements on the quota-mem-cpu-example namespace:
Here is the configuration file for a Pod:
apiVersion: v1
kind: Pod
metadata:
name: quota-mem-cpu-demo
spec:
containers:
- name: quota-mem-cpu-demo-ctr
image: nginx
resources:
limits:
memory: "800Mi"
cpu: "800m"
requests:
memory: "600Mi"
cpu: "400m"
Create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-mem-cpu-pod.yaml --namespace=quota-mem-cpu-example
Verify that the Pod's Container is running:
kubectl get pod quota-mem-cpu-demo --namespace=quota-mem-cpu-example
Once again, view detailed information about the ResourceQuota:
kubectl get resourcequota mem-cpu-demo --namespace=quota-mem-cpu-example --output=yaml
The output shows the quota along with how much of the quota has been used. You can see that the memory and CPU requests and limits for your Pod do not exceed the quota.
status:
hard:
limits.cpu: "2"
limits.memory: 2Gi
requests.cpu: "1"
requests.memory: 1Gi
used:
limits.cpu: 800m
limits.memory: 800Mi
requests.cpu: 400m
requests.memory: 600Mi
Here is the configuration file for a second Pod:
apiVersion: v1
kind: Pod
metadata:
name: quota-mem-cpu-demo-2
spec:
containers:
- name: quota-mem-cpu-demo-2-ctr
image: redis
resources:
limits:
memory: "1Gi"
cpu: "800m"
requests:
memory: "700Mi"
cpu: "400m"
In the configuration file, you can see that the Pod has a memory request of 700 MiB. Notice that the sum of the used memory request and this new memory request exceeds the memory request quota. 600 MiB + 700 MiB > 1 GiB.
Attempt to create the Pod:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-mem-cpu-pod-2.yaml --namespace=quota-mem-cpu-example
The second Pod does not get created. The output shows that creating the second Pod would cause the memory request total to exceed the memory request quota.
Error from server (Forbidden): error when creating "examples/admin/resource/quota-mem-cpu-pod-2.yaml":
pods "quota-mem-cpu-demo-2" is forbidden: exceeded quota: mem-cpu-demo,
requested: requests.memory=700Mi,used: requests.memory=600Mi, limited: requests.memory=1Gi
As you have seen in this exercise, you can use a ResourceQuota to restrict the memory request total for all Containers running in a namespace. You can also restrict the totals for memory limit, cpu request, and cpu limit.
If you want to restrict individual Containers, instead of totals for all Containers, use a LimitRange.
Delete your namespace:
kubectl delete namespace quota-mem-cpu-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum Memory Constraints for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
This page shows how to set a quota for the total number of Pods that can run in a namespace. You specify quotas in a ResourceQuota object.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace quota-pod-example
Here is the configuration file for a ResourceQuota object:
apiVersion: v1
kind: ResourceQuota
metadata:
name: pod-demo
spec:
hard:
pods: "2"
Create the ResourceQuota:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-pod.yaml --namespace=quota-pod-example
View detailed information about the ResourceQuota:
kubectl get resourcequota pod-demo --namespace=quota-pod-example --output=yaml
The output shows that the namespace has a quota of two Pods, and that currently there are no Pods; that is, none of the quota is used.
spec:
hard:
pods: "2"
status:
hard:
pods: "2"
used:
pods: "0"
Here is the configuration file for a Deployment:
apiVersion: apps/v1
kind: Deployment
metadata:
name: pod-quota-demo
spec:
selector:
matchLabels:
purpose: quota-demo
replicas: 3
template:
metadata:
labels:
purpose: quota-demo
spec:
containers:
- name: pod-quota-demo
image: nginx
In the configuration file, replicas: 3 tells Kubernetes to attempt to create three Pods, all running the same application.
Create the Deployment:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-pod-deployment.yaml --namespace=quota-pod-example
View detailed information about the Deployment:
kubectl get deployment pod-quota-demo --namespace=quota-pod-example --output=yaml
The output shows that even though the Deployment specifies three replicas, only two Pods were created because of the quota.
spec:
...
replicas: 3
...
status:
availableReplicas: 2
...
lastUpdateTime: 2017-07-07T20:57:05Z
message: 'unable to create pods: pods "pod-quota-demo-1650323038-" is forbidden:
exceeded quota: pod-demo, requested: pods=1, used: pods=2, limited: pods=2'
Delete your namespace:
kubectl delete namespace quota-pod-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum Memory Constraints for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
This page shows how to install and use Antrea CNI plugin on Kubernetes. For background on Project Antrea, read the Introduction to Antrea.
You need to have a Kubernetes cluster. Follow the kubeadm getting started guide to bootstrap one.
Follow Getting Started guide to deploy Antrea for kubeadm.
Once your cluster is running, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy.
This page shows a couple of quick ways to create a Calico cluster on Kubernetes.
Decide whether you want to deploy a cloud or local cluster.
Prerequisite: gcloud.
To launch a GKE cluster with Calico, include the --enable-network-policy flag.
Syntax
gcloud container clusters create [CLUSTER_NAME] --enable-network-policy
Example
gcloud container clusters create my-calico-cluster --enable-network-policy
To verify the deployment, use the following command.
kubectl get pods --namespace=kube-system
The Calico pods begin with calico. Check to make sure each one has a status of Running.
To get a local single-host Calico cluster in fifteen minutes using kubeadm, refer to the Calico Quickstart.
Once your cluster is running, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy.
This page shows how to use Cilium for NetworkPolicy.
For background on Cilium, read the Introduction to Cilium.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
To get familiar with Cilium easily you can follow the Cilium Kubernetes Getting Started Guide to perform a basic DaemonSet installation of Cilium in minikube.
To start minikube, minimal version required is >= v1.5.2, run the with the following arguments:
minikube version
minikube version: v1.5.2
minikube start --network-plugin=cni
For minikube you can install Cilium using its CLI tool. Cilium will automatically detect the cluster configuration and will install the appropriate components for a successful installation:
curl -LO https://github.com/cilium/cilium-cli/releases/latest/download/cilium-linux-amd64.tar.gz
sudo tar xzvfC cilium-linux-amd64.tar.gz /usr/local/bin
rm cilium-linux-amd64.tar.gz
cilium install
🔮 Auto-detected Kubernetes kind: minikube
✨ Running "minikube" validation checks
✅ Detected minikube version "1.20.0"
ℹ️ Cilium version not set, using default version "v1.10.0"
🔮 Auto-detected cluster name: minikube
🔮 Auto-detected IPAM mode: cluster-pool
🔮 Auto-detected datapath mode: tunnel
🔑 Generating CA...
2021/05/27 02:54:44 [INFO] generate received request
2021/05/27 02:54:44 [INFO] received CSR
2021/05/27 02:54:44 [INFO] generating key: ecdsa-256
2021/05/27 02:54:44 [INFO] encoded CSR
2021/05/27 02:54:44 [INFO] signed certificate with serial number 48713764918856674401136471229482703021230538642
🔑 Generating certificates for Hubble...
2021/05/27 02:54:44 [INFO] generate received request
2021/05/27 02:54:44 [INFO] received CSR
2021/05/27 02:54:44 [INFO] generating key: ecdsa-256
2021/05/27 02:54:44 [INFO] encoded CSR
2021/05/27 02:54:44 [INFO] signed certificate with serial number 3514109734025784310086389188421560613333279574
🚀 Creating Service accounts...
🚀 Creating Cluster roles...
🚀 Creating ConfigMap...
🚀 Creating Agent DaemonSet...
🚀 Creating Operator Deployment...
⌛ Waiting for Cilium to be installed...
The remainder of the Getting Started Guide explains how to enforce both L3/L4 (i.e., IP address + port) security policies, as well as L7 (e.g., HTTP) security policies using an example application.
For detailed instructions around deploying Cilium for production, see: Cilium Kubernetes Installation Guide This documentation includes detailed requirements, instructions and example production DaemonSet files.
Deploying a cluster with Cilium adds Pods to the kube-system namespace. To see
this list of Pods run:
kubectl get pods --namespace=kube-system -l k8s-app=cilium
You'll see a list of Pods similar to this:
NAME READY STATUS RESTARTS AGE
cilium-kkdhz 1/1 Running 0 3m23s
...
A cilium Pod runs on each node in your cluster and enforces network policy
on the traffic to/from Pods on that node using Linux BPF.
Once your cluster is running, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy with Cilium. Have fun, and if you have questions, contact us using the Cilium Slack Channel.
This page shows how to use Kube-router for NetworkPolicy.
You need to have a Kubernetes cluster running. If you do not already have a cluster, you can create one by using any of the cluster installers like Kops, Bootkube, Kubeadm etc.
The Kube-router Addon comes with a Network Policy Controller that watches Kubernetes API server for any NetworkPolicy and pods updated and configures iptables rules and ipsets to allow or block traffic as directed by the policies. Please follow the trying Kube-router with cluster installers guide to install Kube-router addon.
Once you have installed the Kube-router addon, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy.
This page shows how to use Romana for NetworkPolicy.
Complete steps 1, 2, and 3 of the kubeadm getting started guide.
Follow the containerized installation guide for kubeadm.
To apply network policies use one of the following:
Once you have installed Romana, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy.
This page shows how to use Weave Net for NetworkPolicy.
You need to have a Kubernetes cluster. Follow the kubeadm getting started guide to bootstrap one.
Follow the Integrating Kubernetes via the Addon guide.
The Weave Net addon for Kubernetes comes with a
Network Policy Controller
that automatically monitors Kubernetes for any NetworkPolicy annotations on all
namespaces and configures iptables rules to allow or block traffic as directed by the policies.
Verify that the weave works.
Enter the following command:
kubectl get pods -n kube-system -o wide
The output is similar to this:
NAME READY STATUS RESTARTS AGE IP NODE
weave-net-1t1qg 2/2 Running 0 9d 192.168.2.10 worknode3
weave-net-231d7 2/2 Running 1 7d 10.2.0.17 worknodegpu
weave-net-7nmwt 2/2 Running 3 9d 192.168.2.131 masternode
weave-net-pmw8w 2/2 Running 0 9d 192.168.2.216 worknode2
Each Node has a weave Pod, and all Pods are Running and 2/2 READY. (2/2 means that each Pod has weave and weave-npc.)
Once you have installed the Weave Net addon, you can follow the Declare Network Policy to try out Kubernetes NetworkPolicy. If you have any question, contact us at #weave-community on Slack or Weave User Group.
This page shows how to access clusters using the Kubernetes API.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
When accessing the Kubernetes API for the first time, use the
Kubernetes command-line tool, kubectl.
To access a cluster, you need to know the location of the cluster and have credentials to access it. Typically, this is automatically set-up when you work through a Getting started guide, or someone else setup the cluster and provided you with credentials and a location.
Check the location and credentials that kubectl knows about with this command:
kubectl config view
Many of the examples provide an introduction to using kubectl. Complete documentation is found in the kubectl manual.
kubectl handles locating and authenticating to the API server. If you want to directly access the REST API with an http client like
curl or wget, or a browser, there are multiple ways you can locate and authenticate against the API server:
Using the Go or Python client libraries provides accessing kubectl in proxy mode.
The following command runs kubectl in a mode where it acts as a reverse proxy. It handles locating the API server and authenticating.
Run it like this:
kubectl proxy --port=8080 &
See kubectl proxy for more details.
Then you can explore the API with curl, wget, or a browser, like so:
curl http://localhost:8080/api/
The output is similar to this:
{
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
It is possible to avoid using kubectl proxy by passing an authentication token directly to the API server, like this:
Using grep/cut approach:
# Check all possible clusters, as your .KUBECONFIG may have multiple contexts:
kubectl config view -o jsonpath='{"Cluster name\tServer\n"}{range .clusters[*]}{.name}{"\t"}{.cluster.server}{"\n"}{end}'
# Select name of cluster you want to interact with from above output:
export CLUSTER_NAME="some_server_name"
# Point to the API server referring the cluster name
APISERVER=$(kubectl config view -o jsonpath="{.clusters[?(@.name==\"$CLUSTER_NAME\")].cluster.server}")
# Gets the token value
TOKEN=$(kubectl get secrets -o jsonpath="{.items[?(@.metadata.annotations['kubernetes\.io/service-account\.name']=='default')].data.token}"|base64 --decode)
# Explore the API with TOKEN
curl -X GET $APISERVER/api --header "Authorization: Bearer $TOKEN" --insecure
The output is similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
Using jsonpath approach:
APISERVER=$(kubectl config view --minify -o jsonpath='{.clusters[0].cluster.server}')
TOKEN=$(kubectl get secret $(kubectl get serviceaccount default -o jsonpath='{.secrets[0].name}') -o jsonpath='{.data.token}' | base64 --decode )
curl $APISERVER/api --header "Authorization: Bearer $TOKEN" --insecure
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
The above example uses the --insecure flag. This leaves it subject to MITM
attacks. When kubectl accesses the cluster it uses a stored root certificate
and client certificates to access the server. (These are installed in the
~/.kube directory). Since cluster certificates are typically self-signed, it
may take special configuration to get your http client to use root
certificate.
On some clusters, the API server does not require authentication; it may serve on localhost, or be protected by a firewall. There is not a standard for this. Controlling Access to the Kubernetes API describes how you can configure this as a cluster administrator.
Kubernetes officially supports client libraries for Go, Python, Java, dotnet, Javascript, and Haskell. There are other client libraries that are provided and maintained by their authors, not the Kubernetes team. See client libraries for accessing the API from other languages and how they authenticate.
go get k8s.io/client-go@kubernetes-<kubernetes-version-number> See https://github.com/kubernetes/client-go/releases to see which versions are supported.import "k8s.io/client-go/kubernetes" is correct.
The Go client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
package main
import (
"context"
"fmt"
"k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/client-go/kubernetes"
"k8s.io/client-go/tools/clientcmd"
)
func main() {
// uses the current context in kubeconfig
// path-to-kubeconfig -- for example, /root/.kube/config
config, _ := clientcmd.BuildConfigFromFlags("", "<path-to-kubeconfig>")
// creates the clientset
clientset, _ := kubernetes.NewForConfig(config)
// access the API to list pods
pods, _ := clientset.CoreV1().Pods("").List(context.TODO(), v1.ListOptions{})
fmt.Printf("There are %d pods in the cluster\n", len(pods.Items))
}
If the application is deployed as a Pod in the cluster, see Accessing the API from within a Pod.
To use Python client, run the following command: pip install kubernetes See Python Client Library page for more installation options.
The Python client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
from kubernetes import client, config
config.load_kube_config()
v1=client.CoreV1Api()
print("Listing pods with their IPs:")
ret = v1.list_pod_for_all_namespaces(watch=False)
for i in ret.items:
print("%s\t%s\t%s" % (i.status.pod_ip, i.metadata.namespace, i.metadata.name))
To install the Java Client, run:
# Clone java library
git clone --recursive https://github.com/kubernetes-client/java
# Installing project artifacts, POM etc:
cd java
mvn install
See https://github.com/kubernetes-client/java/releases to see which versions are supported.
The Java client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
package io.kubernetes.client.examples;
import io.kubernetes.client.ApiClient;
import io.kubernetes.client.ApiException;
import io.kubernetes.client.Configuration;
import io.kubernetes.client.apis.CoreV1Api;
import io.kubernetes.client.models.V1Pod;
import io.kubernetes.client.models.V1PodList;
import io.kubernetes.client.util.ClientBuilder;
import io.kubernetes.client.util.KubeConfig;
import java.io.FileReader;
import java.io.IOException;
/**
* A simple example of how to use the Java API from an application outside a kubernetes cluster
*
* <p>Easiest way to run this: mvn exec:java
* -Dexec.mainClass="io.kubernetes.client.examples.KubeConfigFileClientExample"
*
*/
public class KubeConfigFileClientExample {
public static void main(String[] args) throws IOException, ApiException {
// file path to your KubeConfig
String kubeConfigPath = "~/.kube/config";
// loading the out-of-cluster config, a kubeconfig from file-system
ApiClient client =
ClientBuilder.kubeconfig(KubeConfig.loadKubeConfig(new FileReader(kubeConfigPath))).build();
// set the global default api-client to the in-cluster one from above
Configuration.setDefaultApiClient(client);
// the CoreV1Api loads default api-client from global configuration.
CoreV1Api api = new CoreV1Api();
// invokes the CoreV1Api client
V1PodList list = api.listPodForAllNamespaces(null, null, null, null, null, null, null, null, null);
System.out.println("Listing all pods: ");
for (V1Pod item : list.getItems()) {
System.out.println(item.getMetadata().getName());
}
}
}
To use dotnet client, run the following command: dotnet add package KubernetesClient --version 1.6.1 See dotnet Client Library page for more installation options. See https://github.com/kubernetes-client/csharp/releases to see which versions are supported.
The dotnet client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
using System;
using k8s;
namespace simple
{
internal class PodList
{
private static void Main(string[] args)
{
var config = KubernetesClientConfiguration.BuildDefaultConfig();
IKubernetes client = new Kubernetes(config);
Console.WriteLine("Starting Request!");
var list = client.ListNamespacedPod("default");
foreach (var item in list.Items)
{
Console.WriteLine(item.Metadata.Name);
}
if (list.Items.Count == 0)
{
Console.WriteLine("Empty!");
}
}
}
}
To install JavaScript client, run the following command: npm install @kubernetes/client-node. See https://github.com/kubernetes-client/javascript/releases to see which versions are supported.
The JavaScript client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
const k8s = require('@kubernetes/client-node');
const kc = new k8s.KubeConfig();
kc.loadFromDefault();
const k8sApi = kc.makeApiClient(k8s.CoreV1Api);
k8sApi.listNamespacedPod('default').then((res) => {
console.log(res.body);
});
See https://github.com/kubernetes-client/haskell/releases to see which versions are supported.
The Haskell client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the API server. See this example:
exampleWithKubeConfig :: IO ()
exampleWithKubeConfig = do
oidcCache <- atomically $ newTVar $ Map.fromList []
(mgr, kcfg) <- mkKubeClientConfig oidcCache $ KubeConfigFile "/path/to/kubeconfig"
dispatchMime
mgr
kcfg
(CoreV1.listPodForAllNamespaces (Accept MimeJSON))
>>= print
This page shows how to connect to services running on the Kubernetes cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In Kubernetes, nodes, pods and services all have their own IPs. In many cases, the node IPs, pod IPs, and some service IPs on a cluster will not be routable, so they will not be reachable from a machine outside the cluster, such as your desktop machine.
You have several options for connecting to nodes, pods and services from outside the cluster:
NodePort or LoadBalancer to make the service reachable outside
the cluster. See the services and
kubectl expose documentation.Typically, there are several services which are started on a cluster by kube-system. Get a list of these
with the kubectl cluster-info command:
kubectl cluster-info
The output is similar to this:
Kubernetes master is running at https://104.197.5.247
elasticsearch-logging is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy
kibana-logging is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/kibana-logging/proxy
kube-dns is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/kube-dns/proxy
grafana is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/monitoring-grafana/proxy
heapster is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/monitoring-heapster/proxy
This shows the proxy-verb URL for accessing each service.
For example, this cluster has cluster-level logging enabled (using Elasticsearch), which can be reached
at https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/ if suitable credentials are passed, or through a kubectl proxy at, for example:
http://localhost:8080/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/.
As mentioned above, you use the kubectl cluster-info command to retrieve the service's proxy URL. To create proxy URLs that include service endpoints, suffixes, and parameters, you append to the service's proxy URL:
http://kubernetes_master_address/api/v1/namespaces/namespace_name/services/[https:]service_name[:port_name]/proxy
If you haven't specified a name for your port, you don't have to specify port_name in the URL.
To access the Elasticsearch service endpoint _search?q=user:kimchy, you would use:
http://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/_search?q=user:kimchy
To access the Elasticsearch cluster health information _cluster/health?pretty=true, you would use:
https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/_cluster/health?pretty=true
The health information is similar to this:
{
"cluster_name" : "kubernetes_logging",
"status" : "yellow",
"timed_out" : false,
"number_of_nodes" : 1,
"number_of_data_nodes" : 1,
"active_primary_shards" : 5,
"active_shards" : 5,
"relocating_shards" : 0,
"initializing_shards" : 0,
"unassigned_shards" : 5
}
To access the https Elasticsearch service health information _cluster/health?pretty=true, you would use:
https://104.197.5.247/api/v1/namespaces/kube-system/services/https:elasticsearch-logging/proxy/_cluster/health?pretty=true
You may be able to put an apiserver proxy URL into the address bar of a browser. However:
This page shows how to specify extended resources for a Node. Extended resources allow cluster administrators to advertise node-level resources that would otherwise be unknown to Kubernetes.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
kubectl get nodes
Choose one of your Nodes to use for this exercise.
To advertise a new extended resource on a Node, send an HTTP PATCH request to the Kubernetes API server. For example, suppose one of your Nodes has four dongles attached. Here's an example of a PATCH request that advertises four dongle resources for your Node.
PATCH /api/v1/nodes/<your-node-name>/status HTTP/1.1
Accept: application/json
Content-Type: application/json-patch+json
Host: k8s-master:8080
[
{
"op": "add",
"path": "/status/capacity/example.com~1dongle",
"value": "4"
}
]
Note that Kubernetes does not need to know what a dongle is or what a dongle is for. The preceding PATCH request tells Kubernetes that your Node has four things that you call dongles.
Start a proxy, so that you can easily send requests to the Kubernetes API server:
kubectl proxy
In another command window, send the HTTP PATCH request.
Replace <your-node-name> with the name of your Node:
curl --header "Content-Type: application/json-patch+json" \
--request PATCH \
--data '[{"op": "add", "path": "/status/capacity/example.com~1dongle", "value": "4"}]' \
http://localhost:8001/api/v1/nodes/<your-node-name>/status
~1 is the encoding for the character / in
the patch path. The operation path value in JSON-Patch is interpreted as a
JSON-Pointer. For more details, see
IETF RFC 6901, section 3.
The output shows that the Node has a capacity of 4 dongles:
"capacity": {
"cpu": "2",
"memory": "2049008Ki",
"example.com/dongle": "4",
Describe your Node:
kubectl describe node <your-node-name>
Once again, the output shows the dongle resource:
Capacity:
cpu: 2
memory: 2049008Ki
example.com/dongle: 4
Now, application developers can create Pods that request a certain number of dongles. See Assign Extended Resources to a Container.
Extended resources are similar to memory and CPU resources. For example, just as a Node has a certain amount of memory and CPU to be shared by all components running on the Node, it can have a certain number of dongles to be shared by all components running on the Node. And just as application developers can create Pods that request a certain amount of memory and CPU, they can create Pods that request a certain number of dongles.
Extended resources are opaque to Kubernetes; Kubernetes does not know anything about what they are. Kubernetes knows only that a Node has a certain number of them. Extended resources must be advertised in integer amounts. For example, a Node can advertise four dongles, but not 4.5 dongles.
Suppose a Node has 800 GiB of a special kind of disk storage. You could create a name for the special storage, say example.com/special-storage. Then you could advertise it in chunks of a certain size, say 100 GiB. In that case, your Node would advertise that it has eight resources of type example.com/special-storage.
Capacity:
...
example.com/special-storage: 8
If you want to allow arbitrary requests for special storage, you could advertise special storage in chunks of size 1 byte. In that case, you would advertise 800Gi resources of type example.com/special-storage.
Capacity:
...
example.com/special-storage: 800Gi
Then a Container could request any number of bytes of special storage, up to 800Gi.
Here is a PATCH request that removes the dongle advertisement from a Node.
PATCH /api/v1/nodes/<your-node-name>/status HTTP/1.1
Accept: application/json
Content-Type: application/json-patch+json
Host: k8s-master:8080
[
{
"op": "remove",
"path": "/status/capacity/example.com~1dongle",
}
]
Start a proxy, so that you can easily send requests to the Kubernetes API server:
kubectl proxy
In another command window, send the HTTP PATCH request.
Replace <your-node-name> with the name of your Node:
curl --header "Content-Type: application/json-patch+json" \
--request PATCH \
--data '[{"op": "remove", "path": "/status/capacity/example.com~1dongle"}]' \
http://localhost:8001/api/v1/nodes/<your-node-name>/status
Verify that the dongle advertisement has been removed:
kubectl describe node <your-node-name> | grep dongle
(you should not see any output)
This page shows how to enable and configure autoscaling of the DNS service in your Kubernetes cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
This guide assumes your nodes use the AMD64 or Intel 64 CPU architecture.
Make sure Kubernetes DNS is enabled.
List the Deployments in your cluster in the kube-system namespace:
kubectl get deployment --namespace=kube-system
The output is similar to this:
NAME READY UP-TO-DATE AVAILABLE AGE
...
dns-autoscaler 1/1 1 1 ...
...
If you see "dns-autoscaler" in the output, DNS horizontal autoscaling is already enabled, and you can skip to Tuning autoscaling parameters.
List the DNS deployments in your cluster in the kube-system namespace:
kubectl get deployment -l k8s-app=kube-dns --namespace=kube-system
The output is similar to this:
NAME READY UP-TO-DATE AVAILABLE AGE
...
coredns 2/2 2 2 ...
...
If you don't see a Deployment for DNS services, you can also look for it by name:
kubectl get deployment --namespace=kube-system
and look for a deployment named coredns or kube-dns.
Your scale target is
Deployment/<your-deployment-name>
where <your-deployment-name> is the name of your DNS Deployment. For example, if
the name of your Deployment for DNS is coredns, your scale target is Deployment/coredns.
k8s-app=kube-dns so that it can work in clusters that originally used
kube-dns.
In this section, you create a new Deployment. The Pods in the Deployment run a
container based on the cluster-proportional-autoscaler-amd64 image.
Create a file named dns-horizontal-autoscaler.yaml with this content:
kind: ServiceAccount
apiVersion: v1
metadata:
name: kube-dns-autoscaler
namespace: kube-system
---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: system:kube-dns-autoscaler
rules:
- apiGroups: [""]
resources: ["nodes"]
verbs: ["list", "watch"]
- apiGroups: [""]
resources: ["replicationcontrollers/scale"]
verbs: ["get", "update"]
- apiGroups: ["apps"]
resources: ["deployments/scale", "replicasets/scale"]
verbs: ["get", "update"]
# Remove the configmaps rule once below issue is fixed:
# kubernetes-incubator/cluster-proportional-autoscaler#16
- apiGroups: [""]
resources: ["configmaps"]
verbs: ["get", "create"]
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: system:kube-dns-autoscaler
subjects:
- kind: ServiceAccount
name: kube-dns-autoscaler
namespace: kube-system
roleRef:
kind: ClusterRole
name: system:kube-dns-autoscaler
apiGroup: rbac.authorization.k8s.io
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: kube-dns-autoscaler
namespace: kube-system
labels:
k8s-app: kube-dns-autoscaler
kubernetes.io/cluster-service: "true"
spec:
selector:
matchLabels:
k8s-app: kube-dns-autoscaler
template:
metadata:
labels:
k8s-app: kube-dns-autoscaler
spec:
priorityClassName: system-cluster-critical
securityContext:
seccompProfile:
type: RuntimeDefault
supplementalGroups: [ 65534 ]
fsGroup: 65534
nodeSelector:
kubernetes.io/os: linux
containers:
- name: autoscaler
image: k8s.gcr.io/cpa/cluster-proportional-autoscaler:1.8.4
resources:
requests:
cpu: "20m"
memory: "10Mi"
command:
- /cluster-proportional-autoscaler
- --namespace=kube-system
- --configmap=kube-dns-autoscaler
# Should keep target in sync with cluster/addons/dns/kube-dns.yaml.base
- --target=<SCALE_TARGET>
# When cluster is using large nodes(with more cores), "coresPerReplica" should dominate.
# If using small nodes, "nodesPerReplica" should dominate.
- --default-params={"linear":{"coresPerReplica":256,"nodesPerReplica":16,"preventSinglePointFailure":true,"includeUnschedulableNodes":true}}
- --logtostderr=true
- --v=2
tolerations:
- key: "CriticalAddonsOnly"
operator: "Exists"
serviceAccountName: kube-dns-autoscaler
In the file, replace <SCALE_TARGET> with your scale target.
Go to the directory that contains your configuration file, and enter this command to create the Deployment:
kubectl apply -f dns-horizontal-autoscaler.yaml
The output of a successful command is:
deployment.apps/dns-autoscaler created
DNS horizontal autoscaling is now enabled.
Verify that the dns-autoscaler ConfigMap exists:
kubectl get configmap --namespace=kube-system
The output is similar to this:
NAME DATA AGE
...
dns-autoscaler 1 ...
...
Modify the data in the ConfigMap:
kubectl edit configmap dns-autoscaler --namespace=kube-system
Look for this line:
linear: '{"coresPerReplica":256,"min":1,"nodesPerReplica":16}'
Modify the fields according to your needs. The "min" field indicates the minimal number of DNS backends. The actual number of backends is calculated using this equation:
replicas = max( ceil( cores × 1/coresPerReplica ) , ceil( nodes × 1/nodesPerReplica ) )
Note that the values of both coresPerReplica and nodesPerReplica are
floats.
The idea is that when a cluster is using nodes that have many cores,
coresPerReplica dominates. When a cluster is using nodes that have fewer
cores, nodesPerReplica dominates.
There are other supported scaling patterns. For details, see cluster-proportional-autoscaler.
There are a few options for tuning DNS horizontal autoscaling. Which option to use depends on different conditions.
This option works for all situations. Enter this command:
kubectl scale deployment --replicas=0 dns-autoscaler --namespace=kube-system
The output is:
deployment.apps/dns-autoscaler scaled
Verify that the replica count is zero:
kubectl get rs --namespace=kube-system
The output displays 0 in the DESIRED and CURRENT columns:
NAME DESIRED CURRENT READY AGE
...
dns-autoscaler-6b59789fc8 0 0 0 ...
...
This option works if dns-autoscaler is under your own control, which means no one will re-create it:
kubectl delete deployment dns-autoscaler --namespace=kube-system
The output is:
deployment.apps "dns-autoscaler" deleted
This option works if dns-autoscaler is under control of the (deprecated) Addon Manager, and you have write access to the master node.
Sign in to the master node and delete the corresponding manifest file. The common path for this dns-autoscaler is:
/etc/kubernetes/addons/dns-horizontal-autoscaler/dns-horizontal-autoscaler.yaml
After the manifest file is deleted, the Addon Manager will delete the dns-autoscaler Deployment.
The cluster-proportional-autoscaler application is deployed separately from the DNS service.
An autoscaler Pod runs a client that polls the Kubernetes API server for the number of nodes and cores in the cluster.
A desired replica count is calculated and applied to the DNS backends based on the current schedulable nodes and cores and the given scaling parameters.
The scaling parameters and data points are provided via a ConfigMap to the autoscaler, and it refreshes its parameters table every poll interval to be up to date with the latest desired scaling parameters.
Changes to the scaling parameters are allowed without rebuilding or restarting the autoscaler Pod.
The autoscaler provides a controller interface to support two control patterns: linear and ladder.
This page shows how to change the default Storage Class that is used to provision volumes for PersistentVolumeClaims that have no special requirements.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Depending on the installation method, your Kubernetes cluster may be deployed with an existing StorageClass that is marked as default. This default StorageClass is then used to dynamically provision storage for PersistentVolumeClaims that do not require any specific storage class. See PersistentVolumeClaim documentation for details.
The pre-installed default StorageClass may not fit well with your expected workload; for example, it might provision storage that is too expensive. If this is the case, you can either change the default StorageClass or disable it completely to avoid dynamic provisioning of storage.
Deleting the default StorageClass may not work, as it may be re-created automatically by the addon manager running in your cluster. Please consult the docs for your installation for details about addon manager and how to disable individual addons.
List the StorageClasses in your cluster:
kubectl get storageclass
The output is similar to this:
NAME PROVISIONER AGE
standard (default) kubernetes.io/gce-pd 1d
gold kubernetes.io/gce-pd 1d
The default StorageClass is marked by (default).
Mark the default StorageClass as non-default:
The default StorageClass has an annotation
storageclass.kubernetes.io/is-default-class set to true. Any other value
or absence of the annotation is interpreted as false.
To mark a StorageClass as non-default, you need to change its value to false:
kubectl patch storageclass standard -p '{"metadata": {"annotations":{"storageclass.kubernetes.io/is-default-class":"false"}}}'
where standard is the name of your chosen StorageClass.
Mark a StorageClass as default:
Similar to the previous step, you need to add/set the annotation
storageclass.kubernetes.io/is-default-class=true.
kubectl patch storageclass gold -p '{"metadata": {"annotations":{"storageclass.kubernetes.io/is-default-class":"true"}}}'
Please note that at most one StorageClass can be marked as default. If two
or more of them are marked as default, a PersistentVolumeClaim without storageClassName explicitly specified cannot be created.
Verify that your chosen StorageClass is default:
kubectl get storageclass
The output is similar to this:
NAME PROVISIONER AGE
standard kubernetes.io/gce-pd 1d
gold (default) kubernetes.io/gce-pd 1d
This page shows how to change the reclaim policy of a Kubernetes PersistentVolume.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
PersistentVolumes can have various reclaim policies, including "Retain", "Recycle", and "Delete". For dynamically provisioned PersistentVolumes, the default reclaim policy is "Delete". This means that a dynamically provisioned volume is automatically deleted when a user deletes the corresponding PersistentVolumeClaim. This automatic behavior might be inappropriate if the volume contains precious data. In that case, it is more appropriate to use the "Retain" policy. With the "Retain" policy, if a user deletes a PersistentVolumeClaim, the corresponding PersistentVolume will not be deleted. Instead, it is moved to the Released phase, where all of its data can be manually recovered.
List the PersistentVolumes in your cluster:
kubectl get pv
The output is similar to this:
NAME CAPACITY ACCESSMODES RECLAIMPOLICY STATUS CLAIM STORAGECLASS REASON AGE
pvc-b6efd8da-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Delete Bound default/claim1 manual 10s
pvc-b95650f8-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Delete Bound default/claim2 manual 6s
pvc-bb3ca71d-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Delete Bound default/claim3 manual 3s
This list also includes the name of the claims that are bound to each volume for easier identification of dynamically provisioned volumes.
Choose one of your PersistentVolumes and change its reclaim policy:
kubectl patch pv <your-pv-name> -p '{"spec":{"persistentVolumeReclaimPolicy":"Retain"}}'
where <your-pv-name> is the name of your chosen PersistentVolume.
On Windows, you must double quote any JSONPath template that contains spaces (not single quote as shown above for bash). This in turn means that you must use a single quote or escaped double quote around any literals in the template. For example:
kubectl patch pv <your-pv-name> -p "{\"spec\":{\"persistentVolumeReclaimPolicy\":\"Retain\"}}"
Verify that your chosen PersistentVolume has the right policy:
kubectl get pv
The output is similar to this:
NAME CAPACITY ACCESSMODES RECLAIMPOLICY STATUS CLAIM STORAGECLASS REASON AGE
pvc-b6efd8da-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Delete Bound default/claim1 manual 40s
pvc-b95650f8-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Delete Bound default/claim2 manual 36s
pvc-bb3ca71d-b7b5-11e6-9d58-0ed433a7dd94 4Gi RWO Retain Bound default/claim3 manual 33s
In the preceding output, you can see that the volume bound to claim
default/claim3 has reclaim policy Retain. It will not be automatically
deleted when a user deletes claim default/claim3.
.spec.persistentVolumeReclaimPolicy field of PersistentVolume.Kubernetes v1.11 [beta]
Since cloud providers develop and release at a different pace compared to the Kubernetes project, abstracting the provider-specific code to the cloud-controller-manager binary allows cloud vendors to evolve independently from the core Kubernetes code.
The cloud-controller-manager can be linked to any cloud provider that satisfies cloudprovider.Interface. For backwards compatibility, the cloud-controller-manager provided in the core Kubernetes project uses the same cloud libraries as kube-controller-manager. Cloud providers already supported in Kubernetes core are expected to use the in-tree cloud-controller-manager to transition out of Kubernetes core.
Every cloud has their own set of requirements for running their own cloud provider integration, it should not be too different from the requirements when running kube-controller-manager. As a general rule of thumb you'll need:
Successfully running cloud-controller-manager requires some changes to your cluster configuration.
kube-apiserver and kube-controller-manager MUST NOT specify the --cloud-provider flag. This ensures that it does not run any cloud specific loops that would be run by cloud controller manager. In the future, this flag will be deprecated and removed.kubelet must run with --cloud-provider=external. This is to ensure that the kubelet is aware that it must be initialized by the cloud controller manager before it is scheduled any work.Keep in mind that setting up your cluster to use cloud controller manager will change your cluster behaviour in a few ways:
--cloud-provider=external will add a taint node.cloudprovider.kubernetes.io/uninitialized with an effect NoSchedule during initialization. This marks the node as needing a second initialization from an external controller before it can be scheduled work. Note that in the event that cloud controller manager is not available, new nodes in the cluster will be left unschedulable. The taint is important since the scheduler may require cloud specific information about nodes such as their region or type (high cpu, gpu, high memory, spot instance, etc).The cloud controller manager can implement:
If you are using a cloud that is currently supported in Kubernetes core and would like to adopt cloud controller manager, see the cloud controller manager in kubernetes core.
For cloud controller managers not in Kubernetes core, you can find the respective projects in repositories maintained by cloud vendors or by SIGs.
For providers already in Kubernetes core, you can run the in-tree cloud controller manager as a DaemonSet in your cluster, use the following as a guideline:
# This is an example of how to setup cloud-controller-manager as a Daemonset in your cluster.
# It assumes that your masters can run pods and has the role node-role.kubernetes.io/master
# Note that this Daemonset will not work straight out of the box for your cloud, this is
# meant to be a guideline.
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: cloud-controller-manager
namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: system:cloud-controller-manager
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: cluster-admin
subjects:
- kind: ServiceAccount
name: cloud-controller-manager
namespace: kube-system
---
apiVersion: apps/v1
kind: DaemonSet
metadata:
labels:
k8s-app: cloud-controller-manager
name: cloud-controller-manager
namespace: kube-system
spec:
selector:
matchLabels:
k8s-app: cloud-controller-manager
template:
metadata:
labels:
k8s-app: cloud-controller-manager
spec:
serviceAccountName: cloud-controller-manager
containers:
- name: cloud-controller-manager
# for in-tree providers we use k8s.gcr.io/cloud-controller-manager
# this can be replaced with any other image for out-of-tree providers
image: k8s.gcr.io/cloud-controller-manager:v1.8.0
command:
- /usr/local/bin/cloud-controller-manager
- --cloud-provider=[YOUR_CLOUD_PROVIDER] # Add your own cloud provider here!
- --leader-elect=true
- --use-service-account-credentials
# these flags will vary for every cloud provider
- --allocate-node-cidrs=true
- --configure-cloud-routes=true
- --cluster-cidr=172.17.0.0/16
tolerations:
# this is required so CCM can bootstrap itself
- key: node.cloudprovider.kubernetes.io/uninitialized
value: "true"
effect: NoSchedule
# this is to have the daemonset runnable on master nodes
# the taint may vary depending on your cluster setup
- key: node-role.kubernetes.io/master
effect: NoSchedule
# this is to restrict CCM to only run on master nodes
# the node selector may vary depending on your cluster setup
nodeSelector:
node-role.kubernetes.io/master: ""
Running cloud controller manager comes with a few possible limitations. Although these limitations are being addressed in upcoming releases, it's important that you are aware of these limitations for production workloads.
Cloud controller manager does not implement any of the volume controllers found in kube-controller-manager as the volume integrations also require coordination with kubelets. As we evolve CSI (container storage interface) and add stronger support for flex volume plugins, necessary support will be added to cloud controller manager so that clouds can fully integrate with volumes. Learn more about out-of-tree CSI volume plugins here.
The cloud-controller-manager queries your cloud provider's APIs to retrieve information for all nodes. For very large clusters, consider possible bottlenecks such as resource requirements and API rate limiting.
The goal of the cloud controller manager project is to decouple development of cloud features from the core Kubernetes project. Unfortunately, many aspects of the Kubernetes project has assumptions that cloud provider features are tightly integrated into the project. As a result, adopting this new architecture can create several situations where a request is being made for information from a cloud provider, but the cloud controller manager may not be able to return that information without the original request being complete.
A good example of this is the TLS bootstrapping feature in the Kubelet. TLS bootstrapping assumes that the Kubelet has the ability to ask the cloud provider (or a local metadata service) for all its address types (private, public, etc) but cloud controller manager cannot set a node's address types without being initialized in the first place which requires that the kubelet has TLS certificates to communicate with the apiserver.
As this initiative evolves, changes will be made to address these issues in upcoming releases.
To build and develop your own cloud controller manager, read Developing Cloud Controller Manager.
This page shows how to configure quotas for API objects, including PersistentVolumeClaims and Services. A quota restricts the number of objects, of a particular type, that can be created in a namespace. You specify quotas in a ResourceQuota object.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace quota-object-example
Here is the configuration file for a ResourceQuota object:
apiVersion: v1
kind: ResourceQuota
metadata:
name: object-quota-demo
spec:
hard:
persistentvolumeclaims: "1"
services.loadbalancers: "2"
services.nodeports: "0"
Create the ResourceQuota:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-objects.yaml --namespace=quota-object-example
View detailed information about the ResourceQuota:
kubectl get resourcequota object-quota-demo --namespace=quota-object-example --output=yaml
The output shows that in the quota-object-example namespace, there can be at most one PersistentVolumeClaim, at most two Services of type LoadBalancer, and no Services of type NodePort.
status:
hard:
persistentvolumeclaims: "1"
services.loadbalancers: "2"
services.nodeports: "0"
used:
persistentvolumeclaims: "0"
services.loadbalancers: "0"
services.nodeports: "0"
Here is the configuration file for a PersistentVolumeClaim object:
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: pvc-quota-demo
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 3Gi
Create the PersistentVolumeClaim:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-objects-pvc.yaml --namespace=quota-object-example
Verify that the PersistentVolumeClaim was created:
kubectl get persistentvolumeclaims --namespace=quota-object-example
The output shows that the PersistentVolumeClaim exists and has status Pending:
NAME STATUS
pvc-quota-demo Pending
Here is the configuration file for a second PersistentVolumeClaim:
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: pvc-quota-demo-2
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 4Gi
Attempt to create the second PersistentVolumeClaim:
kubectl apply -f https://k8s.io/examples/admin/resource/quota-objects-pvc-2.yaml --namespace=quota-object-example
The output shows that the second PersistentVolumeClaim was not created, because it would have exceeded the quota for the namespace.
persistentvolumeclaims "pvc-quota-demo-2" is forbidden:
exceeded quota: object-quota-demo, requested: persistentvolumeclaims=1,
used: persistentvolumeclaims=1, limited: persistentvolumeclaims=1
These are the strings used to identify API resources that can be constrained by quotas:
| String | API Object |
|---|---|
| "pods" | Pod |
| "services" | Service |
| "replicationcontrollers" | ReplicationController |
| "resourcequotas" | ResourceQuota |
| "secrets" | Secret |
| "configmaps" | ConfigMap |
| "persistentvolumeclaims" | PersistentVolumeClaim |
| "services.nodeports" | Service of type NodePort |
| "services.loadbalancers" | Service of type LoadBalancer |
Delete your namespace:
kubectl delete namespace quota-object-example
Configure Default Memory Requests and Limits for a Namespace
Configure Minimum and Maximum Memory Constraints for a Namespace
Configure Minimum and Maximum CPU Constraints for a Namespace
Kubernetes v1.12 [beta]
Kubernetes keeps many aspects of how pods execute on nodes abstracted from the user. This is by design. However, some workloads require stronger guarantees in terms of latency and/or performance in order to operate acceptably. The kubelet provides methods to enable more complex workload placement policies while keeping the abstraction free from explicit placement directives.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
By default, the kubelet uses CFS quota to enforce pod CPU limits. When the node runs many CPU-bound pods, the workload can move to different CPU cores depending on whether the pod is throttled and which CPU cores are available at scheduling time. Many workloads are not sensitive to this migration and thus work fine without any intervention.
However, in workloads where CPU cache affinity and scheduling latency significantly affect workload performance, the kubelet allows alternative CPU management policies to determine some placement preferences on the node.
The CPU Manager policy is set with the --cpu-manager-policy kubelet
option. There are two supported policies:
none: the default policy.static: allows pods with certain resource characteristics to be
granted increased CPU affinity and exclusivity on the node.The CPU manager periodically writes resource updates through the CRI in
order to reconcile in-memory CPU assignments with cgroupfs. The reconcile
frequency is set through a new Kubelet configuration value
--cpu-manager-reconcile-period. If not specified, it defaults to the same
duration as --node-status-update-frequency.
The behavior of the static policy can be fine-tuned using the --cpu-manager-policy-options flag.
The flag takes a comma-separated list of key=value policy options.
The none policy explicitly enables the existing default CPU
affinity scheme, providing no affinity beyond what the OS scheduler does
automatically. Limits on CPU usage for
Guaranteed pods and
Burstable pods
are enforced using CFS quota.
The static policy allows containers in Guaranteed pods with integer CPU
requests access to exclusive CPUs on the node. This exclusivity is enforced
using the cpuset cgroup controller.
cpu_manager_state in the kubelet root directory.
This policy manages a shared pool of CPUs that initially contains all CPUs in the
node. The amount of exclusively allocatable CPUs is equal to the total
number of CPUs in the node minus any CPU reservations by the kubelet --kube-reserved or
--system-reserved options. From 1.17, the CPU reservation list can be specified
explicitly by kubelet --reserved-cpus option. The explicit CPU list specified by
--reserved-cpus takes precedence over the CPU reservation specified by
--kube-reserved and --system-reserved. CPUs reserved by these options are taken, in
integer quantity, from the initial shared pool in ascending order by physical
core ID. This shared pool is the set of CPUs on which any containers in
BestEffort and Burstable pods run. Containers in Guaranteed pods with fractional
CPU requests also run on CPUs in the shared pool. Only containers that are
both part of a Guaranteed pod and have integer CPU requests are assigned
exclusive CPUs.
--kube-reserved and/or --system-reserved or --reserved-cpus when
the static policy is enabled. This is because zero CPU reservation would allow the shared
pool to become empty.
As Guaranteed pods whose containers fit the requirements for being statically
assigned are scheduled to the node, CPUs are removed from the shared pool and
placed in the cpuset for the container. CFS quota is not used to bound
the CPU usage of these containers as their usage is bound by the scheduling domain
itself. In others words, the number of CPUs in the container cpuset is equal to the integer
CPU limit specified in the pod spec. This static assignment increases CPU
affinity and decreases context switches due to throttling for the CPU-bound
workload.
Consider the containers in the following pod specs:
spec:
containers:
- name: nginx
image: nginx
This pod runs in the BestEffort QoS class because no resource requests or
limits are specified. It runs in the shared pool.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
requests:
memory: "100Mi"
This pod runs in the Burstable QoS class because resource requests do not
equal limits and the cpu quantity is not specified. It runs in the shared
pool.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "2"
requests:
memory: "100Mi"
cpu: "1"
This pod runs in the Burstable QoS class because resource requests do not
equal limits. It runs in the shared pool.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "2"
requests:
memory: "200Mi"
cpu: "2"
This pod runs in the Guaranteed QoS class because requests are equal to limits.
And the container's resource limit for the CPU resource is an integer greater than
or equal to one. The nginx container is granted 2 exclusive CPUs.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "1.5"
requests:
memory: "200Mi"
cpu: "1.5"
This pod runs in the Guaranteed QoS class because requests are equal to limits.
But the container's resource limit for the CPU resource is a fraction. It runs in
the shared pool.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "2"
This pod runs in the Guaranteed QoS class because only limits are specified
and requests are set equal to limits when not explicitly specified. And the
container's resource limit for the CPU resource is an integer greater than or
equal to one. The nginx container is granted 2 exclusive CPUs.
If the full-pcpus-only policy option is specified, the static policy will always allocate full physical cores.
You can enable this option by adding full-pcups-only=true to the CPUManager policy options.
By default, without this option, the static policy allocates CPUs using a topology-aware best-fit allocation.
On SMT enabled systems, the policy can allocate individual virtual cores, which correspond to hardware threads.
This can lead to different containers sharing the same physical cores; this behaviour in turn contributes
to the noisy neighbours problem.
With the option enabled, the pod will be admitted by the kubelet only if the CPU request of all its containers
can be fulfilled by allocating full physical cores.
If the pod does not pass the admission, it will be put in Failed state with the message SMTAlignmentError.
Kubernetes v1.18 [beta]
An increasing number of systems leverage a combination of CPUs and hardware accelerators to support latency-critical execution and high-throughput parallel computation. These include workloads in fields such as telecommunications, scientific computing, machine learning, financial services and data analytics. Such hybrid systems comprise a high performance environment.
In order to extract the best performance, optimizations related to CPU isolation, memory and device locality are required. However, in Kubernetes, these optimizations are handled by a disjoint set of components.
Topology Manager is a Kubelet component that aims to co-ordinate the set of components that are responsible for these optimizations.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.18. To check the version, enterkubectl version.
Prior to the introduction of Topology Manager, the CPU and Device Manager in Kubernetes make resource allocation decisions independently of each other. This can result in undesirable allocations on multiple-socketed systems, performance/latency sensitive applications will suffer due to these undesirable allocations. Undesirable in this case meaning for example, CPUs and devices being allocated from different NUMA Nodes thus, incurring additional latency.
The Topology Manager is a Kubelet component, which acts as a source of truth so that other Kubelet components can make topology aligned resource allocation choices.
The Topology Manager provides an interface for components, called Hint Providers, to send and receive topology information. Topology Manager has a set of node level policies which are explained below.
The Topology manager receives Topology information from the Hint Providers as a bitmask denoting NUMA Nodes available and a preferred allocation indication. The Topology Manager policies perform a set of operations on the hints provided and converge on the hint determined by the policy to give the optimal result, if an undesirable hint is stored the preferred field for the hint will be set to false. In the current policies preferred is the narrowest preferred mask. The selected hint is stored as part of the Topology Manager. Depending on the policy configured the pod can be accepted or rejected from the node based on the selected hint. The hint is then stored in the Topology Manager for use by the Hint Providers when making the resource allocation decisions.
Support for the Topology Manager requires TopologyManager feature gate to be enabled. It is enabled by default starting with Kubernetes 1.18.
The Topology Manager currently:
If these conditions are met, the Topology Manager will align the requested resources.
In order to customise how this alignment is carried out, the Topology Manager provides two distinct knobs: scope and policy.
The scope defines the granularity at which you would like resource alignment to be performed (e.g. at the pod or container level). And the policy defines the actual strategy used to carry out the alignment (e.g. best-effort, restricted, single-numa-node, etc.).
Details on the various scopes and policies available today can be found below.
The Topology Manager can deal with the alignment of resources in a couple of distinct scopes:
container (default)podEither option can be selected at a time of the kubelet startup, with --topology-manager-scope flag.
The container scope is used by default.
Within this scope, the Topology Manager performs a number of sequential resource alignments, i.e., for each container (in a pod) a separate alignment is computed. In other words, there is no notion of grouping the containers to a specific set of NUMA nodes, for this particular scope. In effect, the Topology Manager performs an arbitrary alignment of individual containers to NUMA nodes.
The notion of grouping the containers was endorsed and implemented on purpose in the following scope, for example the pod scope.
To select the pod scope, start the kubelet with the command line option --topology-manager-scope=pod.
This scope allows for grouping all containers in a pod to a common set of NUMA nodes. That is, the Topology Manager treats a pod as a whole and attempts to allocate the entire pod (all containers) to either a single NUMA node or a common set of NUMA nodes. The following examples illustrate the alignments produced by the Topology Manager on different occasions:
The total amount of particular resource demanded for the entire pod is calculated according to effective requests/limits formula, and thus, this total value is equal to the maximum of:
Using the pod scope in tandem with single-numa-node Topology Manager policy is specifically valuable for workloads that are latency sensitive or for high-throughput applications that perform IPC. By combining both options, you are able to place all containers in a pod onto a single NUMA node; hence, the inter-NUMA communication overhead can be eliminated for that pod.
In the case of single-numa-node policy, a pod is accepted only if a suitable set of NUMA nodes is present among possible allocations. Reconsider the example above:
To recap, Topology Manager first computes a set of NUMA nodes and then tests it against Topology Manager policy, which either leads to the rejection or admission of the pod.
Topology Manager supports four allocation policies. You can set a policy via a Kubelet flag, --topology-manager-policy.
There are four supported policies:
none (default)best-effortrestrictedsingle-numa-nodeThis is the default policy and does not perform any topology alignment.
For each container in a Pod, the kubelet, with best-effort topology
management policy, calls each Hint Provider to discover their resource availability.
Using this information, the Topology Manager stores the
preferred NUMA Node affinity for that container. If the affinity is not preferred,
Topology Manager will store this and admit the pod to the node anyway.
The Hint Providers can then use this information when making the resource allocation decision.
For each container in a Pod, the kubelet, with restricted topology
management policy, calls each Hint Provider to discover their resource availability.
Using this information, the Topology Manager stores the
preferred NUMA Node affinity for that container. If the affinity is not preferred,
Topology Manager will reject this pod from the node. This will result in a pod in a Terminated state with a pod admission failure.
Once the pod is in a Terminated state, the Kubernetes scheduler will not attempt to reschedule the pod. It is recommended to use a ReplicaSet or Deployment to trigger a redeploy of the pod.
An external control loop could be also implemented to trigger a redeployment of pods that have the Topology Affinity error.
If the pod is admitted, the Hint Providers can then use this information when making the resource allocation decision.
For each container in a Pod, the kubelet, with single-numa-node topology
management policy, calls each Hint Provider to discover their resource availability.
Using this information, the Topology Manager determines if a single NUMA Node affinity is possible.
If it is, Topology Manager will store this and the Hint Providers can then use this information when making the
resource allocation decision.
If, however, this is not possible then the Topology Manager will reject the pod from the node. This will result in a pod in a Terminated state with a pod admission failure.
Once the pod is in a Terminated state, the Kubernetes scheduler will not attempt to reschedule the pod. It is recommended to use a Deployment with replicas to trigger a redeploy of the Pod.
An external control loop could be also implemented to trigger a redeployment of pods that have the Topology Affinity error.
Consider the containers in the following pod specs:
spec:
containers:
- name: nginx
image: nginx
This pod runs in the BestEffort QoS class because no resource requests or
limits are specified.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
requests:
memory: "100Mi"
This pod runs in the Burstable QoS class because requests are less than limits.
If the selected policy is anything other than none, Topology Manager would consider these Pod specifications. The Topology Manager would consult the Hint Providers to get topology hints. In the case of the static, the CPU Manager policy would return default topology hint, because these Pods do not have explicitly request CPU resources.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "2"
example.com/device: "1"
requests:
memory: "200Mi"
cpu: "2"
example.com/device: "1"
This pod with integer CPU request runs in the Guaranteed QoS class because requests are equal to limits.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "300m"
example.com/device: "1"
requests:
memory: "200Mi"
cpu: "300m"
example.com/device: "1"
This pod with sharing CPU request runs in the Guaranteed QoS class because requests are equal to limits.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
example.com/deviceA: "1"
example.com/deviceB: "1"
requests:
example.com/deviceA: "1"
example.com/deviceB: "1"
This pod runs in the BestEffort QoS class because there are no CPU and memory requests.
The Topology Manager would consider the above pods. The Topology Manager would consult the Hint Providers, which are CPU and Device Manager to get topology hints for the pods.
In the case of the Guaranteed pod with integer CPU request, the static CPU Manager policy would return topology hints relating to the exclusive CPU and the Device Manager would send back hints for the requested device.
In the case of the Guaranteed pod with sharing CPU request, the static CPU Manager policy would return default topology hint as there is no exclusive CPU request and the Device Manager would send back hints for the requested device.
In the above two cases of the Guaranteed pod, the none CPU Manager policy would return default topology hint.
In the case of the BestEffort pod, the static CPU Manager policy would send back the default topology hint as there is no CPU request and the Device Manager would send back the hints for each of the requested devices.
Using this information the Topology Manager calculates the optimal hint for the pod and stores this information, which will be used by the Hint Providers when they are making their resource assignments.
The maximum number of NUMA nodes that Topology Manager allows is 8. With more than 8 NUMA nodes there will be a state explosion when trying to enumerate the possible NUMA affinities and generating their hints.
The scheduler is not topology-aware, so it is possible to be scheduled on a node and then fail on the node due to the Topology Manager.
This page explains how to configure your DNS Pod(s) and customize the DNS resolution process in your cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your cluster must be running the CoreDNS add-on.
Migrating to CoreDNS
explains how to use kubeadm to migrate from kube-dns.
Your Kubernetes server must be at or later than version v1.12.
To check the version, enter kubectl version.
DNS is a built-in Kubernetes service launched automatically using the addon manager cluster add-on.
As of Kubernetes v1.12, CoreDNS is the recommended DNS Server, replacing kube-dns. If your cluster
originally used kube-dns, you may still have kube-dns deployed rather than CoreDNS.
kube-dns in the metadata.name field.kube-dns Service name to resolve addresses internal to the cluster. Using a Service named kube-dns abstracts away the implementation detail of which DNS provider is running behind that common name.
If you are running CoreDNS as a Deployment, it will typically be exposed as a Kubernetes Service with a static IP address.
The kubelet passes DNS resolver information to each container with the --cluster-dns=<dns-service-ip> flag.
DNS names also need domains. You configure the local domain in the kubelet
with the flag --cluster-domain=<default-local-domain>.
The DNS server supports forward lookups (A and AAAA records), port lookups (SRV records), reverse IP address lookups (PTR records), and more. For more information, see DNS for Services and Pods.
If a Pod's dnsPolicy is set to default, it inherits the name resolution
configuration from the node that the Pod runs on. The Pod's DNS resolution
should behave the same as the node.
But see Known issues.
If you don't want this, or if you want a different DNS config for pods, you can
use the kubelet's --resolv-conf flag. Set this flag to "" to prevent Pods from
inheriting DNS. Set it to a valid file path to specify a file other than
/etc/resolv.conf for DNS inheritance.
CoreDNS is a general-purpose authoritative DNS server that can serve as cluster DNS, complying with the dns specifications.
CoreDNS is a DNS server that is modular and pluggable, and each plugin adds new functionality to CoreDNS. This can be configured by maintaining a Corefile, which is the CoreDNS configuration file. As a cluster administrator, you can modify the ConfigMap for the CoreDNS Corefile to change how DNS service discovery behaves for that cluster.
In Kubernetes, CoreDNS is installed with the following default Corefile configuration:
apiVersion: v1
kind: ConfigMap
metadata:
name: coredns
namespace: kube-system
data:
Corefile: |
.:53 {
errors
health {
lameduck 5s
}
ready
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
fallthrough in-addr.arpa ip6.arpa
ttl 30
}
prometheus :9153
forward . /etc/resolv.conf
cache 30
loop
reload
loadbalance
}
The Corefile configuration includes the following plugins of CoreDNS:
http://localhost:8080/health. In this extended syntax lameduck will make the process unhealthy then wait for 5 seconds before the process is shut down.ttl allows you to set a custom TTL for responses. The default is 5 seconds. The minimum TTL allowed is 0 seconds, and the maximum is capped at 3600 seconds. Setting TTL to 0 will prevent records from being cached.pods insecure option is provided for backward compatibility with kube-dns. You can use the pods verified option, which returns an A record only if there exists a pod in same namespace with matching IP. The pods disabled option can be used if you don't use pod records.http://localhost:9153/metrics in Prometheus format (also known as OpenMetrics).You can modify the default CoreDNS behavior by modifying the ConfigMap.
CoreDNS has the ability to configure stubdomains and upstream nameservers using the forward plugin.
If a cluster operator has a Consul domain server located at 10.150.0.1, and all Consul names have the suffix .consul.local. To configure it in CoreDNS, the cluster administrator creates the following stanza in the CoreDNS ConfigMap.
consul.local:53 {
errors
cache 30
forward . 10.150.0.1
}
To explicitly force all non-cluster DNS lookups to go through a specific nameserver at 172.16.0.1, point the forward to the nameserver instead of /etc/resolv.conf
forward . 172.16.0.1
The final ConfigMap along with the default Corefile configuration looks like:
apiVersion: v1
kind: ConfigMap
metadata:
name: coredns
namespace: kube-system
data:
Corefile: |
.:53 {
errors
health
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
fallthrough in-addr.arpa ip6.arpa
}
prometheus :9153
forward . 172.16.0.1
cache 30
loop
reload
loadbalance
}
consul.local:53 {
errors
cache 30
forward . 10.150.0.1
}
The kubeadm tool supports automatic translation from the kube-dns ConfigMap
to the equivalent CoreDNS ConfigMap.
CoreDNS supports the features of kube-dns and more.
A ConfigMap created for kube-dns to support StubDomainsand upstreamNameservers translates to the forward plugin in CoreDNS.
This example ConfigMap for kube-dns specifies stubdomains and upstreamnameservers:
apiVersion: v1
data:
stubDomains: |
{"abc.com" : ["1.2.3.4"], "my.cluster.local" : ["2.3.4.5"]}
upstreamNameservers: |
["8.8.8.8", "8.8.4.4"]
kind: ConfigMap
The equivalent configuration in CoreDNS creates a Corefile:
abc.com:53 {
errors
cache 30
forward . 1.2.3.4
}
my.cluster.local:53 {
errors
cache 30
forward . 2.3.4.5
}
The complete Corefile with the default plugins:
.:53 {
errors
health
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
fallthrough in-addr.arpa ip6.arpa
}
federation cluster.local {
foo foo.feddomain.com
}
prometheus :9153
forward . 8.8.8.8 8.8.4.4
cache 30
}
abc.com:53 {
errors
cache 30
forward . 1.2.3.4
}
my.cluster.local:53 {
errors
cache 30
forward . 2.3.4.5
}
To migrate from kube-dns to CoreDNS, a detailed blog article is available to help users adapt CoreDNS in place of kube-dns.
You can also migrate using the official CoreDNS deploy script.
This page provides hints on diagnosing DNS problems.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.6.
To check the version, enter kubectl version.
apiVersion: v1
kind: Pod
metadata:
name: dnsutils
namespace: default
spec:
containers:
- name: dnsutils
image: gcr.io/kubernetes-e2e-test-images/dnsutils:1.3
command:
- sleep
- "3600"
imagePullPolicy: IfNotPresent
restartPolicy: Always
default namespace. DNS name resolution for
services depends on the namespace of the pod. For more information, review
DNS for Services and Pods.
Use that manifest to create a Pod:
kubectl apply -f https://k8s.io/examples/admin/dns/dnsutils.yaml
pod/dnsutils created
…and verify its status:
kubectl get pods dnsutils
NAME READY STATUS RESTARTS AGE
dnsutils 1/1 Running 0 <some-time>
Once that Pod is running, you can exec nslookup in that environment.
If you see something like the following, DNS is working correctly.
kubectl exec -i -t dnsutils -- nslookup kubernetes.default
Server: 10.0.0.10
Address 1: 10.0.0.10
Name: kubernetes.default
Address 1: 10.0.0.1
If the nslookup command fails, check the following:
Take a look inside the resolv.conf file. (See Customizing DNS Service and Known issues below for more information)
kubectl exec -ti dnsutils -- cat /etc/resolv.conf
Verify that the search path and name server are set up like the following (note that search path may vary for different cloud providers):
search default.svc.cluster.local svc.cluster.local cluster.local google.internal c.gce_project_id.internal
nameserver 10.0.0.10
options ndots:5
Errors such as the following indicate a problem with the CoreDNS (or kube-dns) add-on or with associated Services:
kubectl exec -i -t dnsutils -- nslookup kubernetes.default
Server: 10.0.0.10
Address 1: 10.0.0.10
nslookup: can't resolve 'kubernetes.default'
or
kubectl exec -i -t dnsutils -- nslookup kubernetes.default
Server: 10.0.0.10
Address 1: 10.0.0.10 kube-dns.kube-system.svc.cluster.local
nslookup: can't resolve 'kubernetes.default'
Use the kubectl get pods command to verify that the DNS pod is running.
kubectl get pods --namespace=kube-system -l k8s-app=kube-dns
NAME READY STATUS RESTARTS AGE
...
coredns-7b96bf9f76-5hsxb 1/1 Running 0 1h
coredns-7b96bf9f76-mvmmt 1/1 Running 0 1h
...
k8s-app is kube-dns for both CoreDNS and kube-dns deployments.
If you see that no CoreDNS Pod is running or that the Pod has failed/completed, the DNS add-on may not be deployed by default in your current environment and you will have to deploy it manually.
Use the kubectl logs command to see logs for the DNS containers.
For CoreDNS:
kubectl logs --namespace=kube-system -l k8s-app=kube-dns
Here is an example of a healthy CoreDNS log:
.:53
2018/08/15 14:37:17 [INFO] CoreDNS-1.2.2
2018/08/15 14:37:17 [INFO] linux/amd64, go1.10.3, 2e322f6
CoreDNS-1.2.2
linux/amd64, go1.10.3, 2e322f6
2018/08/15 14:37:17 [INFO] plugin/reload: Running configuration MD5 = 24e6c59e83ce706f07bcc82c31b1ea1c
See if there are any suspicious or unexpected messages in the logs.
Verify that the DNS service is up by using the kubectl get service command.
kubectl get svc --namespace=kube-system
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
...
kube-dns ClusterIP 10.0.0.10 <none> 53/UDP,53/TCP 1h
...
kube-dns for both CoreDNS and kube-dns deployments.
If you have created the Service or in the case it should be created by default but it does not appear, see debugging Services for more information.
You can verify that DNS endpoints are exposed by using the kubectl get endpoints
command.
kubectl get endpoints kube-dns --namespace=kube-system
NAME ENDPOINTS AGE
kube-dns 10.180.3.17:53,10.180.3.17:53 1h
If you do not see the endpoints, see the endpoints section in the debugging Services documentation.
For additional Kubernetes DNS examples, see the cluster-dns examples in the Kubernetes GitHub repository.
You can verify if queries are being received by CoreDNS by adding the log plugin to the CoreDNS configuration (aka Corefile).
The CoreDNS Corefile is held in a ConfigMap named coredns. To edit it, use the command:
kubectl -n kube-system edit configmap coredns
Then add log in the Corefile section per the example below:
apiVersion: v1
kind: ConfigMap
metadata:
name: coredns
namespace: kube-system
data:
Corefile: |
.:53 {
log
errors
health
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
upstream
fallthrough in-addr.arpa ip6.arpa
}
prometheus :9153
forward . /etc/resolv.conf
cache 30
loop
reload
loadbalance
}
After saving the changes, it may take up to minute or two for Kubernetes to propagate these changes to the CoreDNS pods.
Next, make some queries and view the logs per the sections above in this document. If CoreDNS pods are receiving the queries, you should see them in the logs.
Here is an example of a query in the log:
.:53
2018/08/15 14:37:15 [INFO] CoreDNS-1.2.0
2018/08/15 14:37:15 [INFO] linux/amd64, go1.10.3, 2e322f6
CoreDNS-1.2.0
linux/amd64, go1.10.3, 2e322f6
2018/09/07 15:29:04 [INFO] plugin/reload: Running configuration MD5 = 162475cdf272d8aa601e6fe67a6ad42f
2018/09/07 15:29:04 [INFO] Reloading complete
172.17.0.18:41675 - [07/Sep/2018:15:29:11 +0000] 59925 "A IN kubernetes.default.svc.cluster.local. udp 54 false 512" NOERROR qr,aa,rd,ra 106 0.000066649s
DNS queries that don't specify a namespace are limited to the pod's namespace.
If the namespace of the pod and service differ, the DNS query must include the namespace of the service.
This query is limited to the pod's namespace:
kubectl exec -i -t dnsutils -- nslookup <service-name>
This query specifies the namespace:
kubectl exec -i -t dnsutils -- nslookup <service-name>.<namespace>
To learn more about name resolution, see DNS for Services and Pods.
Some Linux distributions (e.g. Ubuntu) use a local DNS resolver by default (systemd-resolved).
Systemd-resolved moves and replaces /etc/resolv.conf with a stub file that can cause a fatal forwarding
loop when resolving names in upstream servers. This can be fixed manually by using kubelet's --resolv-conf flag
to point to the correct resolv.conf (With systemd-resolved, this is /run/systemd/resolve/resolv.conf).
kubeadm automatically detects systemd-resolved, and adjusts the kubelet flags accordingly.
Kubernetes installs do not configure the nodes' resolv.conf files to use the
cluster DNS by default, because that process is inherently distribution-specific.
This should probably be implemented eventually.
Linux's libc (a.k.a. glibc) has a limit for the DNS nameserver records to 3 by default. What's more, for the glibc versions which are older than glibc-2.17-222 (the new versions update see this issue), the allowed number of DNS search records has been limited to 6 (see this bug from 2005). Kubernetes needs to consume 1 nameserver record and 3 search records. This means that if a local installation already uses 3 nameservers or uses more than 3 searches while your glibc version is in the affected list, some of those settings will be lost. To work around the DNS nameserver records limit, the node can run dnsmasq, which will provide more nameserver entries. You can also use kubelet's --resolv-conf flag. To fix the DNS search records limit, consider upgrading your linux distribution or upgrading to an unaffected version of glibc.
search records.
If you are using Alpine version 3.3 or earlier as your base image, DNS may not work properly due to a known issue with Alpine. Kubernetes issue 30215 details more information on this.
This document helps you get started using the Kubernetes NetworkPolicy API to declare network policies that govern how pods communicate with each other.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.8. To check the version, enterkubectl version.
Make sure you've configured a network provider with network policy support. There are a number of network providers that support NetworkPolicy, including:
nginx deployment and expose it via a serviceTo see how Kubernetes network policy works, start off by creating an nginx Deployment.
kubectl create deployment nginx --image=nginx
deployment.apps/nginx created
Expose the Deployment through a Service called nginx.
kubectl expose deployment nginx --port=80
service/nginx exposed
The above commands create a Deployment with an nginx Pod and expose the Deployment through a Service named nginx. The nginx Pod and Deployment are found in the default namespace.
kubectl get svc,pod
NAME CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/kubernetes 10.100.0.1 <none> 443/TCP 46m
service/nginx 10.100.0.16 <none> 80/TCP 33s
NAME READY STATUS RESTARTS AGE
pod/nginx-701339712-e0qfq 1/1 Running 0 35s
You should be able to access the new nginx service from other Pods. To access the nginx Service from another Pod in the default namespace, start a busybox container:
kubectl run busybox --rm -ti --image=busybox -- /bin/sh
In your shell, run the following command:
wget --spider --timeout=1 nginx
Connecting to nginx (10.100.0.16:80)
remote file exists
nginx serviceTo limit the access to the nginx service so that only Pods with the label access: true can query it, create a NetworkPolicy object as follows:
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: access-nginx
spec:
podSelector:
matchLabels:
app: nginx
ingress:
- from:
- podSelector:
matchLabels:
access: "true"
The name of a NetworkPolicy object must be a valid DNS subdomain name.
podSelector which selects the grouping of Pods to which the policy applies. You can see this policy selects Pods with the label app=nginx. The label was automatically added to the Pod in the nginx Deployment. An empty podSelector selects all pods in the namespace.
Use kubectl to create a NetworkPolicy from the above nginx-policy.yaml file:
kubectl apply -f https://k8s.io/examples/service/networking/nginx-policy.yaml
networkpolicy.networking.k8s.io/access-nginx created
When you attempt to access the nginx Service from a Pod without the correct labels, the request times out:
kubectl run busybox --rm -ti --image=busybox -- /bin/sh
In your shell, run the command:
wget --spider --timeout=1 nginx
Connecting to nginx (10.100.0.16:80)
wget: download timed out
You can create a Pod with the correct labels to see that the request is allowed:
kubectl run busybox --rm -ti --labels="access=true" --image=busybox -- /bin/sh
In your shell, run the command:
wget --spider --timeout=1 nginx
Connecting to nginx (10.100.0.16:80)
remote file exists
Kubernetes v1.11 [beta]
The cloud-controller-manager is a Kubernetes control plane component that embeds cloud-specific control logic. The cloud controller manager lets you link your cluster into your cloud provider's API, and separates out the components that interact with that cloud platform from components that only interact with your cluster.
By decoupling the interoperability logic between Kubernetes and the underlying cloud infrastructure, the cloud-controller-manager component enables cloud providers to release features at a different pace compared to the main Kubernetes project.
Since cloud providers develop and release at a different pace compared to the Kubernetes project, abstracting the provider-specific code to the cloud-controller-manager binary allows cloud vendors to evolve independently from the core Kubernetes code.
The Kubernetes project provides skeleton cloud-controller-manager code with Go interfaces to allow you (or your cloud provider) to plug in your own implementations. This means that a cloud provider can implement a cloud-controller-manager by importing packages from Kubernetes core; each cloudprovider will register their own code by calling cloudprovider.RegisterCloudProvider to update a global variable of available cloud providers.
To build an out-of-tree cloud-controller-manager for your cloud:
main.go in cloud-controller-manager from Kubernetes core as a template for your main.go. As mentioned above, the only difference should be the cloud package that will be imported.main.go, ensure your package has an init block to run cloudprovider.RegisterCloudProvider.Many cloud providers publish their controller manager code as open source. If you are creating a new cloud-controller-manager from scratch, you could take an existing out-of-tree cloud controller manager as your starting point.
For in-tree cloud providers, you can run the in-tree cloud controller manager as a DaemonSet in your cluster. See Cloud Controller Manager Administration for more details.
This page shows how to enable or disable an API version from your cluster's control plane.
Specific API versions can be turned on or off by passing --runtime-config=api/<version> as a
command line argument to the API server. The values for this argument are a comma-separated
list of API versions. Later values override earlier values.
The runtime-config command line argument also supports 2 special keys:
api/all, representing all known APIsapi/legacy, representing only legacy APIs. Legacy APIs are any APIs that have been
explicitly deprecated.For example, to turning off all API versions except v1, pass --runtime-config=api/all=false,api/v1=true
to the kube-apiserver.
Read the full documentation
for the kube-apiserver component.
Kubernetes v1.21 [deprecated]
This feature, specifically the alpha topologyKeys field, is deprecated since
Kubernetes v1.21.
Topology Aware Hints,
introduced in Kubernetes v1.21, provide similar functionality.
Service Topology enables a Service to route traffic based upon the Node topology of the cluster. For example, a service can specify that traffic be preferentially routed to endpoints that are on the same Node as the client, or in the same availability zone.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.17. To check the version, enterkubectl version.
The following prerequisites are needed in order to enable topology aware service routing:
Kubernetes v1.21 [deprecated]
To enable service topology, enable the ServiceTopology
feature gate for all Kubernetes components:
--feature-gates="ServiceTopology=true`
topologyKeys field.Kubernetes v1.21 [alpha]
Topology Aware Hints enable topology aware routing with topology hints included in EndpointSlices. This approach tries to keep traffic close to where it originated from; you might do this to reduce costs, or to improve network performance.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.21. To check the version, enterkubectl version.
The following prerequisite is needed in order to enable topology aware hints:
To enable service topology hints, enable the TopologyAwareHints feature
gate for the
kube-apiserver, kube-controller-manager, and kube-proxy:
--feature-gates="TopologyAwareHints=true"
This page shows how to enable and configure encryption of secret data at rest.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.13. To check the version, enterkubectl version.
etcd v3.0 or later is required
The kube-apiserver process accepts an argument --encryption-provider-config
that controls how API data is encrypted in etcd. An example configuration
is provided below.
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- identity: {}
- aesgcm:
keys:
- name: key1
secret: c2VjcmV0IGlzIHNlY3VyZQ==
- name: key2
secret: dGhpcyBpcyBwYXNzd29yZA==
- aescbc:
keys:
- name: key1
secret: c2VjcmV0IGlzIHNlY3VyZQ==
- name: key2
secret: dGhpcyBpcyBwYXNzd29yZA==
- secretbox:
keys:
- name: key1
secret: YWJjZGVmZ2hpamtsbW5vcHFyc3R1dnd4eXoxMjM0NTY=
Each resources array item is a separate config and contains a complete configuration. The
resources.resources field is an array of Kubernetes resource names (resource or resource.group)
that should be encrypted. The providers array is an ordered list of the possible encryption
providers. Only one provider type may be specified per entry (identity or aescbc may be provided, but not both in the same item).
The first provider in the list is used to encrypt resources going into storage. When reading resources from storage each provider that matches the stored data attempts to decrypt the data in order. If no provider can read the stored data due to a mismatch in format or secret key, an error is returned which prevents clients from accessing that resource.
| Name | Encryption | Strength | Speed | Key Length | Other Considerations |
|---|---|---|---|---|---|
identity |
None | N/A | N/A | N/A | Resources written as-is without encryption. When set as the first provider, the resource will be decrypted as new values are written. |
secretbox |
XSalsa20 and Poly1305 | Strong | Faster | 32-byte | A newer standard and may not be considered acceptable in environments that require high levels of review. |
aesgcm |
AES-GCM with random nonce | Must be rotated every 200k writes | Fastest | 16, 24, or 32-byte | Is not recommended for use except when an automated key rotation scheme is implemented. |
aescbc |
AES-CBC with PKCS#7 padding | Weak | Fast | 32-byte | Not recommended due to CBC's vulnerability to padding oracle attacks. |
kms |
Uses envelope encryption scheme: Data is encrypted by data encryption keys (DEKs) using AES-CBC with PKCS#7 padding, DEKs are encrypted by key encryption keys (KEKs) according to configuration in Key Management Service (KMS) | Strongest | Fast | 32-bytes | The recommended choice for using a third party tool for key management. Simplifies key rotation, with a new DEK generated for each encryption, and KEK rotation controlled by the user. Configure the KMS provider |
Each provider supports multiple keys - the keys are tried in order for decryption, and if the provider is the first provider, the first key is used for encryption.
Storing the raw encryption key in the EncryptionConfig only moderately improves your security posture, compared to no encryption.
Please use kms provider for additional security. By default, the identity provider is used to protect secrets in etcd, which
provides no encryption. EncryptionConfiguration was introduced to encrypt secrets locally, with a locally managed key.
Encrypting secrets with a locally managed key protects against an etcd compromise, but it fails to protect against a host compromise. Since the encryption keys are stored on the host in the EncryptionConfig YAML file, a skilled attacker can access that file and extract the encryption keys.
Envelope encryption creates dependence on a separate key, not stored in Kubernetes. In this case, an attacker would need to compromise etcd, the kubeapi-server, and the third-party KMS provider to retrieve the plaintext values, providing a higher level of security than locally-stored encryption keys.
Create a new encryption config file:
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- aescbc:
keys:
- name: key1
secret: <BASE 64 ENCODED SECRET>
- identity: {}
To create a new secret perform the following steps:
Generate a 32 byte random key and base64 encode it. If you're on Linux or macOS, run the following command:
head -c 32 /dev/urandom | base64
Place that value in the secret field.
Set the --encryption-provider-config flag on the kube-apiserver to point to the location of the config file.
Restart your API server.
Data is encrypted when written to etcd. After restarting your kube-apiserver, any newly created or
updated secret should be encrypted when stored. To check, you can use the etcdctl command line
program to retrieve the contents of your secret.
Create a new secret called secret1 in the default namespace:
kubectl create secret generic secret1 -n default --from-literal=mykey=mydata
Using the etcdctl commandline, read that secret out of etcd:
ETCDCTL_API=3 etcdctl get /registry/secrets/default/secret1 [...] | hexdump -C
where [...] must be the additional arguments for connecting to the etcd server.
Verify the stored secret is prefixed with k8s:enc:aescbc:v1: which indicates the aescbc provider has encrypted the resulting data.
Verify the secret is correctly decrypted when retrieved via the API:
kubectl describe secret secret1 -n default
should match mykey: bXlkYXRh, mydata is encoded, check decoding a secret to
completely decode the secret.
Since secrets are encrypted on write, performing an update on a secret will encrypt that content.
kubectl get secrets --all-namespaces -o json | kubectl replace -f -
The command above reads all secrets and then updates them to apply server side encryption.
Changing the secret without incurring downtime requires a multi step operation, especially in
the presence of a highly available deployment where multiple kube-apiserver processes are running.
kube-apiserver processes to ensure each server can decrypt using the new keykeys array so that it is used for encryption in the configkube-apiserver processes to ensure each server now encrypts using the new keykubectl get secrets --all-namespaces -o json | kubectl replace -f - to encrypt all existing secrets with the new keyWith a single kube-apiserver, step 2 may be skipped.
To disable encryption at rest place the identity provider as the first entry in the config:
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- identity: {}
- aescbc:
keys:
- name: key1
secret: <BASE 64 ENCODED SECRET>
and restart all kube-apiserver processes. Then run:
kubectl get secrets --all-namespaces -o json | kubectl replace -f -
to force all secrets to be decrypted.
Kubernetes core components such as the API server, scheduler, and controller-manager run on a control plane node. However, add-ons must run on a regular cluster node. Some of these add-ons are critical to a fully functional cluster, such as metrics-server, DNS, and UI. A cluster may stop working properly if a critical add-on is evicted (either manually or as a side effect of another operation like upgrade) and becomes pending (for example when the cluster is highly utilized and either there are other pending pods that schedule into the space vacated by the evicted critical add-on pod or the amount of resources available on the node changed for some other reason).
Note that marking a pod as critical is not meant to prevent evictions entirely; it only prevents the pod from becoming permanently unavailable. A static pod marked as critical, can't be evicted. However, a non-static pods marked as critical are always rescheduled.
To mark a Pod as critical, set priorityClassName for that Pod to system-cluster-critical or system-node-critical. system-node-critical is the highest available priority, even higher than system-cluster-critical.
This page shows how to configure and enable the ip-masq-agent.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The ip-masq-agent configures iptables rules to hide a pod's IP address behind the cluster node's IP address. This is typically done when sending traffic to destinations outside the cluster's pod CIDR range.
The ip-masq-agent configures iptables rules to handle masquerading node/pod IP addresses when sending traffic to destinations outside the cluster node's IP and the Cluster IP range. This essentially hides pod IP addresses behind the cluster node's IP address. In some environments, traffic to "external" addresses must come from a known machine address. For example, in Google Cloud, any traffic to the internet must come from a VM's IP. When containers are used, as in Google Kubernetes Engine, the Pod IP will be rejected for egress. To avoid this, we must hide the Pod IP behind the VM's own IP address - generally known as "masquerade". By default, the agent is configured to treat the three private IP ranges specified by RFC 1918 as non-masquerade CIDR. These ranges are 10.0.0.0/8, 172.16.0.0/12, and 192.168.0.0/16. The agent will also treat link-local (169.254.0.0/16) as a non-masquerade CIDR by default. The agent is configured to reload its configuration from the location /etc/config/ip-masq-agent every 60 seconds, which is also configurable.
The agent configuration file must be written in YAML or JSON syntax, and may contain three optional keys:
Traffic to 10.0.0.0/8, 172.16.0.0/12 and 192.168.0.0/16) ranges will NOT be masqueraded. Any other traffic (assumed to be internet) will be masqueraded. An example of a local destination from a pod could be its Node's IP address as well as another node's address or one of the IP addresses in Cluster's IP range. Any other traffic will be masqueraded by default. The below entries show the default set of rules that are applied by the ip-masq-agent:
iptables -t nat -L IP-MASQ-AGENT
RETURN all -- anywhere 169.254.0.0/16 /* ip-masq-agent: cluster-local traffic should not be subject to MASQUERADE */ ADDRTYPE match dst-type !LOCAL
RETURN all -- anywhere 10.0.0.0/8 /* ip-masq-agent: cluster-local traffic should not be subject to MASQUERADE */ ADDRTYPE match dst-type !LOCAL
RETURN all -- anywhere 172.16.0.0/12 /* ip-masq-agent: cluster-local traffic should not be subject to MASQUERADE */ ADDRTYPE match dst-type !LOCAL
RETURN all -- anywhere 192.168.0.0/16 /* ip-masq-agent: cluster-local traffic should not be subject to MASQUERADE */ ADDRTYPE match dst-type !LOCAL
MASQUERADE all -- anywhere anywhere /* ip-masq-agent: outbound traffic should be subject to MASQUERADE (this match must come after cluster-local CIDR matches) */ ADDRTYPE match dst-type !LOCAL
By default, in GCE/Google Kubernetes Engine starting with Kubernetes version 1.7.0, if network policy is enabled or you are using a cluster CIDR not in the 10.0.0.0/8 range, the ip-masq-agent will run in your cluster. If you are running in another environment, you can add the ip-masq-agent DaemonSet to your cluster:
To create an ip-masq-agent, run the following kubectl command:
kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/ip-masq-agent/master/ip-masq-agent.yaml
You must also apply the appropriate node label to any nodes in your cluster that you want the agent to run on.
kubectl label nodes my-node beta.kubernetes.io/masq-agent-ds-ready=true
More information can be found in the ip-masq-agent documentation here
In most cases, the default set of rules should be sufficient; however, if this is not the case for your cluster, you can create and apply a ConfigMap to customize the IP ranges that are affected. For example, to allow only 10.0.0.0/8 to be considered by the ip-masq-agent, you can create the following ConfigMap in a file called "config".
It is important that the file is called config since, by default, that will be used as the key for lookup by the ip-masq-agent:
nonMasqueradeCIDRs:
- 10.0.0.0/8
resyncInterval: 60s
Run the following command to add the config map to your cluster:
kubectl create configmap ip-masq-agent --from-file=config --namespace=kube-system
This will update a file located at /etc/config/ip-masq-agent which is periodically checked every resyncInterval and applied to the cluster node. After the resync interval has expired, you should see the iptables rules reflect your changes:
iptables -t nat -L IP-MASQ-AGENT
Chain IP-MASQ-AGENT (1 references)
target prot opt source destination
RETURN all -- anywhere 169.254.0.0/16 /* ip-masq-agent: cluster-local traffic should not be subject to MASQUERADE */ ADDRTYPE match dst-type !LOCAL
RETURN all -- anywhere 10.0.0.0/8 /* ip-masq-agent: cluster-local
MASQUERADE all -- anywhere anywhere /* ip-masq-agent: outbound traffic should be subject to MASQUERADE (this match must come after cluster-local CIDR matches) */ ADDRTYPE match dst-type !LOCAL
By default, the link local range (169.254.0.0/16) is also handled by the ip-masq agent, which sets up the appropriate iptables rules. To have the ip-masq-agent ignore link local, you can set masqLinkLocal to true in the config map.
nonMasqueradeCIDRs:
- 10.0.0.0/8
resyncInterval: 60s
masqLinkLocal: true
This example demonstrates how to limit the amount of storage consumed in a namespace.
The following resources are used in the demonstration: ResourceQuota, LimitRange, and PersistentVolumeClaim.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The cluster-admin is operating a cluster on behalf of a user population and the admin wants to control how much storage a single namespace can consume in order to control cost.
The admin would like to limit:
Adding a LimitRange to a namespace enforces storage request sizes to a minimum and maximum. Storage is requested
via PersistentVolumeClaim. The admission controller that enforces limit ranges will reject any PVC that is above or below
the values set by the admin.
In this example, a PVC requesting 10Gi of storage would be rejected because it exceeds the 2Gi max.
apiVersion: v1
kind: LimitRange
metadata:
name: storagelimits
spec:
limits:
- type: PersistentVolumeClaim
max:
storage: 2Gi
min:
storage: 1Gi
Minimum storage requests are used when the underlying storage provider requires certain minimums. For example, AWS EBS volumes have a 1Gi minimum requirement.
Admins can limit the number of PVCs in a namespace as well as the cumulative capacity of those PVCs. New PVCs that exceed either maximum value will be rejected.
In this example, a 6th PVC in the namespace would be rejected because it exceeds the maximum count of 5. Alternatively, a 5Gi maximum quota when combined with the 2Gi max limit above, cannot have 3 PVCs where each has 2Gi. That would be 6Gi requested for a namespace capped at 5Gi.
apiVersion: v1
kind: ResourceQuota
metadata:
name: storagequota
spec:
hard:
persistentvolumeclaims: "5"
requests.storage: "5Gi"
A limit range can put a ceiling on how much storage is requested while a resource quota can effectively cap the storage consumed by a namespace through claim counts and cumulative storage capacity. The allows a cluster-admin to plan their cluster's storage budget without risk of any one project going over their allotment.
Kubernetes v1.22 [beta]
The cloud-controller-manager is a Kubernetes control plane component that embeds cloud-specific control logic. The cloud controller manager lets you link your cluster into your cloud provider's API, and separates out the components that interact with that cloud platform from components that only interact with your cluster.
By decoupling the interoperability logic between Kubernetes and the underlying cloud infrastructure, the cloud-controller-manager component enables cloud providers to release features at a different pace compared to the main Kubernetes project.
As part of the cloud provider extraction effort, all cloud specific controllers must be moved out of the kube-controller-manager. All existing clusters that run cloud controllers in the kube-controller-manager must migrate to instead run the controllers in a cloud provider specific cloud-controller-manager.
Leader Migration provides a mechanism in which HA clusters can safely migrate "cloud specific" controllers between the kube-controller-manager and the cloud-controller-manager via a shared resource lock between the two components while upgrading the replicated control plane. For a single-node control plane, or if unavailability of controller managers can be tolerated during the upgrade, Leader Migration is not needed and this guide can be ignored.
Leader Migration can be enabled by setting --enable-leader-migration on kube-controller-manager or cloud-controller-manager. Leader Migration only applies during the upgrade and can be safely disabled or left enabled after the upgrade is complete.
This guide walks you through the manual process of upgrading the control plane from kube-controller-manager with built-in cloud provider to running both kube-controller-manager and cloud-controller-manager. If you use a tool to administrator the cluster, please refer to the documentation of the tool and the cloud provider for more details.
It is assumed that the control plane is running Kubernetes version N and to be upgraded to version N + 1. Although it is possible to migrate within the same version, ideally the migration should be performed as part of an upgrade so that changes of configuration can be aligned to each release. The exact versions of N and N + 1 depend on each cloud provider. For example, if a cloud provider builds a cloud-controller-manager to work with Kubernetes 1.22, then N can be 1.21 and N + 1 can be 1.22.
The control plane nodes should run kube-controller-manager with Leader Election enabled through --leader-elect=true. As of version N, an in-tree cloud privider must be set with --cloud-provider flag and cloud-controller-manager should not yet be deployed.
The out-of-tree cloud provider must have built a cloud-controller-manager with Leader Migration implementation. If the cloud provider imports k8s.io/cloud-provider and k8s.io/controller-manager of version v0.21.0 or later, Leader Migration will be available. However, for version before v0.22.0, Leader Migration is alpha and requires feature gate ControllerManagerLeaderMigration to be enabled.
This guide assumes that kubelet of each control plane node starts kube-controller-manager and cloud-controller-manager as static pods defined by their manifests. If the components run in a different setting, please adjust the steps accordingly.
For authorization, this guide assumes that the cluster uses RBAC. If another authorization mode grants permissions to kube-controller-manager and cloud-controller-manager components, please grant the needed access in a way that matches the mode.
The default permissions of the controller manager allow only accesses to their main Lease. In order for the migration to work, accesses to another Lease are required.
You can grant kube-controller-manager full access to the leases API by modifying the system::leader-locking-kube-controller-manager role. This task guide assumes that the name of the migration lease is cloud-provider-extraction-migration.
kubectl patch -n kube-system role 'system::leader-locking-kube-controller-manager' -p '{"rules": [ {"apiGroups":[ "coordination.k8s.io"], "resources": ["leases"], "resourceNames": ["cloud-provider-extraction-migration"], "verbs": ["create", "list", "get", "update"] } ]}' --type=merge
Do the same to the system::leader-locking-cloud-controller-manager role.
kubectl patch -n kube-system role 'system::leader-locking-cloud-controller-manager' -p '{"rules": [ {"apiGroups":[ "coordination.k8s.io"], "resources": ["leases"], "resourceNames": ["cloud-provider-extraction-migration"], "verbs": ["create", "list", "get", "update"] } ]}' --type=merge
Leader Migration optionally takes a configuration file representing the state of controller-to-manager assignment. At this moment, with in-tree cloud provider, kube-controller-manager runs route, service, and cloud-node-lifecycle. The following example configuration shows the assignment.
Leader Migration can be enabled without a configuration. Please see Default Configuration for details.
kind: LeaderMigrationConfiguration
apiVersion: controllermanager.config.k8s.io/v1beta1
leaderName: cloud-provider-extraction-migration
resourceLock: leases
controllerLeaders:
- name: route
component: kube-controller-manager
- name: service
component: kube-controller-manager
- name: cloud-node-lifecycle
component: kube-controller-manager
On each control plane node, save the content to /etc/leadermigration.conf, and update the manifest of kube-controller-manager so that the file is mounted inside the container at the same location. Also, update the same manifest to add the following arguments:
--enable-leader-migration to enable Leader Migration on the controller manager--leader-migration-config=/etc/leadermigration.conf to set configuration fileRestart kube-controller-manager on each node. At this moment, kube-controller-manager has leader migration enabled and is ready for the migration.
In version N + 1, the desired state of controller-to-manager assignment can be represented by a new configuration file, shown as follows. Please note component field of each controllerLeaders changing from kube-controller-manager to cloud-controller-manager.
kind: LeaderMigrationConfiguration
apiVersion: controllermanager.config.k8s.io/v1beta1
leaderName: cloud-provider-extraction-migration
resourceLock: leases
controllerLeaders:
- name: route
component: cloud-controller-manager
- name: service
component: cloud-controller-manager
- name: cloud-node-lifecycle
component: cloud-controller-manager
When creating control plane nodes of version N + 1, the content should be deploy to /etc/leadermigration.conf. The manifest of cloud-controller-manager should be updated to mount the configuration file in the same manner as kube-controller-manager of version N. Similarly, add --feature-gates=ControllerManagerLeaderMigration=true,--enable-leader-migration, and --leader-migration-config=/etc/leadermigration.conf to the arguments of cloud-controller-manager.
Create a new control plane node of version N + 1 with the updated cloud-controller-manager manifest, and with the --cloud-provider flag unset for kube-controller-manager. kube-controller-manager of version N + 1 MUST NOT have Leader Migration enabled because, with an external cloud provider, it does not run the migrated controllers anymore and thus it is not involved in the migration.
Please refer to Cloud Controller Manager Administration for more detail on how to deploy cloud-controller-manager.
The control plane now contains nodes of both version N and N + 1. The nodes of version N run kube-controller-manager only, and these of version N + 1 run both kube-controller-manager and cloud-controller-manager. The migrated controllers, as specified in the configuration, are running under either kube-controller-manager of version N or cloud-controller-manager of version N + 1 depending on which controller manager holds the migration lease. No controller will ever be running under both controller managers at any time.
In a rolling manner, create a new control plane node of version N + 1 and bring down one of version N + 1 until the control plane contains only nodes of version N + 1.
If a rollback from version N + 1 to N is required, add nodes of version N with Leader Migration enabled for kube-controller-manager back to the control plane, replacing one of version N + 1 each time until there are only nodes of version N.
Now that the control plane has been upgraded to run both kube-controller-manager and cloud-controller-manager of version N + 1, Leader Migration has finished its job and can be safely disabled to save one Lease resource. It is safe to re-enable Leader Migration for the rollback in the future.
In a rolling manager, update manifest of cloud-controller-manager to unset both --enable-leader-migration and --leader-migration-config= flag, also remove the mount of /etc/leadermigration.conf, and finally remove /etc/leadermigration.conf. To re-enable Leader Migration, recreate the configuration file and add its mount and the flags that enable Leader Migration back to cloud-controller-manager.
Starting Kubernetes 1.22, Leader Migration provides a default configuration suitable for the default controller-to-manager assignment.
The default configuration can be enabled by setting --enable-leader-migration but without --leader-migration-config=.
For kube-controller-manager and cloud-controller-manager, if there are no flags that enable any in-tree cloud provider or change ownership of controllers, the default configuration can be used to avoid manual creation of the configuration file.
Kubernetes namespaces help different projects, teams, or customers to share a Kubernetes cluster.
It does this by providing the following:
Use of multiple namespaces is optional.
This example demonstrates how to use Kubernetes namespaces to subdivide your cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
This example assumes the following:
By default, a Kubernetes cluster will instantiate a default namespace when provisioning the cluster to hold the default set of Pods, Services, and Deployments used by the cluster.
Assuming you have a fresh cluster, you can inspect the available namespaces by doing the following:
kubectl get namespaces
NAME STATUS AGE
default Active 13m
For this exercise, we will create two additional Kubernetes namespaces to hold our content.
Let's imagine a scenario where an organization is using a shared Kubernetes cluster for development and production use cases.
The development team would like to maintain a space in the cluster where they can get a view on the list of Pods, Services, and Deployments they use to build and run their application. In this space, Kubernetes resources come and go, and the restrictions on who can or cannot modify resources are relaxed to enable agile development.
The operations team would like to maintain a space in the cluster where they can enforce strict procedures on who can or cannot manipulate the set of Pods, Services, and Deployments that run the production site.
One pattern this organization could follow is to partition the Kubernetes cluster into two namespaces: development and production.
Let's create two new namespaces to hold our work.
Use the file namespace-dev.json which describes a development namespace:
{
"apiVersion": "v1",
"kind": "Namespace",
"metadata": {
"name": "development",
"labels": {
"name": "development"
}
}
}
Create the development namespace using kubectl.
kubectl create -f https://k8s.io/examples/admin/namespace-dev.json
Save the following contents into file namespace-prod.json which describes a production namespace:
{
"apiVersion": "v1",
"kind": "Namespace",
"metadata": {
"name": "production",
"labels": {
"name": "production"
}
}
}
And then let's create the production namespace using kubectl.
kubectl create -f https://k8s.io/examples/admin/namespace-prod.json
To be sure things are right, let's list all of the namespaces in our cluster.
kubectl get namespaces --show-labels
NAME STATUS AGE LABELS
default Active 32m <none>
development Active 29s name=development
production Active 23s name=production
A Kubernetes namespace provides the scope for Pods, Services, and Deployments in the cluster.
Users interacting with one namespace do not see the content in another namespace.
To demonstrate this, let's spin up a simple Deployment and Pods in the development namespace.
We first check what is the current context:
kubectl config view
apiVersion: v1
clusters:
- cluster:
certificate-authority-data: REDACTED
server: https://130.211.122.180
name: lithe-cocoa-92103_kubernetes
contexts:
- context:
cluster: lithe-cocoa-92103_kubernetes
user: lithe-cocoa-92103_kubernetes
name: lithe-cocoa-92103_kubernetes
current-context: lithe-cocoa-92103_kubernetes
kind: Config
preferences: {}
users:
- name: lithe-cocoa-92103_kubernetes
user:
client-certificate-data: REDACTED
client-key-data: REDACTED
token: 65rZW78y8HbwXXtSXuUw9DbP4FLjHi4b
- name: lithe-cocoa-92103_kubernetes-basic-auth
user:
password: h5M0FtUUIflBSdI7
username: admin
kubectl config current-context
lithe-cocoa-92103_kubernetes
The next step is to define a context for the kubectl client to work in each namespace. The value of "cluster" and "user" fields are copied from the current context.
kubectl config set-context dev --namespace=development \
--cluster=lithe-cocoa-92103_kubernetes \
--user=lithe-cocoa-92103_kubernetes
kubectl config set-context prod --namespace=production \
--cluster=lithe-cocoa-92103_kubernetes \
--user=lithe-cocoa-92103_kubernetes
By default, the above commands adds two contexts that are saved into file
.kube/config. You can now view the contexts and alternate against the two
new request contexts depending on which namespace you wish to work against.
To view the new contexts:
kubectl config view
apiVersion: v1
clusters:
- cluster:
certificate-authority-data: REDACTED
server: https://130.211.122.180
name: lithe-cocoa-92103_kubernetes
contexts:
- context:
cluster: lithe-cocoa-92103_kubernetes
user: lithe-cocoa-92103_kubernetes
name: lithe-cocoa-92103_kubernetes
- context:
cluster: lithe-cocoa-92103_kubernetes
namespace: development
user: lithe-cocoa-92103_kubernetes
name: dev
- context:
cluster: lithe-cocoa-92103_kubernetes
namespace: production
user: lithe-cocoa-92103_kubernetes
name: prod
current-context: lithe-cocoa-92103_kubernetes
kind: Config
preferences: {}
users:
- name: lithe-cocoa-92103_kubernetes
user:
client-certificate-data: REDACTED
client-key-data: REDACTED
token: 65rZW78y8HbwXXtSXuUw9DbP4FLjHi4b
- name: lithe-cocoa-92103_kubernetes-basic-auth
user:
password: h5M0FtUUIflBSdI7
username: admin
Let's switch to operate in the development namespace.
kubectl config use-context dev
You can verify your current context by doing the following:
kubectl config current-context
dev
At this point, all requests we make to the Kubernetes cluster from the command line are scoped to the development namespace.
Let's create some contents.
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: snowflake
name: snowflake
spec:
replicas: 2
selector:
matchLabels:
app: snowflake
template:
metadata:
labels:
app: snowflake
spec:
containers:
- image: k8s.gcr.io/serve_hostname
imagePullPolicy: Always
name: snowflake
Apply the manifest to create a Deployment
kubectl apply -f https://k8s.io/examples/admin/snowflake-deployment.yaml
We have created a deployment whose replica size is 2 that is running the pod called snowflake with a basic container that serves the hostname.
kubectl get deployment
NAME READY UP-TO-DATE AVAILABLE AGE
snowflake 2/2 2 2 2m
kubectl get pods -l app=snowflake
NAME READY STATUS RESTARTS AGE
snowflake-3968820950-9dgr8 1/1 Running 0 2m
snowflake-3968820950-vgc4n 1/1 Running 0 2m
And this is great, developers are able to do what they want, and they do not have to worry about affecting content in the production namespace.
Let's switch to the production namespace and show how resources in one namespace are hidden from the other.
kubectl config use-context prod
The production namespace should be empty, and the following commands should return nothing.
kubectl get deployment
kubectl get pods
Production likes to run cattle, so let's create some cattle pods.
kubectl create deployment cattle --image=k8s.gcr.io/serve_hostname --replicas=5
kubectl get deployment
NAME READY UP-TO-DATE AVAILABLE AGE
cattle 5/5 5 5 10s
kubectl get pods -l app=cattle
NAME READY STATUS RESTARTS AGE
cattle-2263376956-41xy6 1/1 Running 0 34s
cattle-2263376956-kw466 1/1 Running 0 34s
cattle-2263376956-n4v97 1/1 Running 0 34s
cattle-2263376956-p5p3i 1/1 Running 0 34s
cattle-2263376956-sxpth 1/1 Running 0 34s
At this point, it should be clear that the resources users create in one namespace are hidden from the other namespace.
As the policy support in Kubernetes evolves, we will extend this scenario to show how you can provide different authorization rules for each namespace.
etcd is a consistent and highly-available key value store used as Kubernetes' backing store for all cluster data.
If your Kubernetes cluster uses etcd as its backing store, make sure you have a back up plan for those data.
You can find in-depth information about etcd in the official documentation.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Run etcd as a cluster of odd members.
etcd is a leader-based distributed system. Ensure that the leader periodically send heartbeats on time to all followers to keep the cluster stable.
Ensure that no resource starvation occurs.
Performance and stability of the cluster is sensitive to network and disk I/O. Any resource starvation can lead to heartbeat timeout, causing instability of the cluster. An unstable etcd indicates that no leader is elected. Under such circumstances, a cluster cannot make any changes to its current state, which implies no new pods can be scheduled.
Keeping etcd clusters stable is critical to the stability of Kubernetes clusters. Therefore, run etcd clusters on dedicated machines or isolated environments for guaranteed resource requirements.
The minimum recommended version of etcd to run in production is 3.2.10+.
Operating etcd with limited resources is suitable only for testing purposes. For deploying in production, advanced hardware configuration is required. Before deploying etcd in production, see resource requirement reference.
This section covers starting a single-node and multi-node etcd cluster.
Use a single-node etcd cluster only for testing purpose.
Run the following:
etcd --listen-client-urls=http://$PRIVATE_IP:2379 \
--advertise-client-urls=http://$PRIVATE_IP:2379
Start the Kubernetes API server with the flag
--etcd-servers=$PRIVATE_IP:2379.
Make sure PRIVATE_IP is set to your etcd client IP.
For durability and high availability, run etcd as a multi-node cluster in production and back it up periodically. A five-member cluster is recommended in production. For more information, see FAQ documentation.
Configure an etcd cluster either by static member information or by dynamic discovery. For more information on clustering, see etcd clustering documentation.
For an example, consider a five-member etcd cluster running with the following
client URLs: http://$IP1:2379, http://$IP2:2379, http://$IP3:2379,
http://$IP4:2379, and http://$IP5:2379. To start a Kubernetes API server:
Run the following:
etcd --listen-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379 --advertise-client-urls=http://$IP1:2379,http://$IP2:2379,http://$IP3:2379,http://$IP4:2379,http://$IP5:2379
Start the Kubernetes API servers with the flag
--etcd-servers=$IP1:2379,$IP2:2379,$IP3:2379,$IP4:2379,$IP5:2379.
Make sure the IP<n> variables are set to your client IP addresses.
To run a load balancing etcd cluster:
$LB.--etcd-servers=$LB:2379.Access to etcd is equivalent to root permission in the cluster so ideally only the API server should have access to it. Considering the sensitivity of the data, it is recommended to grant permission to only those nodes that require access to etcd clusters.
To secure etcd, either set up firewall rules or use the security features
provided by etcd. etcd security features depend on x509 Public Key
Infrastructure (PKI). To begin, establish secure communication channels by
generating a key and certificate pair. For example, use key pairs peer.key
and peer.cert for securing communication between etcd members, and
client.key and client.cert for securing communication between etcd and its
clients. See the example scripts
provided by the etcd project to generate key pairs and CA files for client
authentication.
To configure etcd with secure peer communication, specify flags
--peer-key-file=peer.key and --peer-cert-file=peer.cert, and use HTTPS as
the URL schema.
Similarly, to configure etcd with secure client communication, specify flags
--key-file=k8sclient.key and --cert-file=k8sclient.cert, and use HTTPS as
the URL schema. Here is an example on a client command that uses secure
communication:
ETCDCTL_API=3 etcdctl --endpoints 10.2.0.9:2379 \
--cert=/etc/kubernetes/pki/etcd/server.crt \
--key=/etc/kubernetes/pki/etcd/server.key \
--cacert=/etc/kubernetes/pki/etcd/ca.crt \
member list
After configuring secure communication, restrict the access of etcd cluster to only the Kubernetes API servers. Use TLS authentication to do so.
For example, consider key pairs k8sclient.key and k8sclient.cert that are
trusted by the CA etcd.ca. When etcd is configured with --client-cert-auth
along with TLS, it verifies the certificates from clients by using system CAs
or the CA passed in by --trusted-ca-file flag. Specifying flags
--client-cert-auth=true and --trusted-ca-file=etcd.ca will restrict the
access to clients with the certificate k8sclient.cert.
Once etcd is configured correctly, only clients with valid certificates can
access it. To give Kubernetes API servers the access, configure them with the
flags --etcd-certfile=k8sclient.cert,--etcd-keyfile=k8sclient.key and
--etcd-cafile=ca.cert.
etcd cluster achieves high availability by tolerating minor member failures. However, to improve the overall health of the cluster, replace failed members immediately. When multiple members fail, replace them one by one. Replacing a failed member involves two steps: removing the failed member and adding a new member.
Though etcd keeps unique member IDs internally, it is recommended to use a
unique name for each member to avoid human errors. For example, consider a
three-member etcd cluster. Let the URLs be, member1=http://10.0.0.1,
member2=http://10.0.0.2, and member3=http://10.0.0.3. When member1 fails,
replace it with member4=http://10.0.0.4.
Get the member ID of the failed member1:
etcdctl --endpoints=http://10.0.0.2,http://10.0.0.3 member list
The following message is displayed:
8211f1d0f64f3269, started, member1, http://10.0.0.1:2380, http://10.0.0.1:2379
91bc3c398fb3c146, started, member2, http://10.0.0.2:2380, http://10.0.0.2:2379
fd422379fda50e48, started, member3, http://10.0.0.3:2380, http://10.0.0.3:2379
Remove the failed member:
etcdctl member remove 8211f1d0f64f3269
The following message is displayed:
Removed member 8211f1d0f64f3269 from cluster
Add the new member:
etcdctl member add member4 --peer-urls=http://10.0.0.4:2380
The following message is displayed:
Member 2be1eb8f84b7f63e added to cluster ef37ad9dc622a7c4
Start the newly added member on a machine with the IP 10.0.0.4:
export ETCD_NAME="member4"
export ETCD_INITIAL_CLUSTER="member2=http://10.0.0.2:2380,member3=http://10.0.0.3:2380,member4=http://10.0.0.4:2380"
export ETCD_INITIAL_CLUSTER_STATE=existing
etcd [flags]
Do either of the following:
--etcd-servers flag for the Kubernetes API servers to make
Kubernetes aware of the configuration changes, then restart the
Kubernetes API servers.For more information on cluster reconfiguration, see etcd reconfiguration documentation.
All Kubernetes objects are stored on etcd. Periodically backing up the etcd cluster data is important to recover Kubernetes clusters under disaster scenarios, such as losing all control plane nodes. The snapshot file contains all the Kubernetes states and critical information. In order to keep the sensitive Kubernetes data safe, encrypt the snapshot files.
Backing up an etcd cluster can be accomplished in two ways: etcd built-in snapshot and volume snapshot.
etcd supports built-in snapshot. A snapshot may either be taken from a live
member with the etcdctl snapshot save command or by copying the
member/snap/db file from an etcd
data directory
that is not currently used by an etcd process. Taking the snapshot will
not affect the performance of the member.
Below is an example for taking a snapshot of the keyspace served by
$ENDPOINT to the file snapshotdb:
ETCDCTL_API=3 etcdctl --endpoints $ENDPOINT snapshot save snapshotdb
Verify the snapshot:
ETCDCTL_API=3 etcdctl --write-out=table snapshot status snapshotdb
+----------+----------+------------+------------+
| HASH | REVISION | TOTAL KEYS | TOTAL SIZE |
+----------+----------+------------+------------+
| fe01cf57 | 10 | 7 | 2.1 MB |
+----------+----------+------------+------------+
If etcd is running on a storage volume that supports backup, such as Amazon Elastic Block Store, back up etcd data by taking a snapshot of the storage volume.
We can also take the snapshot using various options given by etcdctl. For example
ETCDCTL_API=3 etcdctl -h
will list various options available from etcdctl. For example, you can take a snapshot by specifying the endpoint, certificates etc as shown below:
ETCDCTL_API=3 etcdctl --endpoints=https://127.0.0.1:2379 \
--cacert=<trusted-ca-file> --cert=<cert-file> --key=<key-file> \
snapshot save <backup-file-location>
where trusted-ca-file, cert-file and key-file can be obtained from the description of the etcd Pod.
Scaling up etcd clusters increases availability by trading off performance. Scaling does not increase cluster performance nor capability. A general rule is not to scale up or down etcd clusters. Do not configure any auto scaling groups for etcd clusters. It is highly recommended to always run a static five-member etcd cluster for production Kubernetes clusters at any officially supported scale.
A reasonable scaling is to upgrade a three-member cluster to a five-member one, when more reliability is desired. See etcd reconfiguration documentation for information on how to add members into an existing cluster.
etcd supports restoring from snapshots that are taken from an etcd process of the major.minor version. Restoring a version from a different patch version of etcd also is supported. A restore operation is employed to recover the data of a failed cluster.
Before starting the restore operation, a snapshot file must be present. It can either be a snapshot file from a previous backup operation, or from a remaining data directory. Here is an example:
ETCDCTL_API=3 etcdctl --endpoints 10.2.0.9:2379 snapshot restore snapshotdb
Another example for restoring using etcdutl options:
ETCDCTL_API=3 etcdutl --data-dir <data-dir-location> snapshot restore snapshotdb
For more information and examples on restoring a cluster from a snapshot file, see etcd disaster recovery documentation.
If the access URLs of the restored cluster is changed from the previous
cluster, the Kubernetes API server must be reconfigured accordingly. In this
case, restart Kubernetes API servers with the flag
--etcd-servers=$NEW_ETCD_CLUSTER instead of the flag
--etcd-servers=$OLD_ETCD_CLUSTER. Replace $NEW_ETCD_CLUSTER and
$OLD_ETCD_CLUSTER with the respective IP addresses. If a load balancer is
used in front of an etcd cluster, you might need to update the load balancer
instead.
If the majority of etcd members have permanently failed, the etcd cluster is considered failed. In this scenario, Kubernetes cannot make any changes to its current state. Although the scheduled pods might continue to run, no new pods can be scheduled. In such cases, recover the etcd cluster and potentially reconfigure Kubernetes API servers to fix the issue.
If any API servers are running in your cluster, you should not attempt to restore instances of etcd. Instead, follow these steps to restore etcd:
We also recommend restarting any components (e.g. kube-scheduler,
kube-controller-manager, kubelet) to ensure that they don't rely on some
stale data. Note that in practice, the restore takes a bit of time. During the
restoration, critical components will lose leader lock and restart themselves.
Kubernetes v1.22 [deprecated]
Dynamic Kubelet Configuration allows you to change the configuration of each kubelet in a running Kubernetes cluster, by deploying a ConfigMap and configuring each Node to use it.
KubeletConfiguration.
You need to have a Kubernetes cluster.
You also need kubectl v1.11 or higher, configured to communicate with your cluster.
Your Kubernetes server must be at or later than version v1.11.
To check the version, enter kubectl version.
Your cluster API server version (eg v1.12) must be no more than one minor
version away from the version of kubectl that you are using. For example,
if your cluster is running v1.16 then you can use kubectl v1.15, v1.16
or v1.17; other combinations
aren't supported.
Some of the examples use the command line tool
jq. You do not need jq to complete the task,
because there are manual alternatives.
For each node that you're reconfiguring, you must set the kubelet
--dynamic-config-dir flag to a writable directory.
The basic workflow for configuring a kubelet in a live cluster is as follows:
Each kubelet watches a configuration reference on its respective Node object. When this reference changes, the kubelet downloads the new configuration, updates a local reference to refer to the file, and exits. For the feature to work correctly, you must be running an OS-level service manager (such as systemd), which will restart the kubelet if it exits. When the kubelet is restarted, it will begin using the new configuration.
The new configuration completely overrides configuration provided by --config,
and is overridden by command-line flags. Unspecified values in the new configuration
will receive default values appropriate to the configuration version
(e.g. kubelet.config.k8s.io/v1beta1), unless overridden by flags.
The status of the Node's kubelet configuration is reported via
Node.Status.Config. Once you have updated a Node to use the new
ConfigMap, you can observe this status to confirm that the Node is using the
intended configuration.
This document describes editing Nodes using kubectl edit.
There are other ways to modify a Node's spec, including kubectl patch, for
example, which facilitate scripted workflows.
This document only describes a single Node consuming each ConfigMap. Keep in mind that it is also valid for multiple Nodes to consume the same ConfigMap.
kubectl's --append-hash option,
and incrementally roll out updates to Node.Spec.ConfigSource.
Previously, you were required to manually create RBAC rules to allow Nodes to access their assigned ConfigMaps. The Node Authorizer now automatically configures these rules.
The Dynamic Kubelet Configuration feature allows you to provide an override for the entire configuration object, rather than a per-field overlay. This is a simpler model that makes it easier to trace the source of configuration values and debug issues. The compromise, however, is that you must start with knowledge of the existing configuration to ensure that you only change the fields you intend to change.
The kubelet loads settings from its configuration file, but you can set command line flags to override the configuration in the file. This means that if you only know the contents of the configuration file, and you don't know the command line overrides, then you do not know the running configuration either.
Because you need to know the running configuration in order to override it,
you can fetch the running configuration from the kubelet. You can generate a
config file containing a Node's current configuration by accessing the kubelet's
configz endpoint, through kubectl proxy. The next section explains how to
do this.
configz endpoint is there to help with debugging, and is not
a stable part of kubelet behavior.
Do not rely on the behavior of this endpoint for production scenarios or for
use with automated tools.
For more information on configuring the kubelet via a configuration file, see Set kubelet parameters via a config file).
jq command to streamline working with JSON.
To follow the tasks as written, you need to have jq installed. You can
adapt the steps if you prefer to extract the kubeletconfig subobject manually.
Choose a Node to reconfigure. In this example, the name of this Node is
referred to as NODE_NAME.
Start the kubectl proxy in the background using the following command:
kubectl proxy --port=8001 &
Run the following command to download and unpack the configuration from the
configz endpoint. The command is long, so be careful when copying and
pasting. If you use zsh, note that common zsh configurations add backslashes
to escape the opening and closing curly braces around the variable name in the URL.
For example: ${NODE_NAME} will be rewritten as $\{NODE_NAME\} during the paste.
You must remove the backslashes before running the command, or the command will fail.
NODE_NAME="the-name-of-the-node-you-are-reconfiguring"; curl -sSL "http://localhost:8001/api/v1/nodes/${NODE_NAME}/proxy/configz" | jq '.kubeletconfig|.kind="KubeletConfiguration"|.apiVersion="kubelet.config.k8s.io/v1beta1"' > kubelet_configz_${NODE_NAME}
kind and apiVersion to the downloaded
object, because those fields are not reported by the configz endpoint.
Using a text editor, change one of the parameters in the
file generated by the previous procedure. For example, you
might edit the parameter eventRecordQPS, that controls
rate limiting for event recording.
Push the edited configuration file to the control plane with the following command:
kubectl -n kube-system create configmap my-node-config --from-file=kubelet=kubelet_configz_${NODE_NAME} --append-hash -o yaml
This is an example of a valid response:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2017-09-14T20:23:33Z
name: my-node-config-gkt4c2m4b2
namespace: kube-system
resourceVersion: "119980"
uid: 946d785e-998a-11e7-a8dd-42010a800006
data:
kubelet: |
{...}
You created that ConfigMap inside the kube-system namespace because the kubelet
is a Kubernetes system component.
The --append-hash option appends a short checksum of the ConfigMap contents
to the name. This is convenient for an edit-then-push workflow, because it
automatically, yet deterministically, generates new names for new resources.
The name that includes this generated hash is referred to as CONFIG_MAP_NAME
in the following examples.
Edit the Node's reference to point to the new ConfigMap with the following command:
kubectl edit node ${NODE_NAME}
In your text editor, add the following YAML under spec:
configSource:
configMap:
name: CONFIG_MAP_NAME # replace CONFIG_MAP_NAME with the name of the ConfigMap
namespace: kube-system
kubeletConfigKey: kubelet
You must specify all three of name, namespace, and kubeletConfigKey.
The kubeletConfigKey parameter shows the kubelet which key of the ConfigMap
contains its config.
Retrieve the Node using the kubectl get node ${NODE_NAME} -o yaml command and inspect
Node.Status.Config. The config sources corresponding to the active,
assigned, and lastKnownGood configurations are reported in the status.
active configuration is the version the kubelet is currently running with.assigned configuration is the latest version the kubelet has resolved based on
Node.Spec.ConfigSource.lastKnownGood configuration is the version the
kubelet will fall back to if an invalid config is assigned in Node.Spec.ConfigSource.ThelastKnownGood configuration might not be present if it is set to its default value,
the local config deployed with the node. The status will update lastKnownGood to
match a valid assigned config after the kubelet becomes comfortable with the config.
The details of how the kubelet determines a config should become the lastKnownGood are
not guaranteed by the API, but is currently implemented as a 10-minute grace period.
You can use the following command (using jq) to filter down
to the config status:
kubectl get no ${NODE_NAME} -o json | jq '.status.config'
The following is an example response:
{
"active": {
"configMap": {
"kubeletConfigKey": "kubelet",
"name": "my-node-config-9mbkccg2cc",
"namespace": "kube-system",
"resourceVersion": "1326",
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002"
}
},
"assigned": {
"configMap": {
"kubeletConfigKey": "kubelet",
"name": "my-node-config-9mbkccg2cc",
"namespace": "kube-system",
"resourceVersion": "1326",
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002"
}
},
"lastKnownGood": {
"configMap": {
"kubeletConfigKey": "kubelet",
"name": "my-node-config-9mbkccg2cc",
"namespace": "kube-system",
"resourceVersion": "1326",
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002"
}
}
}
(if you do not have jq, you can look at the whole response and find Node.Status.Config
by eye).
If an error occurs, the kubelet reports it in the Node.Status.Config.Error
structure. Possible errors are listed in
Understanding Node.Status.Config.Error messages.
You can search for the identical text in the kubelet log for additional details
and context about the error.
Follow the workflow above to make more changes and push them again. Each time
you push a ConfigMap with new contents, the --append-hash kubectl option creates
the ConfigMap with a new name. The safest rollout strategy is to first create a
new ConfigMap, and then update the Node to use the new ConfigMap.
To reset the Node to use the configuration it was provisioned with, edit the
Node using kubectl edit node ${NODE_NAME} and remove the
Node.Spec.ConfigSource field.
After removing this subfield, Node.Status.Config eventually becomes
empty, since all config sources have been reset to nil, which indicates that
the local default config is assigned, active, and lastKnownGood, and no
error is reported.
kubectl patch exampleYou can change a Node's configSource using several different mechanisms.
This example uses kubectl patch:
kubectl patch node ${NODE_NAME} -p "{\"spec\":{\"configSource\":{\"configMap\":{\"name\":\"${CONFIG_MAP_NAME}\",\"namespace\":\"kube-system\",\"kubeletConfigKey\":\"kubelet\"}}}}"
When a new config is assigned to the Node, the kubelet downloads and unpacks the
config payload as a set of files on the local disk. The kubelet also records metadata
that locally tracks the assigned and last-known-good config sources, so that the
kubelet knows which config to use across restarts, even if the API server becomes
unavailable. After checkpointing a config and the relevant metadata, the kubelet
exits if it detects that the assigned config has changed. When the kubelet is
restarted by the OS-level service manager (such as systemd), it reads the new
metadata and uses the new config.
The recorded metadata is fully resolved, meaning that it contains all necessary
information to choose a specific config version - typically a UID and ResourceVersion.
This is in contrast to Node.Spec.ConfigSource, where the intended config is declared
via the idempotent namespace/name that identifies the target ConfigMap; the kubelet
tries to use the latest version of this ConfigMap.
When you are debugging problems on a node, you can inspect the kubelet's config metadata and checkpoints. The structure of the kubelet's checkpointing directory is:
- --dynamic-config-dir (root for managing dynamic config)
| - meta
| - assigned (encoded kubeletconfig/v1beta1.SerializedNodeConfigSource object, indicating the assigned config)
| - last-known-good (encoded kubeletconfig/v1beta1.SerializedNodeConfigSource object, indicating the last-known-good config)
| - checkpoints
| - uid1 (dir for versions of object identified by uid1)
| - resourceVersion1 (dir for unpacked files from resourceVersion1 of object with uid1)
| - ...
| - ...
Node.Status.Config.Error messagesThe following table describes error messages that can occur when using Dynamic Kubelet Config. You can search for the identical text in the Kubelet log for additional details and context about the error.
| Error Message | Possible Causes |
|---|---|
| failed to load config, see Kubelet log for details | The kubelet likely could not parse the downloaded config payload, or encountered a filesystem error attempting to load the payload from disk. |
| failed to validate config, see Kubelet log for details | The configuration in the payload, combined with any command-line flag overrides, and the sum of feature gates from flags, the config file, and the remote payload, was determined to be invalid by the kubelet. |
| invalid NodeConfigSource, exactly one subfield must be non-nil, but all were nil | Since Node.Spec.ConfigSource is validated by the API server to contain at least one non-nil subfield, this likely means that the kubelet is older than the API server and does not recognize a newer source type. |
| failed to sync: failed to download config, see Kubelet log for details | The kubelet could not download the config. It is possible that Node.Spec.ConfigSource could not be resolved to a concrete API object, or that network errors disrupted the download attempt. The kubelet will retry the download when in this error state. |
| failed to sync: internal failure, see Kubelet log for details | The kubelet encountered some internal problem and failed to update its config as a result. Examples include filesystem errors and reading objects from the internal informer cache. |
| internal failure, see Kubelet log for details | The kubelet encountered some internal problem while manipulating config, outside of the configuration sync loop. |
NodeConfigSourceKubeletConfiguration
reference.Kubernetes nodes can be scheduled to Capacity. Pods can consume all the
available capacity on a node by default. This is an issue because nodes
typically run quite a few system daemons that power the OS and Kubernetes
itself. Unless resources are set aside for these system daemons, pods and system
daemons compete for resources and lead to resource starvation issues on the
node.
The kubelet exposes a feature named 'Node Allocatable' that helps to reserve
compute resources for system daemons. Kubernetes recommends cluster
administrators to configure 'Node Allocatable' based on their workload density
on each node.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.8. To check the version, enterkubectl version.
Your Kubernetes server must be at or later than version 1.17 to use
the kubelet command line option --reserved-cpus to set an
explicitly reserved CPU list.
'Allocatable' on a Kubernetes node is defined as the amount of compute resources that are available for pods. The scheduler does not over-subscribe 'Allocatable'. 'CPU', 'memory' and 'ephemeral-storage' are supported as of now.
Node Allocatable is exposed as part of v1.Node object in the API and as part
of kubectl describe node in the CLI.
Resources can be reserved for two categories of system daemons in the kubelet.
To properly enforce node allocatable constraints on the node, you must
enable the new cgroup hierarchy via the --cgroups-per-qos flag. This flag is
enabled by default. When enabled, the kubelet will parent all end-user pods
under a cgroup hierarchy managed by the kubelet.
The kubelet supports manipulation of the cgroup hierarchy on
the host using a cgroup driver. The driver is configured via the
--cgroup-driver flag.
The supported values are the following:
cgroupfs is the default driver that performs direct manipulation of the
cgroup filesystem on the host in order to manage cgroup sandboxes.systemd is an alternative driver that manages cgroup sandboxes using
transient slices for resources that are supported by that init system.Depending on the configuration of the associated container runtime,
operators may have to choose a particular cgroup driver to ensure
proper system behavior. For example, if operators use the systemd
cgroup driver provided by the docker runtime, the kubelet must
be configured to use the systemd cgroup driver.
--kube-reserved=[cpu=100m][,][memory=100Mi][,][ephemeral-storage=1Gi][,][pid=1000]--kube-reserved-cgroup=kube-reserved is meant to capture resource reservation for kubernetes system
daemons like the kubelet, container runtime, node problem detector, etc.
It is not meant to reserve resources for system daemons that are run as pods.
kube-reserved is typically a function of pod density on the nodes.
In addition to cpu, memory, and ephemeral-storage, pid may be
specified to reserve the specified number of process IDs for
kubernetes system daemons.
To optionally enforce kube-reserved on kubernetes system daemons, specify the parent
control group for kube daemons as the value for --kube-reserved-cgroup kubelet
flag.
It is recommended that the kubernetes system daemons are placed under a top
level control group (runtime.slice on systemd machines for example). Each
system daemon should ideally run within its own child control group. Refer to
the design proposal
for more details on recommended control group hierarchy.
Note that Kubelet does not create --kube-reserved-cgroup if it doesn't
exist. Kubelet will fail if an invalid cgroup is specified.
--system-reserved=[cpu=100m][,][memory=100Mi][,][ephemeral-storage=1Gi][,][pid=1000]--system-reserved-cgroup=system-reserved is meant to capture resource reservation for OS system daemons
like sshd, udev, etc. system-reserved should reserve memory for the
kernel too since kernel memory is not accounted to pods in Kubernetes at this time.
Reserving resources for user login sessions is also recommended (user.slice in
systemd world).
In addition to cpu, memory, and ephemeral-storage, pid may be
specified to reserve the specified number of process IDs for OS system
daemons.
To optionally enforce system-reserved on system daemons, specify the parent
control group for OS system daemons as the value for --system-reserved-cgroup
kubelet flag.
It is recommended that the OS system daemons are placed under a top level
control group (system.slice on systemd machines for example).
Note that kubelet does not create --system-reserved-cgroup if it doesn't
exist. kubelet will fail if an invalid cgroup is specified.
Kubernetes v1.17 [stable]
Kubelet Flag: --reserved-cpus=0-3
reserved-cpus is meant to define an explicit CPU set for OS system daemons and
kubernetes system daemons. reserved-cpus is for systems that do not intend to
define separate top level cgroups for OS system daemons and kubernetes system daemons
with regard to cpuset resource.
If the Kubelet does not have --system-reserved-cgroup and --kube-reserved-cgroup,
the explicit cpuset provided by reserved-cpus will take precedence over the CPUs
defined by --kube-reserved and --system-reserved options.
This option is specifically designed for Telco/NFV use cases where uncontrolled interrupts/timers may impact the workload performance. you can use this option to define the explicit cpuset for the system/kubernetes daemons as well as the interrupts/timers, so the rest CPUs on the system can be used exclusively for workloads, with less impact from uncontrolled interrupts/timers. To move the system daemon, kubernetes daemons and interrupts/timers to the explicit cpuset defined by this option, other mechanism outside Kubernetes should be used. For example: in Centos, you can do this using the tuned toolset.
Kubelet Flag: --eviction-hard=[memory.available<500Mi]
Memory pressure at the node level leads to System OOMs which affects the entire
node and all pods running on it. Nodes can go offline temporarily until memory
has been reclaimed. To avoid (or reduce the probability of) system OOMs kubelet
provides out of resource
management. Evictions are
supported for memory and ephemeral-storage only. By reserving some memory via
--eviction-hard flag, the kubelet attempts to evict pods whenever memory
availability on the node drops below the reserved value. Hypothetically, if
system daemons did not exist on a node, pods cannot use more than capacity - eviction-hard. For this reason, resources reserved for evictions are not
available for pods.
Kubelet Flag: --enforce-node-allocatable=pods[,][system-reserved][,][kube-reserved]
The scheduler treats 'Allocatable' as the available capacity for pods.
kubelet enforce 'Allocatable' across pods by default. Enforcement is performed
by evicting pods whenever the overall usage across all pods exceeds
'Allocatable'. More details on eviction policy can be found
on the node pressure eviction
page. This enforcement is controlled by
specifying pods value to the kubelet flag --enforce-node-allocatable.
Optionally, kubelet can be made to enforce kube-reserved and
system-reserved by specifying kube-reserved & system-reserved values in
the same flag. Note that to enforce kube-reserved or system-reserved,
--kube-reserved-cgroup or --system-reserved-cgroup needs to be specified
respectively.
System daemons are expected to be treated similar to 'Guaranteed' pods. System
daemons can burst within their bounding control groups and this behavior needs
to be managed as part of kubernetes deployments. For example, kubelet should
have its own control group and share kube-reserved resources with the
container runtime. However, Kubelet cannot burst and use up all available Node
resources if kube-reserved is enforced.
Be extra careful while enforcing system-reserved reservation since it can lead
to critical system services being CPU starved, OOM killed, or unable
to fork on the node. The
recommendation is to enforce system-reserved only if a user has profiled their
nodes exhaustively to come up with precise estimates and is confident in their
ability to recover if any process in that group is oom-killed.
pods.kube-reserved based on usage heuristics.system-reserved over time.The resource requirements of kube system daemons may grow over time as more and
more features are added. Over time, kubernetes project will attempt to bring
down utilization of node system daemons, but that is not a priority as of now.
So expect a drop in Allocatable capacity in future releases.
Here is an example to illustrate Node Allocatable computation:
32Gi of memory, 16 CPUs and 100Gi of Storage--kube-reserved is set to cpu=1,memory=2Gi,ephemeral-storage=1Gi--system-reserved is set to cpu=500m,memory=1Gi,ephemeral-storage=1Gi--eviction-hard is set to memory.available<500Mi,nodefs.available<10%Under this scenario, 'Allocatable' will be 14.5 CPUs, 28.5Gi of memory and
88Gi of local storage.
Scheduler ensures that the total memory requests across all pods on this node does
not exceed 28.5Gi and storage doesn't exceed 88Gi.
Kubelet evicts pods whenever the overall memory usage across pods exceeds 28.5Gi,
or if overall disk usage exceeds 88Gi If all processes on the node consume as
much CPU as they can, pods together cannot consume more than 14.5 CPUs.
If kube-reserved and/or system-reserved is not enforced and system daemons
exceed their reservation, kubelet evicts pods whenever the overall node memory
usage is higher than 31.5Gi or storage is greater than 90Gi.
Kubernetes v1.22 [alpha]
This document describes how to run Kubernetes Node components such as kubelet, CRI, OCI, and CNI without root privileges, by using a user namespace.
This technique is also known as rootless mode.
This document describes how to run Kubernetes Node components (and hence pods) a non-root user.
If you are just looking for how to run a pod as a non-root user, see SecurityContext.
Your Kubernetes server must be at or later than version 1.22.
To check the version, enter kubectl version.
/etc/subuid and /etc/subgidKubeletInUserNamespace feature gatekind supports running Kubernetes inside Rootless Docker or Rootless Podman.
See Running kind with Rootless Docker.
minikube also supports running Kubernetes inside Rootless Docker.
See the page about the docker driver in the Minikube documentation.
Rootless Podman is not supported.
K3s experimentally supports rootless mode.
See Running K3s with Rootless mode for the usage.
Usernetes is a reference distribution of Kubernetes that can be installed under $HOME directory without the root privilege.
Usernetes supports both containerd and CRI-O as CRI runtimes. Usernetes supports multi-node clusters using Flannel (VXLAN).
See the Usernetes repo for the usage.
This section provides hints for running Kubernetes in a user namespace manually.
The first step is to create a user namespace.
If you are trying to run Kubernetes in a user-namespaced container such as Rootless Docker/Podman or LXC/LXD, you are all set, and you can go to the next subsection.
Otherwise you have to create a user namespace by yourself, by calling unshare(2) with CLONE_NEWUSER.
A user namespace can be also unshared by using command line tools such as:
After unsharing the user namespace, you will also have to unshare other namespaces such as mount namespace.
You do not need to call chroot() nor pivot_root() after unsharing the mount namespace,
however, you have to mount writable filesystems on several directories in the namespace.
At least, the following directories need to be writable in the namespace (not outside the namespace):
/etc/run/var/logs/var/lib/kubelet/var/lib/cni/var/lib/containerd (for containerd)/var/lib/containers (for CRI-O)In addition to the user namespace, you also need to have a writable cgroup tree with cgroup v2.
If you are trying to run Kubernetes in Rootless Docker/Podman or LXC/LXD on a systemd-based host, you are all set.
Otherwise you have to create a systemd unit with Delegate=yes property to delegate a cgroup tree with writable permission.
On your node, systemd must already be configured to allow delegation; for more details, see cgroup v2 in the Rootless Containers documentation.
The network namespace of the Node components has to have a non-loopback interface, which can be for example configured with slirp4netns, VPNKit, or lxc-user-nic(1).
The network namespaces of the Pods can be configured with regular CNI plugins. For multi-node networking, Flannel (VXLAN, 8472/UDP) is known to work.
Ports such as the kubelet port (10250/TCP) and NodePort service ports have to be exposed from the Node network namespace to
the host with an external port forwarder, such as RootlessKit, slirp4netns, or
socat(1).
You can use the port forwarder from K3s. See Running K3s in Rootless Mode for more details.
The kubelet relies on a container runtime. You should deploy a container runtime such as containerd or CRI-O and ensure that it is running within the user namespace before the kubelet starts.
Running CRI plugin of containerd in a user namespace is supported since containerd 1.4.
Running containerd within a user namespace requires the following configurations
in /etc/containerd/containerd-config.toml.
version = 2
[plugins."io.containerd.grpc.v1.cri"]
# Disable AppArmor
disable_apparmor = true
# Ignore an error during setting oom_score_adj
restrict_oom_score_adj = true
# Disable hugetlb cgroup v2 controller (because systemd does not support delegating hugetlb controller)
disable_hugetlb_controller = true
[plugins."io.containerd.grpc.v1.cri".containerd]
# Using non-fuse overlayfs is also possible for kernel >= 5.11, but requires SELinux to be disabled
snapshotter = "fuse-overlayfs"
[plugins."io.containerd.grpc.v1.cri".containerd.runtimes.runc.options]
# We use cgroupfs that is delegated by systemd, so we do not use SystemdCgroup driver
# (unless you run another systemd in the namespace)
SystemdCgroup = false
Running CRI-O in a user namespace is supported since CRI-O 1.22.
CRI-O requires an environment variable _CRIO_ROOTLESS=1 to be set.
The following configurations (in /etc/crio/crio.conf) are also recommended:
[crio]
storage_driver = "overlay"
# Using non-fuse overlayfs is also possible for kernel >= 5.11, but requires SELinux to be disabled
storage_option = ["overlay.mount_program=/usr/local/bin/fuse-overlayfs"]
[crio.runtime]
# We use cgroupfs that is delegated by systemd, so we do not use "systemd" driver
# (unless you run another systemd in the namespace)
cgroup_manager = "cgroupfs"
Running kubelet in a user namespace requires the following configuration:
apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration
featureGates:
KubeletInUserNamespace: true
# We use cgroupfs that is delegated by systemd, so we do not use "systemd" driver
# (unless you run another systemd in the namespace)
cgroupDriver: "cgroupfs"
When the KubeletInUserNamespace feature gate is enabled, the kubelet ignores errors
that may happen during setting the following sysctl values on the node.
vm.overcommit_memoryvm.panic_on_oomkernel.panickernel.panic_on_oopskernel.keys.root_maxkeyskernel.keys.root_maxbytes.Within a user namespace, the kubelet also ignores any error raised from trying to open /dev/kmsg.
This feature gate also allows kube-proxy to ignore an error during setting RLIMIT_NOFILE.
The KubeletInUserNamespace feature gate was introduced in Kubernetes v1.22 with "alpha" status.
Running kubelet in a user namespace without using this feature gate is also possible by mounting a specially crafted proc filesystem, but not officially supported.
Running kube-proxy in a user namespace requires the following configuration:
apiVersion: kubeproxy.config.k8s.io/v1alpha1
kind: KubeProxyConfiguration
mode: "iptables" # or "userspace"
conntrack:
# Skip setting sysctl value "net.netfilter.nf_conntrack_max"
maxPerCore: 0
# Skip setting "net.netfilter.nf_conntrack_tcp_timeout_established"
tcpEstablishedTimeout: 0s
# Skip setting "net.netfilter.nf_conntrack_tcp_timeout_close"
tcpCloseWaitTimeout: 0s
Most of "non-local" volume drivers such as nfs and iscsi do not work.
Local volumes like local, hostPath, emptyDir, configMap, secret, and downwardAPI are known to work.
Some CNI plugins may not work. Flannel (VXLAN) is known to work.
For more on this, see the Caveats and Future work page on the rootlesscontaine.rs website.
This page shows how to safely drain a node, optionally respecting the PodDisruptionBudget you have defined.
Your Kubernetes server must be at or later than version 1.5.
To check the version, enter kubectl version.
This task also assumes that you have met the following prerequisites:
To ensure that your workloads remain available during maintenance, you can configure a PodDisruptionBudget.
If availability is important for any applications that run or could run on the node(s) that you are draining, configure a PodDisruptionBudgets first and the continue following this guide.
kubectl drain to remove a node from serviceYou can use kubectl drain to safely evict all of your pods from a
node before you perform maintenance on the node (e.g. kernel upgrade,
hardware maintenance, etc.). Safe evictions allow the pod's containers
to gracefully terminate
and will respect the PodDisruptionBudgets you have specified.
kubectl drain ignores certain system pods on the node
that cannot be killed; see
the kubectl drain
documentation for more details.
When kubectl drain returns successfully, that indicates that all of
the pods (except the ones excluded as described in the previous paragraph)
have been safely evicted (respecting the desired graceful termination period,
and respecting the PodDisruptionBudget you have defined). It is then safe to
bring down the node by powering down its physical machine or, if running on a
cloud platform, deleting its virtual machine.
First, identify the name of the node you wish to drain. You can list all of the nodes in your cluster with
kubectl get nodes
Next, tell Kubernetes to drain the node:
kubectl drain <node name>
Once it returns (without giving an error), you can power down the node (or equivalently, if on a cloud platform, delete the virtual machine backing the node). If you leave the node in the cluster during the maintenance operation, you need to run
kubectl uncordon <node name>
afterwards to tell Kubernetes that it can resume scheduling new pods onto the node.
The kubectl drain command should only be issued to a single node at a
time. However, you can run multiple kubectl drain commands for
different nodes in parallel, in different terminals or in the
background. Multiple drain commands running concurrently will still
respect the PodDisruptionBudget you specify.
For example, if you have a StatefulSet with three replicas and have
set a PodDisruptionBudget for that set specifying minAvailable: 2,
kubectl drain only evicts a pod from the StatefulSet if all three
replicas pods are ready; if then you issue multiple drain commands in
parallel, Kubernetes respects the PodDisruptionBudget and ensure
that only 1 (calculated as replicas - minAvailable) Pod is unavailable
at any given time. Any drains that would cause the number of ready
replicas to fall below the specified budget are blocked.
If you prefer not to use kubectl drain (such as to avoid calling to an external command, or to get finer control over the pod eviction process), you can also programmatically cause evictions using the eviction API.
You should first be familiar with using Kubernetes language clients to access the API.
The eviction subresource of a Pod can be thought of as a kind of policy-controlled DELETE operation on the Pod itself. To attempt an eviction (more precisely: to attempt to create an Eviction), you POST an attempted operation. Here's an example:
policy/v1 Eviction is available in v1.22+. Use policy/v1beta1 with prior releases.
{
"apiVersion": "policy/v1",
"kind": "Eviction",
"metadata": {
"name": "quux",
"namespace": "default"
}
}
policy/v1
{
"apiVersion": "policy/v1beta1",
"kind": "Eviction",
"metadata": {
"name": "quux",
"namespace": "default"
}
}
You can attempt an eviction using curl:
curl -v -H 'Content-type: application/json' https://your-cluster-api-endpoint.example/api/v1/namespaces/default/pods/quux/eviction -d @eviction.json
The API can respond in one of three ways:
DELETE request to the Pod's URL and received back 200 OK.429 Too Many Requests. This is
typically used for generic rate limiting of any requests, but here we mean
that this request isn't allowed right now but it may be allowed later.500 Internal Server Error response.For a given eviction request, there are two cases:
200 OK.In some cases, an application may reach a broken state, one where unless you intervene the eviction API will never return anything other than 429 or 500.
For example: this can happen if ReplicaSet is creating Pods for your application but
the replacement Pods do not become Ready. You can also see similar symptoms if the
last Pod evicted has a very long termination grace period.
In this case, there are two potential solutions:
DELETE the Pod from your cluster's control plane, instead
of using the eviction API.Kubernetes does not specify what the behavior should be in this case; it is up to the application owners and cluster owners to establish an agreement on behavior in these cases.
This document covers topics related to protecting a cluster from accidental or malicious access and provides recommendations on overall security.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
As Kubernetes is entirely API-driven, controlling and limiting who can access the cluster and what actions they are allowed to perform is the first line of defense.
Kubernetes expects that all API communication in the cluster is encrypted by default with TLS, and the majority of installation methods will allow the necessary certificates to be created and distributed to the cluster components. Note that some components and installation methods may enable local ports over HTTP and administrators should familiarize themselves with the settings of each component to identify potentially unsecured traffic.
Choose an authentication mechanism for the API servers to use that matches the common access patterns when you install a cluster. For instance, small, single-user clusters may wish to use a simple certificate or static Bearer token approach. Larger clusters may wish to integrate an existing OIDC or LDAP server that allow users to be subdivided into groups.
All API clients must be authenticated, even those that are part of the infrastructure like nodes, proxies, the scheduler, and volume plugins. These clients are typically service accounts or use x509 client certificates, and they are created automatically at cluster startup or are setup as part of the cluster installation.
Consult the authentication reference document for more information.
Once authenticated, every API call is also expected to pass an authorization check. Kubernetes ships an integrated Role-Based Access Control (RBAC) component that matches an incoming user or group to a set of permissions bundled into roles. These permissions combine verbs (get, create, delete) with resources (pods, services, nodes) and can be namespace-scoped or cluster-scoped. A set of out-of-the-box roles are provided that offer reasonable default separation of responsibility depending on what actions a client might want to perform. It is recommended that you use the Node and RBAC authorizers together, in combination with the NodeRestriction admission plugin.
As with authentication, simple and broad roles may be appropriate for smaller clusters, but as more users interact with the cluster, it may become necessary to separate teams into separate namespaces with more limited roles.
With authorization, it is important to understand how updates on one object may cause actions in other places. For instance, a user may not be able to create pods directly, but allowing them to create a deployment, which creates pods on their behalf, will let them create those pods indirectly. Likewise, deleting a node from the API will result in the pods scheduled to that node being terminated and recreated on other nodes. The out-of-the box roles represent a balance between flexibility and common use cases, but more limited roles should be carefully reviewed to prevent accidental escalation. You can make roles specific to your use case if the out-of-box ones don't meet your needs.
Consult the authorization reference section for more information.
Kubelets expose HTTPS endpoints which grant powerful control over the node and containers. By default Kubelets allow unauthenticated access to this API.
Production clusters should enable Kubelet authentication and authorization.
Consult the Kubelet authentication/authorization reference for more information.
Authorization in Kubernetes is intentionally high level, focused on coarse actions on resources. More powerful controls exist as policies to limit by use case how those objects act on the cluster, themselves, and other resources.
Resource quota limits the number or capacity of resources granted to a namespace. This is most often used to limit the amount of CPU, memory, or persistent disk a namespace can allocate, but can also control how many pods, services, or volumes exist in each namespace.
Limit ranges restrict the maximum or minimum size of some of the resources above, to prevent users from requesting unreasonably high or low values for commonly reserved resources like memory, or to provide default limits when none are specified.
A pod definition contains a security context
that allows it to request access to run as a specific Linux user on a node (like root),
access to run privileged or access the host network, and other controls that would otherwise
allow it to run unfettered on a hosting node. Pod security policies
can limit which users or service accounts can provide dangerous security context settings. For example, pod security policies can limit volume mounts, especially hostPath, which are aspects of a pod that should be controlled.
Generally, most application workloads need limited access to host resources so they can successfully run as a root process (uid 0) without access to host information. However, considering the privileges associated with the root user, you should write application containers to run as a non-root user. Similarly, administrators who wish to prevent client applications from escaping their containers should use a restrictive pod security policy.
The Linux kernel automatically loads kernel modules from disk if needed in certain circumstances, such as when a piece of hardware is attached or a filesystem is mounted. Of particular relevance to Kubernetes, even unprivileged processes can cause certain network-protocol-related kernel modules to be loaded, just by creating a socket of the appropriate type. This may allow an attacker to exploit a security hole in a kernel module that the administrator assumed was not in use.
To prevent specific modules from being automatically loaded, you can uninstall them from
the node, or add rules to block them. On most Linux distributions, you can do that by
creating a file such as /etc/modprobe.d/kubernetes-blacklist.conf with contents like:
# DCCP is unlikely to be needed, has had multiple serious
# vulnerabilities, and is not well-maintained.
blacklist dccp
# SCTP is not used in most Kubernetes clusters, and has also had
# vulnerabilities in the past.
blacklist sctp
To block module loading more generically, you can use a Linux Security Module (such as
SELinux) to completely deny the module_request permission to containers, preventing the
kernel from loading modules for containers under any circumstances. (Pods would still be
able to use modules that had been loaded manually, or modules that were loaded by the
kernel on behalf of some more-privileged process.)
The network policies for a namespace allows application authors to restrict which pods in other namespaces may access pods and ports within their namespaces. Many of the supported Kubernetes networking providers now respect network policy.
Quota and limit ranges can also be used to control whether users may request node ports or load-balanced services, which on many clusters can control whether those users applications are visible outside of the cluster.
Additional protections may be available that control network rules on a per-plugin or per- environment basis, such as per-node firewalls, physically separating cluster nodes to prevent cross talk, or advanced networking policy.
Cloud platforms (AWS, Azure, GCE, etc.) often expose metadata services locally to instances. By default these APIs are accessible by pods running on an instance and can contain cloud credentials for that node, or provisioning data such as kubelet credentials. These credentials can be used to escalate within the cluster or to other cloud services under the same account.
When running Kubernetes on a cloud platform, limit permissions given to instance credentials, use network policies to restrict pod access to the metadata API, and avoid using provisioning data to deliver secrets.
By default, there are no restrictions on which nodes may run a pod. Kubernetes offers a rich set of policies for controlling placement of pods onto nodes and the taint-based pod placement and eviction that are available to end users. For many clusters use of these policies to separate workloads can be a convention that authors adopt or enforce via tooling.
As an administrator, a beta admission plugin PodNodeSelector can be used to force pods
within a namespace to default or require a specific node selector, and if end users cannot
alter namespaces, this can strongly limit the placement of all of the pods in a specific workload.
This section describes some common patterns for protecting clusters from compromise.
Write access to the etcd backend for the API is equivalent to gaining root on the entire cluster, and read access can be used to escalate fairly quickly. Administrators should always use strong credentials from the API servers to their etcd server, such as mutual auth via TLS client certificates, and it is often recommended to isolate the etcd servers behind a firewall that only the API servers may access.
The audit logger is a beta feature that records actions taken by the API for later analysis in the event of a compromise. It is recommended to enable audit logging and archive the audit file on a secure server.
Alpha and beta Kubernetes features are in active development and may have limitations or bugs that result in security vulnerabilities. Always assess the value an alpha or beta feature may provide against the possible risk to your security posture. When in doubt, disable features you do not use.
The shorter the lifetime of a secret or credential the harder it is for an attacker to make use of that credential. Set short lifetimes on certificates and automate their rotation. Use an authentication provider that can control how long issued tokens are available and use short lifetimes where possible. If you use service-account tokens in external integrations, plan to rotate those tokens frequently. For example, once the bootstrap phase is complete, a bootstrap token used for setting up nodes should be revoked or its authorization removed.
Many third party integrations to Kubernetes may alter the security profile of your cluster. When enabling an integration, always review the permissions that an extension requests before granting it access. For example, many security integrations may request access to view all secrets on your cluster which is effectively making that component a cluster admin. When in doubt, restrict the integration to functioning in a single namespace if possible.
Components that create pods may also be unexpectedly powerful if they can do so inside namespaces
like the kube-system namespace, because those pods can gain access to service account secrets
or run with elevated permissions if those service accounts are granted access to permissive
pod security policies.
In general, the etcd database will contain any information accessible via the Kubernetes API and may grant an attacker significant visibility into the state of your cluster. Always encrypt your backups using a well reviewed backup and encryption solution, and consider using full disk encryption where possible.
Kubernetes supports encryption at rest, a feature
introduced in 1.7, and beta since 1.13. This will encrypt Secret resources in etcd, preventing
parties that gain access to your etcd backups from viewing the content of those secrets. While
this feature is currently beta, it offers an additional level of defense when backups
are not encrypted or an attacker gains read access to etcd.
Join the kubernetes-announce group for emails about security announcements. See the security reporting page for more on how to report vulnerabilities.
A subset of the Kubelet's configuration parameters may be set via an on-disk config file, as a substitute for command-line flags.
Providing parameters via a config file is the recommended approach because it simplifies node deployment and configuration management.
The subset of the Kubelet's configuration that can be configured via a file
is defined by the
KubeletConfiguration
struct.
The configuration file must be a JSON or YAML representation of the parameters in this struct. Make sure the Kubelet has read permissions on the file.
Here is an example of what this file might look like:
apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration
address: "192.168.0.8",
port: 20250,
serializeImagePulls: false,
evictionHard:
memory.available: "200Mi"
In the example, the Kubelet is configured to serve on IP address 192.168.0.8 and port 20250, pull images in parallel, and evict Pods when available memory drops below 200Mi. All other Kubelet configuration values are left at their built-in defaults, unless overridden by flags. Command line flags which target the same value as a config file will override that value.
kubeadmin init.
See configuring kubelet using kubeadm for details.
Start the Kubelet with the --config flag set to the path of the Kubelet's config file.
The Kubelet will then load its config from this file.
Note that command line flags which target the same value as a config file will override that value. This helps ensure backwards compatibility with the command-line API.
Note that relative file paths in the Kubelet config file are resolved relative to the location of the Kubelet config file, whereas relative paths in command line flags are resolved relative to the Kubelet's current working directory.
Note that some default values differ between command-line flags and the Kubelet config file.
If --config is provided and the values are not specified via the command line, the
defaults for the KubeletConfiguration version apply.
In the above example, this version is kubelet.config.k8s.io/v1beta1.
KubeletConfiguration
reference.Kubernetes v1.5 [alpha]
You can replicate Kubernetes control plane nodes in kube-up or kube-down scripts for Google Compute Engine. However this scripts are not suitable for any sort of production use, it's widely used in the project's CI.
This document describes how to use kube-up/down scripts to manage a highly available (HA) control plane and how HA control planes are implemented for use with GCE.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
To create a new HA-compatible cluster, you must set the following flags in your kube-up script:
MULTIZONE=true - to prevent removal of control plane kubelets from zones different than server's default zone.
Required if you want to run control plane nodes in different zones, which is recommended.
ENABLE_ETCD_QUORUM_READ=true - to ensure that reads from all API servers will return most up-to-date data.
If true, reads will be directed to leader etcd replica.
Setting this value to true is optional: reads will be more reliable but will also be slower.
Optionally, you can specify a GCE zone where the first control plane node is to be created. Set the following flag:
KUBE_GCE_ZONE=zone - zone where the first control plane node will run.The following sample command sets up a HA-compatible cluster in the GCE zone europe-west1-b:
MULTIZONE=true KUBE_GCE_ZONE=europe-west1-b ENABLE_ETCD_QUORUM_READS=true ./cluster/kube-up.sh
Note that the commands above create a cluster with one control plane node; however, you can add new control plane nodes to the cluster with subsequent commands.
After you have created an HA-compatible cluster, you can add control plane nodes to it.
You add control plane nodes by using a kube-up script with the following flags:
KUBE_REPLICATE_EXISTING_MASTER=true - to create a replica of an existing control plane
node.
KUBE_GCE_ZONE=zone - zone where the control plane node will run.
Must be in the same region as other control plane nodes' zones.
You don't need to set the MULTIZONE or ENABLE_ETCD_QUORUM_READS flags,
as those are inherited from when you started your HA-compatible cluster.
The following sample command replicates the control plane node on an existing HA-compatible cluster:
KUBE_GCE_ZONE=europe-west1-c KUBE_REPLICATE_EXISTING_MASTER=true ./cluster/kube-up.sh
You can remove a control plane node from an HA cluster by using a kube-down script with the following flags:
KUBE_DELETE_NODES=false - to restrain deletion of kubelets.
KUBE_GCE_ZONE=zone - the zone from where the control plane node will be removed.
KUBE_REPLICA_NAME=replica_name - (optional) the name of control plane node to
remove. If empty: any replica from the given zone will be removed.
The following sample command removes a control plane node from an existing HA cluster:
KUBE_DELETE_NODES=false KUBE_GCE_ZONE=europe-west1-c ./cluster/kube-down.sh
If one of the control plane nodes in your HA cluster fails, the best practice is to remove the node from your cluster and add a new control plane node in the same zone. The following sample commands demonstrate this process:
KUBE_DELETE_NODES=false KUBE_GCE_ZONE=replica_zone KUBE_REPLICA_NAME=replica_name ./cluster/kube-down.sh
KUBE_GCE_ZONE=replica-zone KUBE_REPLICATE_EXISTING_MASTER=true ./cluster/kube-up.sh
Try to place control plane nodes in different zones. During a zone failure, all control plane nodes placed inside the zone will fail. To survive zone failure, also place nodes in multiple zones (see multiple-zones for details).
Do not use a cluster with two control plane nodes. Consensus on a two-node control plane requires both nodes running when changing persistent state. As a result, both nodes are needed and a failure of any node turns the cluster into majority failure state. A two-node control plane is thus inferior, in terms of HA, to a cluster with one control plane node.
When you add a control plane node, cluster state (etcd) is copied to a new instance. If the cluster is large, it may take a long time to duplicate its state. This operation may be sped up by migrating the etcd data directory, as described in the etcd administration guide (we are considering adding support for etcd data dir migration in the future).
The figure above illustrates three control plane nodes and their components in a highly available cluster. The control plane node’s components employ the following methods:
etcd: instances are clustered together using consensus.
Controllers, scheduler and cluster auto-scaler: only one instance of each will be active in a cluster using a lease mechanism.
Add-on manager: each works independently to keep add-ons in sync.
In addition, a load balancer operating in front of the API servers routes external and internal traffic to the control plane nodes. Each of the control plane nodes will run the following components in the following mode:
etcd instance: all instances will be clustered together using consensus;
API server: each server will talk to local etcd - all API servers in the cluster will be available;
controllers, scheduler, and cluster auto-scaler: will use lease mechanism - only one instance of each of them will be active in the cluster;
add-on manager: each manager will work independently trying to keep add-ons in sync.
In addition, there will be a load balancer in front of API servers that will route external and internal traffic to them.
When starting the second control plane node, a load balancer containing the two replicas will be created and the IP address of the first replica will be promoted to IP address of load balancer. Similarly, after removal of the penultimate control plane node, the load balancer will be removed and its IP address will be assigned to the last remaining replica. Please note that creation and removal of load balancer are complex operations and it may take some time (~20 minutes) for them to propagate.
Instead of trying to keep an up-to-date list of Kubernetes apiserver in the Kubernetes service, the system directs all traffic to the external IP:
in case of a single node control plane, the IP points to the control plane node,
in case of an HA control plane, the IP points to the load balancer in-front of the control plane nodes.
Similarly, the external IP will be used by kubelets to communicate with the control plane.
Kubernetes generates TLS certificates for the external public IP and local IP for each control plane node. There are no certificates for the ephemeral public IP for control plane nodes; to access a control plane node via its ephemeral public IP, you must skip TLS verification.
To allow etcd clustering, ports needed to communicate between etcd instances will be opened (for inside cluster communication). To make such deployment secure, communication between etcd instances is authorized using SSL.
Kubernetes v1.20 [alpha]
The API Server Identity feature is controlled by a
feature gate
and is not enabled by default. You can activate API Server Identity by enabling
the feature gate named APIServerIdentity when you start the
API Server:
kube-apiserver \
--feature-gates=APIServerIdentity=true \
# …and other flags as usual
During bootstrap, each kube-apiserver assigns a unique ID to itself. The ID is
in the format of kube-apiserver-{UUID}. Each kube-apiserver creates a
Lease
in the kube-system namespaces.
The Lease name is the unique ID for the kube-apiserver. The Lease contains a
label k8s.io/component=kube-apiserver. Each kube-apiserver refreshes its
Lease every IdentityLeaseRenewIntervalSeconds (defaults to 10s). Each
kube-apiserver also checks all the kube-apiserver identity Leases every
IdentityLeaseDurationSeconds (defaults to 3600s), and deletes Leases that
hasn't got refreshed for more than IdentityLeaseDurationSeconds.
IdentityLeaseRenewIntervalSeconds and IdentityLeaseDurationSeconds can be
configured by kube-apiserver flags identity-lease-renew-interval-seconds
and identity-lease-duration-seconds.
Enabling this feature is a prerequisite for using features that involve HA API
server coordination (for example, the StorageVersionAPI feature gate).
This page shows how to view, work in, and delete namespaces. The page also shows how to use Kubernetes namespaces to subdivide your cluster.
kubectl get namespaces
NAME STATUS AGE
default Active 11d
kube-system Active 11d
kube-public Active 11d
Kubernetes starts with three initial namespaces:
default The default namespace for objects with no other namespacekube-system The namespace for objects created by the Kubernetes systemkube-public This namespace is created automatically and is readable by all users (including those not authenticated). This namespace is mostly reserved for cluster usage, in case that some resources should be visible and readable publicly throughout the whole cluster. The public aspect of this namespace is only a convention, not a requirement.You can also get the summary of a specific namespace using:
kubectl get namespaces <name>
Or you can get detailed information with:
kubectl describe namespaces <name>
Name: default
Labels: <none>
Annotations: <none>
Status: Active
No resource quota.
Resource Limits
Type Resource Min Max Default
---- -------- --- --- ---
Container cpu - - 100m
Note that these details show both resource quota (if present) as well as resource limit ranges.
Resource quota tracks aggregate usage of resources in the Namespace and allows cluster operators to define Hard resource usage limits that a Namespace may consume.
A limit range defines min/max constraints on the amount of resources a single entity can consume in a Namespace.
See Admission control: Limit Range
A namespace can be in one of two phases:
Active the namespace is in useTerminating the namespace is being deleted, and can not be used for new objectsFor more details, see Namespace in the API reference.
kube-, since it is reserved for Kubernetes system namespaces.
Create a new YAML file called my-namespace.yaml with the contents:
apiVersion: v1
kind: Namespace
metadata:
name: <insert-namespace-name-here>
Then run:
kubectl create -f ./my-namespace.yaml
Alternatively, you can create namespace using below command:
kubectl create namespace <insert-namespace-name-here>
The name of your namespace must be a valid DNS label.
There's an optional field finalizers, which allows observables to purge resources whenever the namespace is deleted. Keep in mind that if you specify a nonexistent finalizer, the namespace will be created but will get stuck in the Terminating state if the user tries to delete it.
More information on finalizers can be found in the namespace design doc.
Delete a namespace with
kubectl delete namespaces <insert-some-namespace-name>
This delete is asynchronous, so for a time you will see the namespace in the Terminating state.
Understand the default namespace
By default, a Kubernetes cluster will instantiate a default namespace when provisioning the cluster to hold the default set of Pods, Services, and Deployments used by the cluster.
Assuming you have a fresh cluster, you can introspect the available namespaces by doing the following:
kubectl get namespaces
NAME STATUS AGE
default Active 13m
Create new namespaces
For this exercise, we will create two additional Kubernetes namespaces to hold our content.
In a scenario where an organization is using a shared Kubernetes cluster for development and production use cases:
The development team would like to maintain a space in the cluster where they can get a view on the list of Pods, Services, and Deployments they use to build and run their application. In this space, Kubernetes resources come and go, and the restrictions on who can or cannot modify resources are relaxed to enable agile development.
The operations team would like to maintain a space in the cluster where they can enforce strict procedures on who can or cannot manipulate the set of Pods, Services, and Deployments that run the production site.
One pattern this organization could follow is to partition the Kubernetes cluster into two namespaces: development and production.
Let's create two new namespaces to hold our work.
Create the development namespace using kubectl:
kubectl create -f https://k8s.io/examples/admin/namespace-dev.json
And then let's create the production namespace using kubectl:
kubectl create -f https://k8s.io/examples/admin/namespace-prod.json
To be sure things are right, list all of the namespaces in our cluster.
kubectl get namespaces --show-labels
NAME STATUS AGE LABELS
default Active 32m <none>
development Active 29s name=development
production Active 23s name=production
Create pods in each namespace
A Kubernetes namespace provides the scope for Pods, Services, and Deployments in the cluster.
Users interacting with one namespace do not see the content in another namespace.
To demonstrate this, let's spin up a simple Deployment and Pods in the development namespace.
kubectl create deployment snowflake --image=k8s.gcr.io/serve_hostname -n=development --replicas=2
We have created a deployment whose replica size is 2 that is running the pod called snowflake with a basic container that serves the hostname.
kubectl get deployment -n=development
NAME READY UP-TO-DATE AVAILABLE AGE
snowflake 2/2 2 2 2m
kubectl get pods -l app=snowflake -n=development
NAME READY STATUS RESTARTS AGE
snowflake-3968820950-9dgr8 1/1 Running 0 2m
snowflake-3968820950-vgc4n 1/1 Running 0 2m
And this is great, developers are able to do what they want, and they do not have to worry about affecting content in the production namespace.
Let's switch to the production namespace and show how resources in one namespace are hidden from the other.
The production namespace should be empty, and the following commands should return nothing.
kubectl get deployment -n=production
kubectl get pods -n=production
Production likes to run cattle, so let's create some cattle pods.
kubectl create deployment cattle --image=k8s.gcr.io/serve_hostname -n=production
kubectl scale deployment cattle --replicas=5 -n=production
kubectl get deployment -n=production
NAME READY UP-TO-DATE AVAILABLE AGE
cattle 5/5 5 5 10s
kubectl get pods -l app=cattle -n=production
NAME READY STATUS RESTARTS AGE
cattle-2263376956-41xy6 1/1 Running 0 34s
cattle-2263376956-kw466 1/1 Running 0 34s
cattle-2263376956-n4v97 1/1 Running 0 34s
cattle-2263376956-p5p3i 1/1 Running 0 34s
cattle-2263376956-sxpth 1/1 Running 0 34s
At this point, it should be clear that the resources users create in one namespace are hidden from the other namespace.
As the policy support in Kubernetes evolves, we will extend this scenario to show how you can provide different authorization rules for each namespace.
A single cluster should be able to satisfy the needs of multiple users or groups of users (henceforth a 'user community').
Kubernetes namespaces help different projects, teams, or customers to share a Kubernetes cluster.
It does this by providing the following:
Use of multiple namespaces is optional.
Each user community wants to be able to work in isolation from other communities.
Each user community has its own:
A cluster operator may create a Namespace for each unique user community.
The Namespace provides a unique scope for:
Use cases include:
When you create a Service, it creates a corresponding DNS entry.
This entry is of the form <service-name>.<namespace-name>.svc.cluster.local, which means
that if a container uses <service-name> it will resolve to the service which
is local to a namespace. This is useful for using the same configuration across
multiple namespaces such as Development, Staging and Production. If you want to reach
across namespaces, you need to use the fully qualified domain name (FQDN).
This page provides an overview of the steps you should follow to upgrade a Kubernetes cluster.
The way that you upgrade a cluster depends on how you initially deployed it and on any subsequent changes.
At a high level, the steps you perform are:
You must have an existing cluster. This page is about upgrading from Kubernetes 1.21 to Kubernetes 1.22. If your cluster is not currently running Kubernetes 1.21 then please check the documentation for the version of Kubernetes that you plan to upgrade to.
If your cluster was deployed using the kubeadm tool, refer to
Upgrading kubeadm clusters
for detailed information on how to upgrade the cluster.
Once you have upgraded the cluster, remember to
install the latest version of kubectl.
You should manually update the control plane following this sequence:
At this point you should
install the latest version of kubectl.
For each node in your cluster, drain that node and then either replace it with a new node that uses the 1.22 kubelet, or upgrade the kubelet on that node and bring the node back into service.
Refer to the documentation for your cluster deployment tool to learn the recommended set up steps for maintenance.
The objects that are serialized into etcd for a cluster's internal representation of the Kubernetes resources active in the cluster are written using a particular version of the API.
When the supported API changes, these objects may need to be rewritten in the newer API. Failure to do this will eventually result in resources that are no longer decodable or usable by the Kubernetes API server.
For each affected object, fetch it using the latest supported API and then write it back also using the latest supported API.
Upgrading to a new Kubernetes version can provide new APIs.
You can use kubectl convert command to convert manifests between different API versions.
For example:
kubectl convert -f pod.yaml --output-version v1
The kubectl tool replaces the contents of pod.yaml with a manifest that sets kind to
Pod (unchanged), but with a revised apiVersion.
This page shows you how to specify the type of cascading deletion to use in your cluster during garbage collection.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
You also need to create a sample Deployment to experiment with the different types of cascading deletion. You will need to recreate the Deployment for each type.
Check that the ownerReferences field is present on your pods:
kubectl get pods -l app=nginx --output=yaml
The output has an ownerReferences field similar to this:
apiVersion: v1
...
ownerReferences:
- apiVersion: apps/v1
blockOwnerDeletion: true
controller: true
kind: ReplicaSet
name: nginx-deployment-6b474476c4
uid: 4fdcd81c-bd5d-41f7-97af-3a3b759af9a7
...
By default, Kubernetes uses background cascading deletion
to delete dependents of an object. You can switch to foreground cascading deletion
using either kubectl or the Kubernetes API, depending on the Kubernetes
version your cluster runs.
To check the version, enter kubectl version.
You can delete objects using foreground cascading deletion using kubectl or the
Kubernetes API.
Using kubectl
Run the following command:
kubectl delete deployment nginx-deployment --cascade=foreground
Using the Kubernetes API
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Foreground"}' \
-H "Content-Type: application/json"
The output contains a foregroundDeletion finalizer
like this:
"kind": "Deployment",
"apiVersion": "apps/v1",
"metadata": {
"name": "nginx-deployment",
"namespace": "default",
"uid": "d1ce1b02-cae8-4288-8a53-30e84d8fa505",
"resourceVersion": "1363097",
"creationTimestamp": "2021-07-08T20:24:37Z",
"deletionTimestamp": "2021-07-08T20:27:39Z",
"finalizers": [
"foregroundDeletion"
]
...
You can delete objects using foreground cascading deletion by calling the Kubernetes API.
For details, read the documentation for your Kubernetes version.
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Foreground"}' \
-H "Content-Type: application/json"
The output contains a foregroundDeletion finalizer
like this:
"kind": "Deployment",
"apiVersion": "apps/v1",
"metadata": {
"name": "nginx-deployment",
"namespace": "default",
"uid": "d1ce1b02-cae8-4288-8a53-30e84d8fa505",
"resourceVersion": "1363097",
"creationTimestamp": "2021-07-08T20:24:37Z",
"deletionTimestamp": "2021-07-08T20:27:39Z",
"finalizers": [
"foregroundDeletion"
]
...
kubectl or the Kubernetes API to delete the Deployment,
depending on the Kubernetes version your cluster runs.
To check the version, enter kubectl version.
You can delete objects using background cascading deletion using kubectl
or the Kubernetes API.
Kubernetes uses background cascading deletion by default, and does so
even if you run the following commands without the --cascade flag or the
propagationPolicy argument.
Using kubectl
Run the following command:
kubectl delete deployment nginx-deployment --cascade=background
Using the Kubernetes API
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Background"}' \
-H "Content-Type: application/json"
The output is similar to this:
"kind": "Status",
"apiVersion": "v1",
...
"status": "Success",
"details": {
"name": "nginx-deployment",
"group": "apps",
"kind": "deployments",
"uid": "cc9eefb9-2d49-4445-b1c1-d261c9396456"
}
Kubernetes uses background cascading deletion by default, and does so
even if you run the following commands without the --cascade flag or the
propagationPolicy: Background argument.
For details, read the documentation for your Kubernetes version.
Using kubectl
Run the following command:
kubectl delete deployment nginx-deployment --cascade=true
Using the Kubernetes API
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Background"}' \
-H "Content-Type: application/json"
The output is similar to this:
"kind": "Status",
"apiVersion": "v1",
...
"status": "Success",
"details": {
"name": "nginx-deployment",
"group": "apps",
"kind": "deployments",
"uid": "cc9eefb9-2d49-4445-b1c1-d261c9396456"
}
By default, when you tell Kubernetes to delete an object, the
controller also deletes
dependent objects. You can make Kubernetes orphan these dependents using
kubectl or the Kubernetes API, depending on the Kubernetes version your
cluster runs.
To check the version, enter kubectl version.
Using kubectl
Run the following command:
kubectl delete deployment nginx-deployment --cascade=orphan
Using the Kubernetes API
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Orphan"}' \
-H "Content-Type: application/json"
The output contains orphan in the finalizers field, similar to this:
"kind": "Deployment",
"apiVersion": "apps/v1",
"namespace": "default",
"uid": "6f577034-42a0-479d-be21-78018c466f1f",
"creationTimestamp": "2021-07-09T16:46:37Z",
"deletionTimestamp": "2021-07-09T16:47:08Z",
"deletionGracePeriodSeconds": 0,
"finalizers": [
"orphan"
],
...
For details, read the documentation for your Kubernetes version.
Using kubectl
Run the following command:
kubectl delete deployment nginx-deployment --cascade=orphan
Using the Kubernetes API
Start a local proxy session:
kubectl proxy --port=8080
Use curl to trigger deletion:
curl -X DELETE localhost:8080/apis/apps/v1/namespaces/default/deployments/nginx-deployment \
-d '{"kind":"DeleteOptions","apiVersion":"v1","propagationPolicy":"Orphan"}' \
-H "Content-Type: application/json"
The output contains orphan in the finalizers field, similar to this:
"kind": "Deployment",
"apiVersion": "apps/v1",
"namespace": "default",
"uid": "6f577034-42a0-479d-be21-78018c466f1f",
"creationTimestamp": "2021-07-09T16:46:37Z",
"deletionTimestamp": "2021-07-09T16:47:08Z",
"deletionGracePeriodSeconds": 0,
"finalizers": [
"orphan"
],
...
You can check that the Pods managed by the Deployment are still running:
kubectl get pods -l app=nginx
This page shows how to configure a Key Management Service (KMS) provider and plugin to enable secret data encryption.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Kubernetes version 1.10.0 or later is required
etcd v3 or later is required
Kubernetes v1.12 [beta]
The KMS encryption provider uses an envelope encryption scheme to encrypt data in etcd. The data is encrypted using a data encryption key (DEK); a new DEK is generated for each encryption. The DEKs are encrypted with a key encryption key (KEK) that is stored and managed in a remote KMS. The KMS provider uses gRPC to communicate with a specific KMS plugin. The KMS plugin, which is implemented as a gRPC server and deployed on the same host(s) as the Kubernetes master(s), is responsible for all communication with the remote KMS.
To configure a KMS provider on the API server, include a provider of type kms in the providers array in the encryption configuration file and set the following properties:
name: Display name of the KMS plugin.endpoint: Listen address of the gRPC server (KMS plugin). The endpoint is a UNIX domain socket.cachesize: Number of data encryption keys (DEKs) to be cached in the clear.
When cached, DEKs can be used without another call to the KMS;
whereas DEKs that are not cached require a call to the KMS to unwrap.timeout: How long should kube-apiserver wait for kms-plugin to respond before returning an error (default is 3 seconds).See Understanding the encryption at rest configuration.
To implement a KMS plugin, you can develop a new plugin gRPC server or enable a KMS plugin already provided by your cloud provider. You then integrate the plugin with the remote KMS and deploy it on the Kubernetes master.
Refer to your cloud provider for instructions on enabling the cloud provider-specific KMS plugin.
You can develop a KMS plugin gRPC server using a stub file available for Go. For other languages, you use a proto file to create a stub file that you can use to develop the gRPC server code.
Using Go: Use the functions and data structures in the stub file: service.pb.go to develop the gRPC server code
Using languages other than Go: Use the protoc compiler with the proto file: service.proto to generate a stub file for the specific language
Then use the functions and data structures in the stub file to develop the server code.
Notes:
kms plugin version: v1beta1
In response to procedure call Version, a compatible KMS plugin should return v1beta1 as VersionResponse.version.
message version: v1beta1
All messages from KMS provider have the version field set to current version v1beta1.
protocol: UNIX domain socket (unix)
The gRPC server should listen at UNIX domain socket.
The KMS plugin can communicate with the remote KMS using any protocol supported by the KMS. All configuration data, including authentication credentials the KMS plugin uses to communicate with the remote KMS, are stored and managed by the KMS plugin independently. The KMS plugin can encode the ciphertext with additional metadata that may be required before sending it to the KMS for decryption.
Ensure that the KMS plugin runs on the same host(s) as the Kubernetes master(s).
To encrypt the data:
Create a new encryption configuration file using the appropriate properties for the kms provider:
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- kms:
name: myKmsPlugin
endpoint: unix:///tmp/socketfile.sock
cachesize: 100
timeout: 3s
- identity: {}
Set the --encryption-provider-config flag on the kube-apiserver to point to the location of the configuration file.
Restart your API server.
Data is encrypted when written to etcd. After restarting your kube-apiserver,
any newly created or updated secret should be encrypted when stored. To verify,
you can use the etcdctl command line program to retrieve the contents of your secret.
Create a new secret called secret1 in the default namespace:
kubectl create secret generic secret1 -n default --from-literal=mykey=mydata
Using the etcdctl command line, read that secret out of etcd:
ETCDCTL_API=3 etcdctl get /kubernetes.io/secrets/default/secret1 [...] | hexdump -C
where [...] must be the additional arguments for connecting to the etcd server.
Verify the stored secret is prefixed with k8s:enc:kms:v1:, which indicates that the kms provider has encrypted the resulting data.
Verify that the secret is correctly decrypted when retrieved via the API:
kubectl describe secret secret1 -n default
should match mykey: mydata
Because secrets are encrypted on write, performing an update on a secret encrypts that content.
The following command reads all secrets and then updates them to apply server side encryption. If an error occurs due to a conflicting write, retry the command. For larger clusters, you may wish to subdivide the secrets by namespace or script an update.
kubectl get secrets --all-namespaces -o json | kubectl replace -f -
To switch from a local encryption provider to the kms provider and re-encrypt all of the secrets:
Add the kms provider as the first entry in the configuration file as shown in the following example.
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- kms:
name : myKmsPlugin
endpoint: unix:///tmp/socketfile.sock
cachesize: 100
- aescbc:
keys:
- name: key1
secret: <BASE 64 ENCODED SECRET>
Restart all kube-apiserver processes.
Run the following command to force all secrets to be re-encrypted using the kms provider.
kubectl get secrets --all-namespaces -o json| kubectl replace -f -
To disable encryption at rest:
Place the identity provider as the first entry in the configuration file:
apiVersion: apiserver.config.k8s.io/v1
kind: EncryptionConfiguration
resources:
- resources:
- secrets
providers:
- identity: {}
- kms:
name : myKmsPlugin
endpoint: unix:///tmp/socketfile.sock
cachesize: 100
Restart all kube-apiserver processes.
Run the following command to force all secrets to be decrypted.
kubectl get secrets --all-namespaces -o json | kubectl replace -f -
This page describes the CoreDNS upgrade process and how to install CoreDNS instead of kube-dns.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.9. To check the version, enterkubectl version.
CoreDNS is a flexible, extensible DNS server that can serve as the Kubernetes cluster DNS. Like Kubernetes, the CoreDNS project is hosted by the CNCF.
You can use CoreDNS instead of kube-dns in your cluster by replacing kube-dns in an existing deployment, or by using tools like kubeadm that will deploy and upgrade the cluster for you.
For manual deployment or replacement of kube-dns, see the documentation at the CoreDNS GitHub project.
In Kubernetes version 1.10 and later, you can also move to CoreDNS when you use kubeadm to upgrade
a cluster that is using kube-dns. In this case, kubeadm will generate the CoreDNS configuration
("Corefile") based upon the kube-dns ConfigMap, preserving configurations for
stub domains, and upstream name server.
If you are moving from kube-dns to CoreDNS, make sure to set the CoreDNS feature gate to true
during an upgrade. For example, here is what a v1.11.0 upgrade would look like:
kubeadm upgrade apply v1.11.0 --feature-gates=CoreDNS=true
In Kubernetes version 1.13 and later the CoreDNS feature gate is removed and CoreDNS
is used by default.
In versions prior to 1.11 the Corefile will be overwritten by the one created during upgrade. You should save your existing ConfigMap if you have customized it. You may re-apply your customizations after the new ConfigMap is up and running.
If you are running CoreDNS in Kubernetes version 1.11 and later, during upgrade, your existing Corefile will be retained.
In Kubernetes version 1.21, support for kube-dns is removed from kubeadm.
CoreDNS is available in Kubernetes since v1.9. You can check the version of CoreDNS shipped with Kubernetes and the changes made to CoreDNS here.
CoreDNS can be upgraded manually in case you want to only upgrade CoreDNS or use your own custom image. There is a helpful guideline and walkthrough available to ensure a smooth upgrade.
When resource utilisation is a concern, it may be useful to tune the configuration of CoreDNS. For more details, check out the documentation on scaling CoreDNS.
You can configure CoreDNS to support many more use cases than
kube-dns by modifying the Corefile. For more information, see the
CoreDNS site.
Kubernetes v1.18 [stable]
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
NodeLocal DNSCache improves Cluster DNS performance by running a dns caching agent on cluster nodes as a DaemonSet. In today's architecture, Pods in ClusterFirst DNS mode reach out to a kube-dns serviceIP for DNS queries. This is translated to a kube-dns/CoreDNS endpoint via iptables rules added by kube-proxy. With this new architecture, Pods will reach out to the dns caching agent running on the same node, thereby avoiding iptables DNAT rules and connection tracking. The local caching agent will query kube-dns service for cache misses of cluster hostnames(cluster.local suffix by default).
With the current DNS architecture, it is possible that Pods with the highest DNS QPS have to reach out to a different node, if there is no local kube-dns/CoreDNS instance. Having a local cache will help improve the latency in such scenarios.
Skipping iptables DNAT and connection tracking will help reduce conntrack races and avoid UDP DNS entries filling up conntrack table.
Connections from local caching agent to kube-dns service can be upgraded to TCP. TCP conntrack entries will be removed on connection close in contrast with UDP entries that have to timeout (default nf_conntrack_udp_timeout is 30 seconds)
Upgrading DNS queries from UDP to TCP would reduce tail latency attributed to dropped UDP packets and DNS timeouts usually up to 30s (3 retries + 10s timeout). Since the nodelocal cache listens for UDP DNS queries, applications don't need to be changed.
Metrics & visibility into dns requests at a node level.
Negative caching can be re-enabled, thereby reducing number of queries to kube-dns service.
This is the path followed by DNS Queries after NodeLocal DNSCache is enabled:
This image shows how NodeLocal DNSCache handles DNS queries.
This feature can be enabled using the following steps:
Prepare a manifest similar to the sample nodelocaldns.yaml and save it as nodelocaldns.yaml.
If using IPv6, the CoreDNS configuration file need to enclose all the IPv6 addresses into square brackets if used in IP:Port format.
If you are using the sample manifest from the previous point, this will require to modify the configuration line L70 like this health [__PILLAR__LOCAL__DNS__]:8080
Substitute the variables in the manifest with the right values:
kubedns=kubectl get svc kube-dns -n kube-system -o jsonpath={.spec.clusterIP}
domain=<cluster-domain>
localdns=<node-local-address>
<cluster-domain> is "cluster.local" by default. <node-local-address> is the local listen IP address chosen for NodeLocal DNSCache.
If kube-proxy is running in IPTABLES mode:
sed -i "s/__PILLAR__LOCAL__DNS__/$localdns/g; s/__PILLAR__DNS__DOMAIN__/$domain/g; s/__PILLAR__DNS__SERVER__/$kubedns/g" nodelocaldns.yaml
__PILLAR__CLUSTER__DNS__ and __PILLAR__UPSTREAM__SERVERS__ will be populated by the node-local-dns pods.
In this mode, node-local-dns pods listen on both the kube-dns service IP as well as <node-local-address>, so pods can lookup DNS records using either IP address.
If kube-proxy is running in IPVS mode:
sed -i "s/__PILLAR__LOCAL__DNS__/$localdns/g; s/__PILLAR__DNS__DOMAIN__/$domain/g; s/,__PILLAR__DNS__SERVER__//g; s/__PILLAR__CLUSTER__DNS__/$kubedns/g" nodelocaldns.yaml
In this mode, node-local-dns pods listen only on <node-local-address>. The node-local-dns interface cannot bind the kube-dns cluster IP since the interface used for IPVS loadbalancing already uses this address.
__PILLAR__UPSTREAM__SERVERS__ will be populated by the node-local-dns pods.
Run kubectl create -f nodelocaldns.yaml
If using kube-proxy in IPVS mode, --cluster-dns flag to kubelet needs to be modified to use <node-local-address> that NodeLocal DNSCache is listening on.
Otherwise, there is no need to modify the value of the --cluster-dns flag, since NodeLocal DNSCache listens on both the kube-dns service IP as well as <node-local-address>.
Once enabled, node-local-dns Pods will run in the kube-system namespace on each of the cluster nodes. This Pod runs CoreDNS in cache mode, so all CoreDNS metrics exposed by the different plugins will be available on a per-node basis.
You can disable this feature by removing the DaemonSet, using kubectl delete -f <manifest> . You should also revert any changes you made to the kubelet configuration.
Kubernetes v1.21 [stable]
This document describes how to configure and use kernel parameters within a Kubernetes cluster using the sysctl interface.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
For some steps, you also need to be able to reconfigure the command line options for the kubelets running on your cluster.
In Linux, the sysctl interface allows an administrator to modify kernel
parameters at runtime. Parameters are available via the /proc/sys/ virtual
process file system. The parameters cover various subsystems such as:
kernel.)net.)vm.)dev.)To get a list of all parameters, you can run
sudo sysctl -a
Sysctls are grouped into safe and unsafe sysctls. In addition to proper namespacing, a safe sysctl must be properly isolated between pods on the same node. This means that setting a safe sysctl for one pod
By far, most of the namespaced sysctls are not necessarily considered safe. The following sysctls are supported in the safe set:
kernel.shm_rmid_forced,net.ipv4.ip_local_port_range,net.ipv4.tcp_syncookies,net.ipv4.ping_group_range (since Kubernetes 1.18).net.ipv4.tcp_syncookies is not namespaced on Linux kernel version 4.4 or lower.
This list will be extended in future Kubernetes versions when the kubelet supports better isolation mechanisms.
All safe sysctls are enabled by default.
All unsafe sysctls are disabled by default and must be allowed manually by the cluster admin on a per-node basis. Pods with disabled unsafe sysctls will be scheduled, but will fail to launch.
With the warning above in mind, the cluster admin can allow certain unsafe sysctls for very special situations such as high-performance or real-time application tuning. Unsafe sysctls are enabled on a node-by-node basis with a flag of the kubelet; for example:
kubelet --allowed-unsafe-sysctls \
'kernel.msg*,net.core.somaxconn' ...
For Minikube, this can be done via the extra-config flag:
minikube start --extra-config="kubelet.allowed-unsafe-sysctls=kernel.msg*,net.core.somaxconn"...
Only namespaced sysctls can be enabled this way.
A number of sysctls are namespaced in today's Linux kernels. This means that they can be set independently for each pod on a node. Only namespaced sysctls are configurable via the pod securityContext within Kubernetes.
The following sysctls are known to be namespaced. This list could change in future versions of the Linux kernel.
kernel.shm*,kernel.msg*,kernel.sem,fs.mqueue.*,net.* that can be set in container networking
namespace. However, there are exceptions (e.g.,
net.netfilter.nf_conntrack_max and net.netfilter.nf_conntrack_expect_max
can be set in container networking namespace but they are unnamespaced).Sysctls with no namespace are called node-level sysctls. If you need to set them, you must manually configure them on each node's operating system, or by using a DaemonSet with privileged containers.
Use the pod securityContext to configure namespaced sysctls. The securityContext applies to all containers in the same pod.
This example uses the pod securityContext to set a safe sysctl
kernel.shm_rmid_forced and two unsafe sysctls net.core.somaxconn and
kernel.msgmax. There is no distinction between safe and unsafe sysctls in
the specification.
apiVersion: v1
kind: Pod
metadata:
name: sysctl-example
spec:
securityContext:
sysctls:
- name: kernel.shm_rmid_forced
value: "0"
- name: net.core.somaxconn
value: "1024"
- name: kernel.msgmax
value: "65536"
...
It is good practice to consider nodes with special sysctl settings as tainted within a cluster, and only schedule pods onto them which need those sysctl settings. It is suggested to use the Kubernetes taints and toleration feature to implement this.
A pod with the unsafe sysctls will fail to launch on any node which has not enabled those two unsafe sysctls explicitly. As with node-level sysctls it is recommended to use taints and toleration feature or taints on nodes to schedule those pods onto the right nodes.
Kubernetes v1.21 [deprecated]
You can further control which sysctls can be set in pods by specifying lists of
sysctls or sysctl patterns in the forbiddenSysctls and/or
allowedUnsafeSysctls fields of the PodSecurityPolicy. A sysctl pattern ends
with a * character, such as kernel.*. A * character on its own matches
all sysctls.
By default, all safe sysctls are allowed.
Both forbiddenSysctls and allowedUnsafeSysctls are lists of plain sysctl names
or sysctl patterns (which end with *). The string * matches all sysctls.
The forbiddenSysctls field excludes specific sysctls. You can forbid a
combination of safe and unsafe sysctls in the list. To forbid setting any
sysctls, use * on its own.
If you specify any unsafe sysctl in the allowedUnsafeSysctls field and it is
not present in the forbiddenSysctls field, that sysctl can be used in Pods
using this PodSecurityPolicy. To allow all unsafe sysctls in the
PodSecurityPolicy to be set, use * on its own.
Do not configure these two fields such that there is overlap, meaning that a given sysctl is both allowed and forbidden.
allowedUnsafeSysctls field
in a PodSecurityPolicy, any pod using such a sysctl will fail to start
if the sysctl is not allowed via the --allowed-unsafe-sysctls kubelet
flag as well on that node.
This example allows unsafe sysctls prefixed with kernel.msg to be set and
disallows setting of the kernel.shm_rmid_forced sysctl.
apiVersion: policy/v1beta1
kind: PodSecurityPolicy
metadata:
name: sysctl-psp
spec:
allowedUnsafeSysctls:
- kernel.msg*
forbiddenSysctls:
- kernel.shm_rmid_forced
...
Kubernetes v1.22 [beta]
The Kubernetes Memory Manager enables the feature of guaranteed memory (and hugepages) allocation for pods in the Guaranteed QoS class.
The Memory Manager employs hint generation protocol to yield the most suitable NUMA affinity for a pod. The Memory Manager feeds the central manager (Topology Manager) with these affinity hints. Based on both the hints and Topology Manager policy, the pod is rejected or admitted to the node.
Moreover, the Memory Manager ensures that the memory which a pod requests is allocated from a minimum number of NUMA nodes.
The Memory Manager is only pertinent to Linux based hosts.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.21. To check the version, enterkubectl version.
To align memory resources with other requested resources in a Pod Spec:
Starting from v1.22, the Memory Manager is enabled by default through MemoryManager feature gate.
Preceding v1.22, the kubelet must be started with the following flag:
--feature-gates=MemoryManager=true
in order to enable the Memory Manager feature.
The Memory Manager currently offers the guaranteed memory (and hugepages) allocation for Pods in Guaranteed QoS class. To immediately put the Memory Manager into operation follow the guidelines in the section Memory Manager configuration, and subsequently, prepare and deploy a Guaranteed pod as illustrated in the section Placing a Pod in the Guaranteed QoS class.
The Memory Manager is a Hint Provider, and it provides topology hints for the Topology Manager which then aligns the requested resources according to these topology hints. It also enforces cgroups (i.e. cpuset.mems) for pods. The complete flow diagram concerning pod admission and deployment process is illustrated in Memory Manager KEP: Design Overview and below:
During this process, the Memory Manager updates its internal counters stored in Node Map and Memory Maps to manage guaranteed memory allocation.
The Memory Manager updates the Node Map during the startup and runtime as follows.
This occurs once a node administrator employs --reserved-memory (section Reserved memory flag). In this case, the Node Map becomes updated to reflect this reservation as illustrated in Memory Manager KEP: Memory Maps at start-up (with examples).
The administrator must provide --reserved-memory flag when Static policy is configured.
Reference Memory Manager KEP: Memory Maps at runtime (with examples) illustrates how a successful pod deployment affects the Node Map, and it also relates to how potential Out-of-Memory (OOM) situations are handled further by Kubernetes or operating system.
Important topic in the context of Memory Manager operation is the management of NUMA groups. Each time pod's memory request is in excess of single NUMA node capacity, the Memory Manager attempts to create a group that comprises several NUMA nodes and features extend memory capacity. The problem has been solved as elaborated in Memory Manager KEP: How to enable the guaranteed memory allocation over many NUMA nodes?. Also, reference Memory Manager KEP: Simulation - how the Memory Manager works? (by examples) illustrates how the management of groups occurs.
Other Managers should be first pre-configured (section Pre-configuration). Next, the Memory Manger feature should be enabled (section Enable the Memory Manager feature) and be run with Static policy (section Static policy). Optionally, some amount of memory can be reserved for system or kubelet processes to increase node stability (section Reserved memory flag).
Memory Manager supports two policies. You can select a policy via a kubelet flag --memory-manager-policy.
Two policies can be selected:
None (default)StaticThis is the default policy and does not affect the memory allocation in any way. It acts the same as if the Memory Manager is not present at all.
The None policy returns default topology hint. This special hint denotes that Hint Provider (Memory Manger in this case) has no preference for NUMA affinity with any resource.
In the case of the Guaranteed pod, the Static Memory Manger policy returns topology hints relating to the set of NUMA nodes where the memory can be guaranteed, and reserves the memory through updating the internal NodeMap object.
In the case of the BestEffort or Burstable pod, the Static Memory Manager policy sends back the default topology hint as there is no request for the guaranteed memory, and does not reserve the memory in the internal NodeMap object.
The Node Allocatable mechanism is commonly used by node administrators to reserve K8S node system resources for the kubelet or operating system processes in order to enhance the node stability. A dedicated set of flags can be used for this purpose to set the total amount of reserved memory for a node. This pre-configured value is subsequently utilized to calculate the real amount of node's "allocatable" memory available to pods.
The Kubernetes scheduler incorporates "allocatable" to optimise pod scheduling process. The foregoing flags include --kube-reserved, --system-reserved and --eviction-threshold. The sum of their values will account for the total amount of reserved memory.
A new --reserved-memory flag was added to Memory Manager to allow for this total reserved memory to be split (by a node administrator) and accordingly reserved across many NUMA nodes.
The flag specifies a comma-separated list of memory reservations per NUMA node. This parameter is only useful in the context of the Memory Manager feature. The Memory Manager will not use this reserved memory for the allocation of container workloads.
For example, if you have a NUMA node "NUMA0" with 10Gi of memory available, and the --reserved-memory was specified to reserve 1Gi of memory at "NUMA0", the Memory Manager assumes that only 9Gi is available for containers.
You can omit this parameter, however, you should be aware that the quantity of reserved memory from all NUMA nodes should be equal to the quantity of memory specified by the Node Allocatable feature. If at least one node allocatable parameter is non-zero, you will need to specify --reserved-memory for at least one NUMA node. In fact, eviction-hard threshold value is equal to 100Mi by default, so if Static policy is used, --reserved-memory is obligatory.
Also, avoid the following configurations:
memory or hugepages-<size> (hugepages of particular <size> should also exist).Syntax:
--reserved-memory N:memory-type1=value1,memory-type2=value2,...
N (integer) - NUMA node index, e.g. 0memory-type (string) - represents memory type:
memory - conventional memoryhugepages-2Mi or hugepages-1Gi - hugepagesvalue (string) - the quantity of reserved memory, e.g. 1GiExample usage:
--reserved-memory 0:memory=1Gi,hugepages-1Gi=2Gi
or
--reserved-memory 0:memory=1Gi --reserved-memory 1:memory=2Gi
When you specify values for --reserved-memory flag, you must comply with the setting that you prior provided via Node Allocatable Feature flags. That is, the following rule must be obeyed for each memory type:
sum(reserved-memory(i)) = kube-reserved + system-reserved + eviction-threshold,
where i is an index of a NUMA node.
If you do not follow the formula above, the Memory Manager will show an error on startup.
In other words, the example above illustrates that for the conventional memory (type=memory), we reserve 3Gi in total, i.e.:
sum(reserved-memory(i)) = reserved-memory(0) + reserved-memory(1) = 1Gi + 2Gi = 3Gi
An example of kubelet command-line arguments relevant to the node Allocatable configuration:
--kube-reserved=cpu=500m,memory=50Mi--system-reserved=cpu=123m,memory=333Mi--eviction-hard=memory.available<500Mi--reserved-memory by that hard eviction threshold. Otherwise, the kubelet will not start Memory Manager and display an error.
Here is an example of a correct configuration:
--feature-gates=MemoryManager=true
--kube-reserved=cpu=4,memory=4Gi
--system-reserved=cpu=1,memory=1Gi
--memory-manager-policy=Static
--reserved-memory 0:memory=3Gi --reserved-memory 1:memory=2148Mi
Let us validate the configuration above:
kube-reserved + system-reserved + eviction-hard(default) = reserved-memory(0) + reserved-memory(1)4GiB + 1GiB + 100MiB = 3GiB + 2148MiB5120MiB + 100MiB = 3072MiB + 2148MiB5220MiB = 5220MiB (which is correct)If the selected policy is anything other than None, the Memory Manager identifies pods that are in the Guaranteed QoS class. The Memory Manager provides specific topology hints to the Topology Manager for each Guaranteed pod. For pods in a QoS class other than Guaranteed, the Memory Manager provides default topology hints to the Topology Manager.
The following excerpts from pod manifests assign a pod to the Guaranteed QoS class.
Pod with integer CPU(s) runs in the Guaranteed QoS class, when requests are equal to limits:
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "2"
example.com/device: "1"
requests:
memory: "200Mi"
cpu: "2"
example.com/device: "1"
Also, a pod sharing CPU(s) runs in the Guaranteed QoS class, when requests are equal to limits.
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "300m"
example.com/device: "1"
requests:
memory: "200Mi"
cpu: "300m"
example.com/device: "1"
Notice that both CPU and memory requests must be specified for a Pod to lend it to Guaranteed QoS class.
The following means can be used to troubleshoot the reason why a pod could not be deployed or became rejected at a node:
This error typically occurs in the following situations:
The error appears in the status of a pod:
# kubectl get pods
NAME READY STATUS RESTARTS AGE
guaranteed 0/1 TopologyAffinityError 0 113s
Use kubectl describe pod <id> or kubectl get events to obtain detailed error message:
Warning TopologyAffinityError 10m kubelet, dell8 Resources cannot be allocated with Topology locality
Search system logs with respect to a particular pod.
The set of hints that Memory Manager generated for the pod can be found in the logs. Also, the set of hints generated by CPU Manager should be present in the logs.
Topology Manager merges these hints to calculate a single best hint. The best hint should be also present in the logs.
The best hint indicates where to allocate all the resources. Topology Manager tests this hint against its current policy, and based on the verdict, it either admits the pod to the node or rejects it.
Also, search the logs for occurrences associated with the Memory Manager, e.g. to find out information about cgroups and cpuset.mems updates.
Let us first deploy a sample Guaranteed pod whose specification is as follows:
apiVersion: v1
kind: Pod
metadata:
name: guaranteed
spec:
containers:
- name: guaranteed
image: consumer
imagePullPolicy: Never
resources:
limits:
cpu: "2"
memory: 150Gi
requests:
cpu: "2"
memory: 150Gi
command: ["sleep","infinity"]
Next, let us log into the node where it was deployed and examine the state file in /var/lib/kubelet/memory_manager_state:
{
"policyName":"Static",
"machineState":{
"0":{
"numberOfAssignments":1,
"memoryMap":{
"hugepages-1Gi":{
"total":0,
"systemReserved":0,
"allocatable":0,
"reserved":0,
"free":0
},
"memory":{
"total":134987354112,
"systemReserved":3221225472,
"allocatable":131766128640,
"reserved":131766128640,
"free":0
}
},
"nodes":[
0,
1
]
},
"1":{
"numberOfAssignments":1,
"memoryMap":{
"hugepages-1Gi":{
"total":0,
"systemReserved":0,
"allocatable":0,
"reserved":0,
"free":0
},
"memory":{
"total":135286722560,
"systemReserved":2252341248,
"allocatable":133034381312,
"reserved":29295144960,
"free":103739236352
}
},
"nodes":[
0,
1
]
}
},
"entries":{
"fa9bdd38-6df9-4cf9-aa67-8c4814da37a8":{
"guaranteed":[
{
"numaAffinity":[
0,
1
],
"type":"memory",
"size":161061273600
}
]
}
},
"checksum":4142013182
}
It can be deduced from the state file that the pod was pinned to both NUMA nodes, i.e.:
"numaAffinity":[
0,
1
],
Pinned term means that pod's memory consumption is constrained (through cgroups configuration) to these NUMA nodes.
This automatically implies that Memory Manager instantiated a new group that comprises these two NUMA nodes, i.e. 0 and 1 indexed NUMA nodes.
Notice that the management of groups is handled in a relatively complex manner, and further elaboration is provided in Memory Manager KEP in this and this sections.
In order to analyse memory resources available in a group, the corresponding entries from NUMA nodes belonging to the group must be added up.
For example, the total amount of free "conventional" memory in the group can be computed by adding up the free memory available at every NUMA node in the group, i.e., in the "memory" section of NUMA node 0 ("free":0) and NUMA node 1 ("free":103739236352). So, the total amount of free "conventional" memory in this group is equal to 0 + 103739236352 bytes.
The line "systemReserved":3221225472 indicates that the administrator of this node reserved 3221225472 bytes (i.e. 3Gi) to serve kubelet and system processes at NUMA node 0, by using --reserved-memory flag.
By employing the API, the information about reserved memory for each container can be retrieved, which is contained in protobuf ContainerMemory message. This information can be retrieved solely for pods in Guaranteed QoS class.
This page shows how to assign a memory request and a memory limit to a Container. A Container is guaranteed to have as much memory as it requests, but is not allowed to use more memory than its limit.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Each node in your cluster must have at least 300 MiB of memory.
A few of the steps on this page require you to run the metrics-server service in your cluster. If you have the metrics-server running, you can skip those steps.
If you are running Minikube, run the following command to enable the metrics-server:
minikube addons enable metrics-server
To see whether the metrics-server is running, or another provider of the resource metrics
API (metrics.k8s.io), run the following command:
kubectl get apiservices
If the resource metrics API is available, the output includes a
reference to metrics.k8s.io.
NAME
v1beta1.metrics.k8s.io
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace mem-example
To specify a memory request for a Container, include the resources:requests field
in the Container's resource manifest. To specify a memory limit, include resources:limits.
In this exercise, you create a Pod that has one Container. The Container has a memory request of 100 MiB and a memory limit of 200 MiB. Here's the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: memory-demo
namespace: mem-example
spec:
containers:
- name: memory-demo-ctr
image: polinux/stress
resources:
limits:
memory: "200Mi"
requests:
memory: "100Mi"
command: ["stress"]
args: ["--vm", "1", "--vm-bytes", "150M", "--vm-hang", "1"]
The args section in the configuration file provides arguments for the Container when it starts.
The "--vm-bytes", "150M" arguments tell the Container to attempt to allocate 150 MiB of memory.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/memory-request-limit.yaml --namespace=mem-example
Verify that the Pod Container is running:
kubectl get pod memory-demo --namespace=mem-example
View detailed information about the Pod:
kubectl get pod memory-demo --output=yaml --namespace=mem-example
The output shows that the one Container in the Pod has a memory request of 100 MiB and a memory limit of 200 MiB.
...
resources:
limits:
memory: 200Mi
requests:
memory: 100Mi
...
Run kubectl top to fetch the metrics for the pod:
kubectl top pod memory-demo --namespace=mem-example
The output shows that the Pod is using about 162,900,000 bytes of memory, which is about 150 MiB. This is greater than the Pod's 100 MiB request, but within the Pod's 200 MiB limit.
NAME CPU(cores) MEMORY(bytes)
memory-demo <something> 162856960
Delete your Pod:
kubectl delete pod memory-demo --namespace=mem-example
A Container can exceed its memory request if the Node has memory available. But a Container is not allowed to use more than its memory limit. If a Container allocates more memory than its limit, the Container becomes a candidate for termination. If the Container continues to consume memory beyond its limit, the Container is terminated. If a terminated Container can be restarted, the kubelet restarts it, as with any other type of runtime failure.
In this exercise, you create a Pod that attempts to allocate more memory than its limit. Here is the configuration file for a Pod that has one Container with a memory request of 50 MiB and a memory limit of 100 MiB:
apiVersion: v1
kind: Pod
metadata:
name: memory-demo-2
namespace: mem-example
spec:
containers:
- name: memory-demo-2-ctr
image: polinux/stress
resources:
requests:
memory: "50Mi"
limits:
memory: "100Mi"
command: ["stress"]
args: ["--vm", "1", "--vm-bytes", "250M", "--vm-hang", "1"]
In the args section of the configuration file, you can see that the Container
will attempt to allocate 250 MiB of memory, which is well above the 100 MiB limit.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/memory-request-limit-2.yaml --namespace=mem-example
View detailed information about the Pod:
kubectl get pod memory-demo-2 --namespace=mem-example
At this point, the Container might be running or killed. Repeat the preceding command until the Container is killed:
NAME READY STATUS RESTARTS AGE
memory-demo-2 0/1 OOMKilled 1 24s
Get a more detailed view of the Container status:
kubectl get pod memory-demo-2 --output=yaml --namespace=mem-example
The output shows that the Container was killed because it is out of memory (OOM):
lastState:
terminated:
containerID: docker://65183c1877aaec2e8427bc95609cc52677a454b56fcb24340dbd22917c23b10f
exitCode: 137
finishedAt: 2017-06-20T20:52:19Z
reason: OOMKilled
startedAt: null
The Container in this exercise can be restarted, so the kubelet restarts it. Repeat this command several times to see that the Container is repeatedly killed and restarted:
kubectl get pod memory-demo-2 --namespace=mem-example
The output shows that the Container is killed, restarted, killed again, restarted again, and so on:
kubectl get pod memory-demo-2 --namespace=mem-example
NAME READY STATUS RESTARTS AGE
memory-demo-2 0/1 OOMKilled 1 37s
kubectl get pod memory-demo-2 --namespace=mem-example
NAME READY STATUS RESTARTS AGE
memory-demo-2 1/1 Running 2 40s
View detailed information about the Pod history:
kubectl describe pod memory-demo-2 --namespace=mem-example
The output shows that the Container starts and fails repeatedly:
... Normal Created Created container with id 66a3a20aa7980e61be4922780bf9d24d1a1d8b7395c09861225b0eba1b1f8511
... Warning BackOff Back-off restarting failed container
View detailed information about your cluster's Nodes:
kubectl describe nodes
The output includes a record of the Container being killed because of an out-of-memory condition:
Warning OOMKilling Memory cgroup out of memory: Kill process 4481 (stress) score 1994 or sacrifice child
Delete your Pod:
kubectl delete pod memory-demo-2 --namespace=mem-example
Memory requests and limits are associated with Containers, but it is useful to think of a Pod as having a memory request and limit. The memory request for the Pod is the sum of the memory requests for all the Containers in the Pod. Likewise, the memory limit for the Pod is the sum of the limits of all the Containers in the Pod.
Pod scheduling is based on requests. A Pod is scheduled to run on a Node only if the Node has enough available memory to satisfy the Pod's memory request.
In this exercise, you create a Pod that has a memory request so big that it exceeds the capacity of any Node in your cluster. Here is the configuration file for a Pod that has one Container with a request for 1000 GiB of memory, which likely exceeds the capacity of any Node in your cluster.
apiVersion: v1
kind: Pod
metadata:
name: memory-demo-3
namespace: mem-example
spec:
containers:
- name: memory-demo-3-ctr
image: polinux/stress
resources:
limits:
memory: "1000Gi"
requests:
memory: "1000Gi"
command: ["stress"]
args: ["--vm", "1", "--vm-bytes", "150M", "--vm-hang", "1"]
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/memory-request-limit-3.yaml --namespace=mem-example
View the Pod status:
kubectl get pod memory-demo-3 --namespace=mem-example
The output shows that the Pod status is PENDING. That is, the Pod is not scheduled to run on any Node, and it will remain in the PENDING state indefinitely:
kubectl get pod memory-demo-3 --namespace=mem-example
NAME READY STATUS RESTARTS AGE
memory-demo-3 0/1 Pending 0 25s
View detailed information about the Pod, including events:
kubectl describe pod memory-demo-3 --namespace=mem-example
The output shows that the Container cannot be scheduled because of insufficient memory on the Nodes:
Events:
... Reason Message
------ -------
... FailedScheduling No nodes are available that match all of the following predicates:: Insufficient memory (3).
The memory resource is measured in bytes. You can express memory as a plain integer or a fixed-point integer with one of these suffixes: E, P, T, G, M, K, Ei, Pi, Ti, Gi, Mi, Ki. For example, the following represent approximately the same value:
128974848, 129e6, 129M , 123Mi
Delete your Pod:
kubectl delete pod memory-demo-3 --namespace=mem-example
If you do not specify a memory limit for a Container, one of the following situations applies:
The Container has no upper bound on the amount of memory it uses. The Container could use all of the memory available on the Node where it is running which in turn could invoke the OOM Killer. Further, in case of an OOM Kill, a container with no resource limits will have a greater chance of being killed.
The Container is running in a namespace that has a default memory limit, and the Container is automatically assigned the default limit. Cluster administrators can use a LimitRange to specify a default value for the memory limit.
By configuring memory requests and limits for the Containers that run in your cluster, you can make efficient use of the memory resources available on your cluster's Nodes. By keeping a Pod's memory request low, you give the Pod a good chance of being scheduled. By having a memory limit that is greater than the memory request, you accomplish two things:
Delete your namespace. This deletes all the Pods that you created for this task:
kubectl delete namespace mem-example
This page shows how to assign a CPU request and a CPU limit to a container. Containers cannot use more CPU than the configured limit. Provided the system has CPU time free, a container is guaranteed to be allocated as much CPU as it requests.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Your cluster must have at least 1 CPU available for use to run the task examples.
A few of the steps on this page require you to run the metrics-server service in your cluster. If you have the metrics-server running, you can skip those steps.
If you are running Minikube, run the following command to enable metrics-server:
minikube addons enable metrics-server
To see whether metrics-server (or another provider of the resource metrics
API, metrics.k8s.io) is running, type the following command:
kubectl get apiservices
If the resource metrics API is available, the output will include a
reference to metrics.k8s.io.
NAME
v1beta1.metrics.k8s.io
Create a Namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace cpu-example
To specify a CPU request for a container, include the resources:requests field
in the Container resource manifest. To specify a CPU limit, include resources:limits.
In this exercise, you create a Pod that has one container. The container has a request of 0.5 CPU and a limit of 1 CPU. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: cpu-demo
namespace: cpu-example
spec:
containers:
- name: cpu-demo-ctr
image: vish/stress
resources:
limits:
cpu: "1"
requests:
cpu: "0.5"
args:
- -cpus
- "2"
The args section of the configuration file provides arguments for the container when it starts.
The -cpus "2" argument tells the Container to attempt to use 2 CPUs.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/cpu-request-limit.yaml --namespace=cpu-example
Verify that the Pod is running:
kubectl get pod cpu-demo --namespace=cpu-example
View detailed information about the Pod:
kubectl get pod cpu-demo --output=yaml --namespace=cpu-example
The output shows that the one container in the Pod has a CPU request of 500 milliCPU and a CPU limit of 1 CPU.
resources:
limits:
cpu: "1"
requests:
cpu: 500m
Use kubectl top to fetch the metrics for the pod:
kubectl top pod cpu-demo --namespace=cpu-example
This example output shows that the Pod is using 974 milliCPU, which is slightly less than the limit of 1 CPU specified in the Pod configuration.
NAME CPU(cores) MEMORY(bytes)
cpu-demo 974m <something>
Recall that by setting -cpu "2", you configured the Container to attempt to use 2 CPUs, but the Container is only being allowed to use about 1 CPU. The container's CPU use is being throttled, because the container is attempting to use more CPU resources than its limit.
The CPU resource is measured in CPU units. One CPU, in Kubernetes, is equivalent to:
Fractional values are allowed. A Container that requests 0.5 CPU is guaranteed half as much CPU as a Container that requests 1 CPU. You can use the suffix m to mean milli. For example 100m CPU, 100 milliCPU, and 0.1 CPU are all the same. Precision finer than 1m is not allowed.
CPU is always requested as an absolute quantity, never as a relative quantity; 0.1 is the same amount of CPU on a single-core, dual-core, or 48-core machine.
Delete your Pod:
kubectl delete pod cpu-demo --namespace=cpu-example
CPU requests and limits are associated with Containers, but it is useful to think of a Pod as having a CPU request and limit. The CPU request for a Pod is the sum of the CPU requests for all the Containers in the Pod. Likewise, the CPU limit for a Pod is the sum of the CPU limits for all the Containers in the Pod.
Pod scheduling is based on requests. A Pod is scheduled to run on a Node only if the Node has enough CPU resources available to satisfy the Pod CPU request.
In this exercise, you create a Pod that has a CPU request so big that it exceeds the capacity of any Node in your cluster. Here is the configuration file for a Pod that has one Container. The Container requests 100 CPU, which is likely to exceed the capacity of any Node in your cluster.
apiVersion: v1
kind: Pod
metadata:
name: cpu-demo-2
namespace: cpu-example
spec:
containers:
- name: cpu-demo-ctr-2
image: vish/stress
resources:
limits:
cpu: "100"
requests:
cpu: "100"
args:
- -cpus
- "2"
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/cpu-request-limit-2.yaml --namespace=cpu-example
View the Pod status:
kubectl get pod cpu-demo-2 --namespace=cpu-example
The output shows that the Pod status is Pending. That is, the Pod has not been scheduled to run on any Node, and it will remain in the Pending state indefinitely:
NAME READY STATUS RESTARTS AGE
cpu-demo-2 0/1 Pending 0 7m
View detailed information about the Pod, including events:
kubectl describe pod cpu-demo-2 --namespace=cpu-example
The output shows that the Container cannot be scheduled because of insufficient CPU resources on the Nodes:
Events:
Reason Message
------ -------
FailedScheduling No nodes are available that match all of the following predicates:: Insufficient cpu (3).
Delete your Pod:
kubectl delete pod cpu-demo-2 --namespace=cpu-example
If you do not specify a CPU limit for a Container, then one of these situations applies:
The Container has no upper bound on the CPU resources it can use. The Container could use all of the CPU resources available on the Node where it is running.
The Container is running in a namespace that has a default CPU limit, and the Container is automatically assigned the default limit. Cluster administrators can use a LimitRange to specify a default value for the CPU limit.
If you specify a CPU limit for a Container but do not specify a CPU request, Kubernetes automatically assigns a CPU request that matches the limit. Similarly, if a Container specifies its own memory limit, but does not specify a memory request, Kubernetes automatically assigns a memory request that matches the limit.
By configuring the CPU requests and limits of the Containers that run in your cluster, you can make efficient use of the CPU resources available on your cluster Nodes. By keeping a Pod CPU request low, you give the Pod a good chance of being scheduled. By having a CPU limit that is greater than the CPU request, you accomplish two things:
Delete your namespace:
kubectl delete namespace cpu-example
Kubernetes v1.18 [stable]
This page shows how to configure Group Managed Service Accounts (GMSA) for Pods and containers that will run on Windows nodes. Group Managed Service Accounts are a specific type of Active Directory account that provides automatic password management, simplified service principal name (SPN) management, and the ability to delegate the management to other administrators across multiple servers.
In Kubernetes, GMSA credential specs are configured at a Kubernetes cluster-wide scope as Custom Resources. Windows Pods, as well as individual containers within a Pod, can be configured to use a GMSA for domain based functions (e.g. Kerberos authentication) when interacting with other Windows services. As of v1.16, the Docker runtime supports GMSA for Windows workloads.
You need to have a Kubernetes cluster and the kubectl command-line tool must be configured to communicate with your cluster. The cluster is expected to have Windows worker nodes. This section covers a set of initial steps required once for each cluster:
A CustomResourceDefinition(CRD) for GMSA credential spec resources needs to be configured on the cluster to define the custom resource type GMSACredentialSpec. Download the GMSA CRD YAML and save it as gmsa-crd.yaml.
Next, install the CRD with kubectl apply -f gmsa-crd.yaml
Two webhooks need to be configured on the Kubernetes cluster to populate and validate GMSA credential spec references at the Pod or container level:
A mutating webhook that expands references to GMSAs (by name from a Pod specification) into the full credential spec in JSON form within the Pod spec.
A validating webhook ensures all references to GMSAs are authorized to be used by the Pod service account.
Installing the above webhooks and associated objects require the steps below:
Create a certificate key pair (that will be used to allow the webhook container to communicate to the cluster)
Install a secret with the certificate from above.
Create a deployment for the core webhook logic.
Create the validating and mutating webhook configurations referring to the deployment.
A script can be used to deploy and configure the GMSA webhooks and associated objects mentioned above. The script can be run with a --dry-run=server option to allow you to review the changes that would be made to your cluster.
The YAML template used by the script may also be used to deploy the webhooks and associated objects manually (with appropriate substitutions for the parameters)
Before Pods in Kubernetes can be configured to use GMSAs, the desired GMSAs need to be provisioned in Active Directory as described in the Windows GMSA documentation. Windows worker nodes (that are part of the Kubernetes cluster) need to be configured in Active Directory to access the secret credentials associated with the desired GMSA as described in the Windows GMSA documentation
With the GMSACredentialSpec CRD installed (as described earlier), custom resources containing GMSA credential specs can be configured. The GMSA credential spec does not contain secret or sensitive data. It is information that a container runtime can use to describe the desired GMSA of a container to Windows. GMSA credential specs can be generated in YAML format with a utility PowerShell script.
Following are the steps for generating a GMSA credential spec YAML manually in JSON format and then converting it:
Import the CredentialSpec module: ipmo CredentialSpec.psm1
Create a credential spec in JSON format using New-CredentialSpec. To create a GMSA credential spec named WebApp1, invoke New-CredentialSpec -Name WebApp1 -AccountName WebApp1 -Domain $(Get-ADDomain -Current LocalComputer)
Use Get-CredentialSpec to show the path of the JSON file.
Convert the credspec file from JSON to YAML format and apply the necessary header fields apiVersion, kind, metadata and credspec to make it a GMSACredentialSpec custom resource that can be configured in Kubernetes.
The following YAML configuration describes a GMSA credential spec named gmsa-WebApp1:
apiVersion: windows.k8s.io/v1alpha1
kind: GMSACredentialSpec
metadata:
name: gmsa-WebApp1 #This is an arbitrary name but it will be used as a reference
credspec:
ActiveDirectoryConfig:
GroupManagedServiceAccounts:
- Name: WebApp1 #Username of the GMSA account
Scope: CONTOSO #NETBIOS Domain Name
- Name: WebApp1 #Username of the GMSA account
Scope: contoso.com #DNS Domain Name
CmsPlugins:
- ActiveDirectory
DomainJoinConfig:
DnsName: contoso.com #DNS Domain Name
DnsTreeName: contoso.com #DNS Domain Name Root
Guid: 244818ae-87ac-4fcd-92ec-e79e5252348a #GUID
MachineAccountName: WebApp1 #Username of the GMSA account
NetBiosName: CONTOSO #NETBIOS Domain Name
Sid: S-1-5-21-2126449477-2524075714-3094792973 #SID of GMSA
The above credential spec resource may be saved as gmsa-Webapp1-credspec.yaml and applied to the cluster using: kubectl apply -f gmsa-Webapp1-credspec.yml
A cluster role needs to be defined for each GMSA credential spec resource. This authorizes the use verb on a specific GMSA resource by a subject which is typically a service account. The following example shows a cluster role that authorizes usage of the gmsa-WebApp1 credential spec from above. Save the file as gmsa-webapp1-role.yaml and apply using kubectl apply -f gmsa-webapp1-role.yaml
#Create the Role to read the credspec
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
name: webapp1-role
rules:
- apiGroups: ["windows.k8s.io"]
resources: ["gmsacredentialspecs"]
verbs: ["use"]
resourceNames: ["gmsa-WebApp1"]
A service account (that Pods will be configured with) needs to be bound to the cluster role create above. This authorizes the service account to use the desired GMSA credential spec resource. The following shows the default service account being bound to a cluster role webapp1-role to use gmsa-WebApp1 credential spec resource created above.
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
name: allow-default-svc-account-read-on-gmsa-WebApp1
namespace: default
subjects:
- kind: ServiceAccount
name: default
namespace: default
roleRef:
kind: ClusterRole
name: webapp1-role
apiGroup: rbac.authorization.k8s.io
The Pod spec field securityContext.windowsOptions.gmsaCredentialSpecName is used to specify references to desired GMSA credential spec custom resources in Pod specs. This configures all containers in the Pod spec to use the specified GMSA. A sample Pod spec with the annotation populated to refer to gmsa-WebApp1:
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
run: with-creds
name: with-creds
namespace: default
spec:
replicas: 1
selector:
matchLabels:
run: with-creds
template:
metadata:
labels:
run: with-creds
spec:
securityContext:
windowsOptions:
gmsaCredentialSpecName: gmsa-webapp1
containers:
- image: mcr.microsoft.com/windows/servercore/iis:windowsservercore-ltsc2019
imagePullPolicy: Always
name: iis
nodeSelector:
kubernetes.io/os: windows
Individual containers in a Pod spec can also specify the desired GMSA credspec using a per-container securityContext.windowsOptions.gmsaCredentialSpecName field. For example:
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
run: with-creds
name: with-creds
namespace: default
spec:
replicas: 1
selector:
matchLabels:
run: with-creds
template:
metadata:
labels:
run: with-creds
spec:
containers:
- image: mcr.microsoft.com/windows/servercore/iis:windowsservercore-ltsc2019
imagePullPolicy: Always
name: iis
securityContext:
windowsOptions:
gmsaCredentialSpecName: gmsa-Webapp1
nodeSelector:
kubernetes.io/os: windows
As Pod specs with GMSA fields populated (as described above) are applied in a cluster, the following sequence of events take place:
The mutating webhook resolves and expands all references to GMSA credential spec resources to the contents of the GMSA credential spec.
The validating webhook ensures the service account associated with the Pod is authorized for the use verb on the specified GMSA credential spec.
The container runtime configures each Windows container with the specified GMSA credential spec so that the container can assume the identity of the GMSA in Active Directory and access services in the domain using that identity.
On Windows Server 2019, in order to use GMSA with containerd, you must be running OS Build 17763.1817 (or later) which can be installed using the patch KB5000822.
There is also a known issue with containerd that occurs when trying to connect to SMB shares from Pods. Once you have configured GMSA, the pod will be unable to connect to the share using the hostname or FQDN, but connecting to the share using an IP address works as expected.
ping adserver.ad.local
and correctly resolves the hostname to an IPv4 address. The output is similar to:
Pinging adserver.ad.local [192.168.111.18] with 32 bytes of data:
Reply from 192.168.111.18: bytes=32 time=6ms TTL=124
Reply from 192.168.111.18: bytes=32 time=5ms TTL=124
Reply from 192.168.111.18: bytes=32 time=5ms TTL=124
Reply from 192.168.111.18: bytes=32 time=5ms TTL=124
However, when attempting to browse the directory using the hostname
cd \\adserver.ad.local\test
you see an error that implies the target share doesn't exist:
cd : Cannot find path '\\adserver.ad.local\test' because it does not exist.
At line:1 char:1
+ cd \\adserver.ad.local\test
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : ObjectNotFound: (\\adserver.ad.local\test:String) [Set-Location], ItemNotFoundException
+ FullyQualifiedErrorId : PathNotFound,Microsoft.PowerShell.Commands.SetLocationCommand
but you notice that the error disappears if you browse to the share using its IPv4 address instead; for example:
cd \\192.168.111.18\test
After you change into a directory within the share, you see a prompt similar to:
Microsoft.PowerShell.Core\FileSystem::\\192.168.111.18\test>
To correct the behaviour you must run the following on the node reg add "HKLM\SYSTEM\CurrentControlSet\Services\hns\State" /v EnableCompartmentNamespace /t REG_DWORD /d 1 to add the required registry key. This node change will only take effect in newly created pods, meaning you must now recreate any running pods which require access to SMB shares.
If you are having difficulties getting GMSA to work in your environment, there are a few troubleshooting steps you can take.
First, make sure the credspec has been passed to the Pod. To do this you will need to exec into one of your Pods and check the output of the nltest.exe /parentdomain command.
In the example below the Pod did not get the credspec correctly:
kubectl exec -it iis-auth-7776966999-n5nzr powershell.exe
nltest.exe /parentdomain results in the following error:
Getting parent domain failed: Status = 1722 0x6ba RPC_S_SERVER_UNAVAILABLE
If your Pod did get the credspec correctly, then next check communication with the domain. First, from inside of your Pod, quickly do an nslookup to find the root of your domain.
This will tell us 3 things:
If the DNS and communication test passes, next you will need to check if the Pod has established secure channel communication with the domain. To do this, again, exec into your Pod and run the nltest.exe /query command.
nltest.exe /query
Results in the following output:
I_NetLogonControl failed: Status = 1722 0x6ba RPC_S_SERVER_UNAVAILABLE
This tells us that for some reason, the Pod was unable to logon to the domain using the account specified in the credspec. You can try to repair the secure channel by running the following:
nltest /sc_reset:domain.example
If the command is successful you will see and output similar to this:
Flags: 30 HAS_IP HAS_TIMESERV
Trusted DC Name \\dc10.domain.example
Trusted DC Connection Status Status = 0 0x0 NERR_Success
The command completed successfully
If the above corrects the error, you can automate the step by adding the following lifecycle hook to your Pod spec. If it did not correct the error, you will need to examine your credspec again and confirm that it is correct and complete.
image: registry.domain.example/iis-auth:1809v1
lifecycle:
postStart:
exec:
command: ["powershell.exe","-command","do { Restart-Service -Name netlogon } while ( $($Result = (nltest.exe /query); if ($Result -like '*0x0 NERR_Success*') {return $true} else {return $false}) -eq $false)"]
imagePullPolicy: IfNotPresent
If you add the lifecycle section show above to your Pod spec, the Pod will execute the commands listed to restart the netlogon service until the nltest.exe /query command exits without error.
Kubernetes v1.18 [stable]
This page shows how to use the runAsUserName setting for Pods and containers that will run on Windows nodes. This is roughly equivalent of the Linux-specific runAsUser setting, allowing you to run applications in a container as a different username than the default.
You need to have a Kubernetes cluster and the kubectl command-line tool must be configured to communicate with your cluster. The cluster is expected to have Windows worker nodes where pods with containers running Windows workloads will get scheduled.
To specify the username with which to execute the Pod's container processes, include the securityContext field (PodSecurityContext) in the Pod specification, and within it, the windowsOptions (WindowsSecurityContextOptions) field containing the runAsUserName field.
The Windows security context options that you specify for a Pod apply to all Containers and init Containers in the Pod.
Here is a configuration file for a Windows Pod that has the runAsUserName field set:
apiVersion: v1
kind: Pod
metadata:
name: run-as-username-pod-demo
spec:
securityContext:
windowsOptions:
runAsUserName: "ContainerUser"
containers:
- name: run-as-username-demo
image: mcr.microsoft.com/windows/servercore:ltsc2019
command: ["ping", "-t", "localhost"]
nodeSelector:
kubernetes.io/os: windows
Create the Pod:
kubectl apply -f https://k8s.io/examples/windows/run-as-username-pod.yaml
Verify that the Pod's Container is running:
kubectl get pod run-as-username-pod-demo
Get a shell to the running Container:
kubectl exec -it run-as-username-pod-demo -- powershell
Check that the shell is running user the correct username:
echo $env:USERNAME
The output should be:
ContainerUser
To specify the username with which to execute a Container's processes, include the securityContext field (SecurityContext) in the Container manifest, and within it, the windowsOptions (WindowsSecurityContextOptions) field containing the runAsUserName field.
The Windows security context options that you specify for a Container apply only to that individual Container, and they override the settings made at the Pod level.
Here is the configuration file for a Pod that has one Container, and the runAsUserName field is set at the Pod level and the Container level:
apiVersion: v1
kind: Pod
metadata:
name: run-as-username-container-demo
spec:
securityContext:
windowsOptions:
runAsUserName: "ContainerUser"
containers:
- name: run-as-username-demo
image: mcr.microsoft.com/windows/servercore:ltsc2019
command: ["ping", "-t", "localhost"]
securityContext:
windowsOptions:
runAsUserName: "ContainerAdministrator"
nodeSelector:
kubernetes.io/os: windows
Create the Pod:
kubectl apply -f https://k8s.io/examples/windows/run-as-username-container.yaml
Verify that the Pod's Container is running:
kubectl get pod run-as-username-container-demo
Get a shell to the running Container:
kubectl exec -it run-as-username-container-demo -- powershell
Check that the shell is running user the correct username (the one set at the Container level):
echo $env:USERNAME
The output should be:
ContainerAdministrator
In order to use this feature, the value set in the runAsUserName field must be a valid username. It must have the following format: DOMAIN\USER, where DOMAIN\ is optional. Windows user names are case insensitive. Additionally, there are some restrictions regarding the DOMAIN and USER:
runAsUserName field cannot be empty, and it cannot contain control characters (ASCII values: 0x00-0x1F, 0x7F)DOMAIN must be either a NetBios name, or a DNS name, each with their own restrictions:
. (dot), and cannot contain the following characters: \ / : * ? " < > |. (dot) or - (dash).USER must have at most 20 characters, it cannot contain only dots or spaces, and it cannot contain the following characters: " / \ [ ] : ; | = , + * ? < > @.Examples of acceptable values for the runAsUserName field: ContainerAdministrator, ContainerUser, NT AUTHORITY\NETWORK SERVICE, NT AUTHORITY\LOCAL SERVICE.
For more information about these limtations, check here and here.
Kubernetes v1.22 [alpha]
Windows HostProcess containers enable you to run containerized workloads on a Windows host. These containers operate as normal processes but have access to the host network namespace, storage, and devices when given the appropriate user privileges. HostProcess containers can be used to deploy network plugins, storage configurations, device plugins, kube-proxy, and other components to Windows nodes without the need for dedicated proxies or the direct installation of host services.
Administrative tasks such as installation of security patches, event log collection, and more can be performed without requiring cluster operators to log onto each Window node. HostProcess containers can run as any user that is available on the host or is in the domain of the host machine, allowing administrators to restrict resource access through user permissions. While neither filesystem or process isolation are supported, a new volume is created on the host upon starting the container to give it a clean and consolidated workspace. HostProcess containers can also be built on top of existing Windows base images and do not inherit the same compatibility requirements as Windows server containers, meaning that the version of the base images does not need to match that of the host. HostProcess containers also support volume mounts within the container volume.
Your Kubernetes server must be at or later than version 1.22.
To check the version, enter kubectl version.
To enable HostProcess containers while in Alpha you need to pass the following feature gate flag to kubelet and kube-apiserver. See Features Gates documentation for more details.
--feature-gates=WindowsHostProcessContainers=true
You can use the latest version of Containerd (v1.5.4+) with the following settings using the containerd v2 configuration. Add these annotations to any runtime configurations were you wish to enable the HostProcess container feature.
[plugins]
[plugins."io.containerd.grpc.v1.cri"]
[plugins."io.containerd.grpc.v1.cri".containerd]
[plugins."io.containerd.grpc.v1.cri".containerd.default_runtime]
container_annotations = ["microsoft.com/hostprocess-container"]
pod_annotations = ["microsoft.com/hostprocess-container"]
[plugins."io.containerd.grpc.v1.cri".containerd.runtimes]
[plugins."io.containerd.grpc.v1.cri".containerd.runtimes.runhcs-wcow-process]
container_annotations = ["microsoft.com/hostprocess-container"]
pod_annotations = ["microsoft.com/hostprocess-container"]
The current versions of containerd ship with a version of hcsshim that does not have support.
You will need to build a version of hcsshim from the main branch following the
instructions in hcsshim.
Once the containerd shim is built you can replace the file in your contianerd installation.
For example if you followed the instructions to
install containerd
replace the containerd-shim-runhcs-v1.exe is installed at $Env:ProgramFiles\containerd with the newly built shim.
Enabling a Windows HostProcess pod requires setting the right configurations in the pod security configuration. Of the policies defined in the Pod Security Standards HostProcess pods are disallowed by the baseline and restricted policies. It is therefore recommended that HostProcess pods run in alignment with the privileged profile.
When running under the privileged policy, here are the configurations which need to be set to enable the creation of a HostProcess pod:
| Control | Policy |
| Windows HostProcess |
Windows pods offer the ability to run HostProcess containers which enables privileged access to the Windows node. Allowed Values
|
| Host Networking |
Will be in host network by default initially. Support to set network to a different compartment may be desirable in the future. Allowed Values
|
| runAsUsername |
Specification of which user the HostProcess container should run as is required for the pod spec. Allowed Values
|
| runAsNonRoot |
Because HostProcess containers have privileged access to the host, the runAsNonRoot field cannot be set to true. Allowed Values
|
spec:
securityContext:
windowsOptions:
hostProcess: true
runAsUserName: "NT AUTHORITY\\Local service"
hostNetwork: true
containers:
- name: test
image: image1:latest
command:
- ping
- -t
- 127.0.0.1
nodeSelector:
"kubernetes.io/os": windows
HostProcess containers support the ability to mount volumes within the container volume space.
Applications running inside the container can access volume mounts directly via relative or
absolute paths. An environment variable $CONTAINER_SANDBOX_MOUNT_POINT is set upon container
creation and provides the absolute host path to the container volume. Relative paths are based
upon the Pod.containers.volumeMounts.mountPath configuration.
To access service account tokens the following path structures are supported within the container:
.\var\run\secrets\kubernetes.io\serviceaccount\
$CONTAINER_SANDBOX_MOUNT_POINT\var\run\secrets\kubernetes.io\serviceaccount\
HostProcess containers support the ability to run as one of three supported Windows service accounts:
You should select an appropriate Windows service account for each HostProcess container, aiming to limit the degree of privileges so as to avoid accidental (or even malicious) damage to the host. The LocalSystem service account has the highest level of privilege of the three and should be used only if absolutely necessary. Where possible, use the LocalService service account as it is the least privileged of the three options.
This page shows how to configure Pods so that they will be assigned particular Quality of Service (QoS) classes. Kubernetes uses QoS classes to make decisions about scheduling and evicting Pods.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
When Kubernetes creates a Pod it assigns one of these QoS classes to the Pod:
Create a namespace so that the resources you create in this exercise are isolated from the rest of your cluster.
kubectl create namespace qos-example
For a Pod to be given a QoS class of Guaranteed:
These restrictions apply to init containers and app containers equally.
Here is the configuration file for a Pod that has one Container. The Container has a memory limit and a memory request, both equal to 200 MiB. The Container has a CPU limit and a CPU request, both equal to 700 milliCPU:
apiVersion: v1
kind: Pod
metadata:
name: qos-demo
namespace: qos-example
spec:
containers:
- name: qos-demo-ctr
image: nginx
resources:
limits:
memory: "200Mi"
cpu: "700m"
requests:
memory: "200Mi"
cpu: "700m"
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/qos/qos-pod.yaml --namespace=qos-example
View detailed information about the Pod:
kubectl get pod qos-demo --namespace=qos-example --output=yaml
The output shows that Kubernetes gave the Pod a QoS class of Guaranteed. The output also verifies that the Pod Container has a memory request that matches its memory limit, and it has a CPU request that matches its CPU limit.
spec:
containers:
...
resources:
limits:
cpu: 700m
memory: 200Mi
requests:
cpu: 700m
memory: 200Mi
...
status:
qosClass: Guaranteed
Delete your Pod:
kubectl delete pod qos-demo --namespace=qos-example
A Pod is given a QoS class of Burstable if:
Here is the configuration file for a Pod that has one Container. The Container has a memory limit of 200 MiB and a memory request of 100 MiB.
apiVersion: v1
kind: Pod
metadata:
name: qos-demo-2
namespace: qos-example
spec:
containers:
- name: qos-demo-2-ctr
image: nginx
resources:
limits:
memory: "200Mi"
requests:
memory: "100Mi"
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/qos/qos-pod-2.yaml --namespace=qos-example
View detailed information about the Pod:
kubectl get pod qos-demo-2 --namespace=qos-example --output=yaml
The output shows that Kubernetes gave the Pod a QoS class of Burstable.
spec:
containers:
- image: nginx
imagePullPolicy: Always
name: qos-demo-2-ctr
resources:
limits:
memory: 200Mi
requests:
memory: 100Mi
...
status:
qosClass: Burstable
Delete your Pod:
kubectl delete pod qos-demo-2 --namespace=qos-example
For a Pod to be given a QoS class of BestEffort, the Containers in the Pod must not have any memory or CPU limits or requests.
Here is the configuration file for a Pod that has one Container. The Container has no memory or CPU limits or requests:
apiVersion: v1
kind: Pod
metadata:
name: qos-demo-3
namespace: qos-example
spec:
containers:
- name: qos-demo-3-ctr
image: nginx
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/qos/qos-pod-3.yaml --namespace=qos-example
View detailed information about the Pod:
kubectl get pod qos-demo-3 --namespace=qos-example --output=yaml
The output shows that Kubernetes gave the Pod a QoS class of BestEffort.
spec:
containers:
...
resources: {}
...
status:
qosClass: BestEffort
Delete your Pod:
kubectl delete pod qos-demo-3 --namespace=qos-example
Here is the configuration file for a Pod that has two Containers. One container specifies a memory request of 200 MiB. The other Container does not specify any requests or limits.
apiVersion: v1
kind: Pod
metadata:
name: qos-demo-4
namespace: qos-example
spec:
containers:
- name: qos-demo-4-ctr-1
image: nginx
resources:
requests:
memory: "200Mi"
- name: qos-demo-4-ctr-2
image: redis
Notice that this Pod meets the criteria for QoS class Burstable. That is, it does not meet the criteria for QoS class Guaranteed, and one of its Containers has a memory request.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/qos/qos-pod-4.yaml --namespace=qos-example
View detailed information about the Pod:
kubectl get pod qos-demo-4 --namespace=qos-example --output=yaml
The output shows that Kubernetes gave the Pod a QoS class of Burstable:
spec:
containers:
...
name: qos-demo-4-ctr-1
resources:
requests:
memory: 200Mi
...
name: qos-demo-4-ctr-2
resources: {}
...
status:
qosClass: Burstable
Delete your Pod:
kubectl delete pod qos-demo-4 --namespace=qos-example
Delete your namespace:
kubectl delete namespace qos-example
Kubernetes v1.22 [stable]
This page shows how to assign extended resources to a Container.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Before you do this exercise, do the exercise in Advertise Extended Resources for a Node. That will configure one of your Nodes to advertise a dongle resource.
To request an extended resource, include the resources:requests field in your
Container manifest. Extended resources are fully qualified with any domain outside of
*.kubernetes.io/. Valid extended resource names have the form example.com/foo where
example.com is replaced with your organization's domain and foo is a
descriptive resource name.
Here is the configuration file for a Pod that has one Container:
apiVersion: v1
kind: Pod
metadata:
name: extended-resource-demo
spec:
containers:
- name: extended-resource-demo-ctr
image: nginx
resources:
requests:
example.com/dongle: 3
limits:
example.com/dongle: 3
In the configuration file, you can see that the Container requests 3 dongles.
Create a Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/extended-resource-pod.yaml
Verify that the Pod is running:
kubectl get pod extended-resource-demo
Describe the Pod:
kubectl describe pod extended-resource-demo
The output shows dongle requests:
Limits:
example.com/dongle: 3
Requests:
example.com/dongle: 3
Here is the configuration file for a Pod that has one Container. The Container requests two dongles.
apiVersion: v1
kind: Pod
metadata:
name: extended-resource-demo-2
spec:
containers:
- name: extended-resource-demo-2-ctr
image: nginx
resources:
requests:
example.com/dongle: 2
limits:
example.com/dongle: 2
Kubernetes will not be able to satisfy the request for two dongles, because the first Pod used three of the four available dongles.
Attempt to create a Pod:
kubectl apply -f https://k8s.io/examples/pods/resource/extended-resource-pod-2.yaml
Describe the Pod
kubectl describe pod extended-resource-demo-2
The output shows that the Pod cannot be scheduled, because there is no Node that has 2 dongles available:
Conditions:
Type Status
PodScheduled False
...
Events:
...
... Warning FailedScheduling pod (extended-resource-demo-2) failed to fit in any node
fit failure summary on nodes : Insufficient example.com/dongle (1)
View the Pod status:
kubectl get pod extended-resource-demo-2
The output shows that the Pod was created, but not scheduled to run on a Node. It has a status of Pending:
NAME READY STATUS RESTARTS AGE
extended-resource-demo-2 0/1 Pending 0 6m
Delete the Pods that you created for this exercise:
kubectl delete pod extended-resource-demo
kubectl delete pod extended-resource-demo-2
This page shows how to configure a Pod to use a Volume for storage.
A Container's file system lives only as long as the Container does. So when a Container terminates and restarts, filesystem changes are lost. For more consistent storage that is independent of the Container, you can use a Volume. This is especially important for stateful applications, such as key-value stores (such as Redis) and databases.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise, you create a Pod that runs one Container. This Pod has a Volume of type emptyDir that lasts for the life of the Pod, even if the Container terminates and restarts. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: redis
spec:
containers:
- name: redis
image: redis
volumeMounts:
- name: redis-storage
mountPath: /data/redis
volumes:
- name: redis-storage
emptyDir: {}
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/storage/redis.yaml
Verify that the Pod's Container is running, and then watch for changes to the Pod:
kubectl get pod redis --watch
The output looks like this:
NAME READY STATUS RESTARTS AGE
redis 1/1 Running 0 13s
In another terminal, get a shell to the running Container:
kubectl exec -it redis -- /bin/bash
In your shell, go to /data/redis, and then create a file:
root@redis:/data# cd /data/redis/
root@redis:/data/redis# echo Hello > test-file
In your shell, list the running processes:
root@redis:/data/redis# apt-get update
root@redis:/data/redis# apt-get install procps
root@redis:/data/redis# ps aux
The output is similar to this:
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
redis 1 0.1 0.1 33308 3828 ? Ssl 00:46 0:00 redis-server *:6379
root 12 0.0 0.0 20228 3020 ? Ss 00:47 0:00 /bin/bash
root 15 0.0 0.0 17500 2072 ? R+ 00:48 0:00 ps aux
In your shell, kill the Redis process:
root@redis:/data/redis# kill <pid>
where <pid> is the Redis process ID (PID).
In your original terminal, watch for changes to the Redis Pod. Eventually, you will see something like this:
NAME READY STATUS RESTARTS AGE
redis 1/1 Running 0 13s
redis 0/1 Completed 0 6m
redis 1/1 Running 1 6m
At this point, the Container has terminated and restarted. This is because the
Redis Pod has a
restartPolicy
of Always.
Get a shell into the restarted Container:
kubectl exec -it redis -- /bin/bash
In your shell, go to /data/redis, and verify that test-file is still there.
root@redis:/data/redis# cd /data/redis/
root@redis:/data/redis# ls
test-file
Delete the Pod that you created for this exercise:
kubectl delete pod redis
See Volume.
See Pod.
In addition to the local disk storage provided by emptyDir, Kubernetes
supports many different network-attached storage solutions, including PD on
GCE and EBS on EC2, which are preferred for critical data and will handle
details such as mounting and unmounting the devices on the nodes. See
Volumes for more details.
This page shows you how to configure a Pod to use a PersistentVolumeClaim for storage. Here is a summary of the process:
You, as cluster administrator, create a PersistentVolume backed by physical storage. You do not associate the volume with any Pod.
You, now taking the role of a developer / cluster user, create a PersistentVolumeClaim that is automatically bound to a suitable PersistentVolume.
You create a Pod that uses the above PersistentVolumeClaim for storage.
You need to have a Kubernetes cluster that has only one Node, and the kubectl command-line tool must be configured to communicate with your cluster. If you do not already have a single-node cluster, you can create one by using Minikube.
Familiarize yourself with the material in Persistent Volumes.
Open a shell to the single Node in your cluster. How you open a shell depends
on how you set up your cluster. For example, if you are using Minikube, you
can open a shell to your Node by entering minikube ssh.
In your shell on that Node, create a /mnt/data directory:
# This assumes that your Node uses "sudo" to run commands
# as the superuser
sudo mkdir /mnt/data
In the /mnt/data directory, create an index.html file:
# This again assumes that your Node uses "sudo" to run commands
# as the superuser
sudo sh -c "echo 'Hello from Kubernetes storage' > /mnt/data/index.html"
sudo, you can
usually make this work if you replace sudo with the name of the other tool.
Test that the index.html file exists:
cat /mnt/data/index.html
The output should be:
Hello from Kubernetes storage
You can now close the shell to your Node.
In this exercise, you create a hostPath PersistentVolume. Kubernetes supports hostPath for development and testing on a single-node cluster. A hostPath PersistentVolume uses a file or directory on the Node to emulate network-attached storage.
In a production cluster, you would not use hostPath. Instead a cluster administrator would provision a network resource like a Google Compute Engine persistent disk, an NFS share, or an Amazon Elastic Block Store volume. Cluster administrators can also use StorageClasses to set up dynamic provisioning.
Here is the configuration file for the hostPath PersistentVolume:
apiVersion: v1
kind: PersistentVolume
metadata:
name: task-pv-volume
labels:
type: local
spec:
storageClassName: manual
capacity:
storage: 10Gi
accessModes:
- ReadWriteOnce
hostPath:
path: "/mnt/data"
The configuration file specifies that the volume is at /mnt/data on the
cluster's Node. The configuration also specifies a size of 10 gibibytes and
an access mode of ReadWriteOnce, which means the volume can be mounted as
read-write by a single Node. It defines the StorageClass name
manual for the PersistentVolume, which will be used to bind
PersistentVolumeClaim requests to this PersistentVolume.
Create the PersistentVolume:
kubectl apply -f https://k8s.io/examples/pods/storage/pv-volume.yaml
View information about the PersistentVolume:
kubectl get pv task-pv-volume
The output shows that the PersistentVolume has a STATUS of Available. This
means it has not yet been bound to a PersistentVolumeClaim.
NAME CAPACITY ACCESSMODES RECLAIMPOLICY STATUS CLAIM STORAGECLASS REASON AGE
task-pv-volume 10Gi RWO Retain Available manual 4s
The next step is to create a PersistentVolumeClaim. Pods use PersistentVolumeClaims to request physical storage. In this exercise, you create a PersistentVolumeClaim that requests a volume of at least three gibibytes that can provide read-write access for at least one Node.
Here is the configuration file for the PersistentVolumeClaim:
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: task-pv-claim
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 3Gi
Create the PersistentVolumeClaim:
kubectl apply -f https://k8s.io/examples/pods/storage/pv-claim.yaml
After you create the PersistentVolumeClaim, the Kubernetes control plane looks for a PersistentVolume that satisfies the claim's requirements. If the control plane finds a suitable PersistentVolume with the same StorageClass, it binds the claim to the volume.
Look again at the PersistentVolume:
kubectl get pv task-pv-volume
Now the output shows a STATUS of Bound.
NAME CAPACITY ACCESSMODES RECLAIMPOLICY STATUS CLAIM STORAGECLASS REASON AGE
task-pv-volume 10Gi RWO Retain Bound default/task-pv-claim manual 2m
Look at the PersistentVolumeClaim:
kubectl get pvc task-pv-claim
The output shows that the PersistentVolumeClaim is bound to your PersistentVolume,
task-pv-volume.
NAME STATUS VOLUME CAPACITY ACCESSMODES STORAGECLASS AGE
task-pv-claim Bound task-pv-volume 10Gi RWO manual 30s
The next step is to create a Pod that uses your PersistentVolumeClaim as a volume.
Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: task-pv-pod
spec:
volumes:
- name: task-pv-storage
persistentVolumeClaim:
claimName: task-pv-claim
containers:
- name: task-pv-container
image: nginx
ports:
- containerPort: 80
name: "http-server"
volumeMounts:
- mountPath: "/usr/share/nginx/html"
name: task-pv-storage
Notice that the Pod's configuration file specifies a PersistentVolumeClaim, but it does not specify a PersistentVolume. From the Pod's point of view, the claim is a volume.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/storage/pv-pod.yaml
Verify that the container in the Pod is running;
kubectl get pod task-pv-pod
Get a shell to the container running in your Pod:
kubectl exec -it task-pv-pod -- /bin/bash
In your shell, verify that nginx is serving the index.html file from the
hostPath volume:
# Be sure to run these 3 commands inside the root shell that comes from
# running "kubectl exec" in the previous step
apt update
apt install curl
curl http://localhost/
The output shows the text that you wrote to the index.html file on the
hostPath volume:
Hello from Kubernetes storage
If you see that message, you have successfully configured a Pod to use storage from a PersistentVolumeClaim.
Delete the Pod, the PersistentVolumeClaim and the PersistentVolume:
kubectl delete pod task-pv-pod
kubectl delete pvc task-pv-claim
kubectl delete pv task-pv-volume
If you don't already have a shell open to the Node in your cluster, open a new shell the same way that you did earlier.
In the shell on your Node, remove the file and directory that you created:
# This assumes that your Node uses "sudo" to run commands
# as the superuser
sudo rm /mnt/data/index.html
sudo rmdir /mnt/data
You can now close the shell to your Node.
Storage configured with a group ID (GID) allows writing only by Pods using the same GID. Mismatched or missing GIDs cause permission denied errors. To reduce the need for coordination with users, an administrator can annotate a PersistentVolume with a GID. Then the GID is automatically added to any Pod that uses the PersistentVolume.
Use the pv.beta.kubernetes.io/gid annotation as follows:
apiVersion: v1
kind: PersistentVolume
metadata:
name: pv1
annotations:
pv.beta.kubernetes.io/gid: "1234"
When a Pod consumes a PersistentVolume that has a GID annotation, the annotated GID is applied to all containers in the Pod in the same way that GIDs specified in the Pod's security context are. Every GID, whether it originates from a PersistentVolume annotation or the Pod's specification, is applied to the first process run in each container.
This page shows how to use a projected Volume to mount
several existing volume sources into the same directory. Currently, secret, configMap, downwardAPI,
and serviceAccountToken volumes can be projected.
serviceAccountToken is not a volume type.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise, you create username and password Secrets from local files. You then create a Pod that runs one container, using a projected Volume to mount the Secrets into the same shared directory.
Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: test-projected-volume
spec:
containers:
- name: test-projected-volume
image: busybox
args:
- sleep
- "86400"
volumeMounts:
- name: all-in-one
mountPath: "/projected-volume"
readOnly: true
volumes:
- name: all-in-one
projected:
sources:
- secret:
name: user
- secret:
name: pass
Create the Secrets:
# Create files containing the username and password:
echo -n "admin" > ./username.txt
echo -n "1f2d1e2e67df" > ./password.txt
# Package these files into secrets:
kubectl create secret generic user --from-file=./username.txt
kubectl create secret generic pass --from-file=./password.txt
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/storage/projected.yaml
Verify that the Pod's container is running, and then watch for changes to the Pod:
kubectl get --watch pod test-projected-volume
The output looks like this:
NAME READY STATUS RESTARTS AGE
test-projected-volume 1/1 Running 0 14s
In another terminal, get a shell to the running container:
kubectl exec -it test-projected-volume -- /bin/sh
In your shell, verify that the projected-volume directory contains your projected sources:
ls /projected-volume/
Delete the Pod and the Secrets:
kubectl delete pod test-projected-volume
kubectl delete secret user pass
projected volumes.A security context defines privilege and access control settings for a Pod or Container. Security context settings include, but are not limited to:
Discretionary Access Control: Permission to access an object, like a file, is based on user ID (UID) and group ID (GID).
Security Enhanced Linux (SELinux): Objects are assigned security labels.
Running as privileged or unprivileged.
Linux Capabilities: Give a process some privileges, but not all the privileges of the root user.
AppArmor: Use program profiles to restrict the capabilities of individual programs.
Seccomp: Filter a process's system calls.
AllowPrivilegeEscalation: Controls whether a process can gain more privileges than its parent process. This bool directly controls whether the no_new_privs flag gets set on the container process. AllowPrivilegeEscalation is true always when the container is: 1) run as Privileged OR 2) has CAP_SYS_ADMIN.
readOnlyRootFilesystem: Mounts the container's root filesystem as read-only.
The above bullets are not a complete set of security context settings -- please see SecurityContext for a comprehensive list.
For more information about security mechanisms in Linux, see Overview of Linux Kernel Security Features
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
To specify security settings for a Pod, include the securityContext field
in the Pod specification. The securityContext field is a
PodSecurityContext object.
The security settings that you specify for a Pod apply to all Containers in the Pod.
Here is a configuration file for a Pod that has a securityContext and an emptyDir volume:
apiVersion: v1
kind: Pod
metadata:
name: security-context-demo
spec:
securityContext:
runAsUser: 1000
runAsGroup: 3000
fsGroup: 2000
volumes:
- name: sec-ctx-vol
emptyDir: {}
containers:
- name: sec-ctx-demo
image: busybox
command: [ "sh", "-c", "sleep 1h" ]
volumeMounts:
- name: sec-ctx-vol
mountPath: /data/demo
securityContext:
allowPrivilegeEscalation: false
In the configuration file, the runAsUser field specifies that for any Containers in
the Pod, all processes run with user ID 1000. The runAsGroup field specifies the primary group ID of 3000 for
all processes within any containers of the Pod. If this field is omitted, the primary group ID of the containers
will be root(0). Any files created will also be owned by user 1000 and group 3000 when runAsGroup is specified.
Since fsGroup field is specified, all processes of the container are also part of the supplementary group ID 2000.
The owner for volume /data/demo and any files created in that volume will be Group ID 2000.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/security/security-context.yaml
Verify that the Pod's Container is running:
kubectl get pod security-context-demo
Get a shell to the running Container:
kubectl exec -it security-context-demo -- sh
In your shell, list the running processes:
ps
The output shows that the processes are running as user 1000, which is the value of runAsUser:
PID USER TIME COMMAND
1 1000 0:00 sleep 1h
6 1000 0:00 sh
...
In your shell, navigate to /data, and list the one directory:
cd /data
ls -l
The output shows that the /data/demo directory has group ID 2000, which is
the value of fsGroup.
drwxrwsrwx 2 root 2000 4096 Jun 6 20:08 demo
In your shell, navigate to /data/demo, and create a file:
cd demo
echo hello > testfile
List the file in the /data/demo directory:
ls -l
The output shows that testfile has group ID 2000, which is the value of fsGroup.
-rw-r--r-- 1 1000 2000 6 Jun 6 20:08 testfile
Run the following command:
$ id
uid=1000 gid=3000 groups=2000
You will see that gid is 3000 which is same as runAsGroup field. If the runAsGroup was omitted the gid would
remain as 0(root) and the process will be able to interact with files that are owned by root(0) group and that have
the required group permissions for root(0) group.
Exit your shell:
exit
Kubernetes v1.20 [beta]
By default, Kubernetes recursively changes ownership and permissions for the contents of each
volume to match the fsGroup specified in a Pod's securityContext when that volume is
mounted.
For large volumes, checking and changing ownership and permissions can take a lot of time,
slowing Pod startup. You can use the fsGroupChangePolicy field inside a securityContext
to control the way that Kubernetes checks and manages ownership and permissions
for a volume.
fsGroupChangePolicy - fsGroupChangePolicy defines behavior for changing ownership and permission of the volume
before being exposed inside a Pod. This field only applies to volume types that support
fsGroup controlled ownership and permissions. This field has two possible values:
For example:
securityContext:
runAsUser: 1000
runAsGroup: 3000
fsGroup: 2000
fsGroupChangePolicy: "OnRootMismatch"
Kubernetes v1.22 [alpha]
If you deploy a Container Storage Interface (CSI)
driver which supports the VOLUME_MOUNT_GROUP NodeServiceCapability, the
process of setting file ownership and permissions based on the
fsGroup specified in the securityContext will be performed by the CSI driver
instead of Kubernetes, provided that the DelegateFSGroupToCSIDriver Kubernetes
feature gate is enabled. In this case, since Kubernetes doesn't perform any
ownership and permission change, fsGroupChangePolicy does not take effect, and
as specified by CSI, the driver is expected to mount the volume with the
provided fsGroup, resulting in a volume that is readable/writable by the
fsGroup.
Please refer to the KEP
and the description of the VolumeCapability.MountVolume.volume_mount_group
field in the CSI spec
for more information.
To specify security settings for a Container, include the securityContext field
in the Container manifest. The securityContext field is a
SecurityContext object.
Security settings that you specify for a Container apply only to
the individual Container, and they override settings made at the Pod level when
there is overlap. Container settings do not affect the Pod's Volumes.
Here is the configuration file for a Pod that has one Container. Both the Pod
and the Container have a securityContext field:
apiVersion: v1
kind: Pod
metadata:
name: security-context-demo-2
spec:
securityContext:
runAsUser: 1000
containers:
- name: sec-ctx-demo-2
image: gcr.io/google-samples/node-hello:1.0
securityContext:
runAsUser: 2000
allowPrivilegeEscalation: false
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/security/security-context-2.yaml
Verify that the Pod's Container is running:
kubectl get pod security-context-demo-2
Get a shell into the running Container:
kubectl exec -it security-context-demo-2 -- sh
In your shell, list the running processes:
ps aux
The output shows that the processes are running as user 2000. This is the value
of runAsUser specified for the Container. It overrides the value 1000 that is
specified for the Pod.
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
2000 1 0.0 0.0 4336 764 ? Ss 20:36 0:00 /bin/sh -c node server.js
2000 8 0.1 0.5 772124 22604 ? Sl 20:36 0:00 node server.js
...
Exit your shell:
exit
With Linux capabilities,
you can grant certain privileges to a process without granting all the privileges
of the root user. To add or remove Linux capabilities for a Container, include the
capabilities field in the securityContext section of the Container manifest.
First, see what happens when you don't include a capabilities field.
Here is configuration file that does not add or remove any Container capabilities:
apiVersion: v1
kind: Pod
metadata:
name: security-context-demo-3
spec:
containers:
- name: sec-ctx-3
image: gcr.io/google-samples/node-hello:1.0
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/security/security-context-3.yaml
Verify that the Pod's Container is running:
kubectl get pod security-context-demo-3
Get a shell into the running Container:
kubectl exec -it security-context-demo-3 -- sh
In your shell, list the running processes:
ps aux
The output shows the process IDs (PIDs) for the Container:
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 1 0.0 0.0 4336 796 ? Ss 18:17 0:00 /bin/sh -c node server.js
root 5 0.1 0.5 772124 22700 ? Sl 18:17 0:00 node server.js
In your shell, view the status for process 1:
cd /proc/1
cat status
The output shows the capabilities bitmap for the process:
...
CapPrm: 00000000a80425fb
CapEff: 00000000a80425fb
...
Make a note of the capabilities bitmap, and then exit your shell:
exit
Next, run a Container that is the same as the preceding container, except that it has additional capabilities set.
Here is the configuration file for a Pod that runs one Container. The configuration
adds the CAP_NET_ADMIN and CAP_SYS_TIME capabilities:
apiVersion: v1
kind: Pod
metadata:
name: security-context-demo-4
spec:
containers:
- name: sec-ctx-4
image: gcr.io/google-samples/node-hello:1.0
securityContext:
capabilities:
add: ["NET_ADMIN", "SYS_TIME"]
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/security/security-context-4.yaml
Get a shell into the running Container:
kubectl exec -it security-context-demo-4 -- sh
In your shell, view the capabilities for process 1:
cd /proc/1
cat status
The output shows capabilities bitmap for the process:
...
CapPrm: 00000000aa0435fb
CapEff: 00000000aa0435fb
...
Compare the capabilities of the two Containers:
00000000a80425fb
00000000aa0435fb
In the capability bitmap of the first container, bits 12 and 25 are clear. In the second container,
bits 12 and 25 are set. Bit 12 is CAP_NET_ADMIN, and bit 25 is CAP_SYS_TIME.
See capability.h
for definitions of the capability constants.
CAP_XXX. But when you list capabilities in your Container manifest, you must omit the CAP_ portion of the constant. For example, to add CAP_SYS_TIME, include SYS_TIME in your list of capabilities.
To set the Seccomp profile for a Container, include the seccompProfile field
in the securityContext section of your Pod or Container manifest. The
seccompProfile field is a
SeccompProfile object consisting of type and localhostProfile.
Valid options for type include RuntimeDefault, Unconfined, and
Localhost. localhostProfile must only be set set if type: Localhost. It
indicates the path of the pre-configured profile on the node, relative to the
kubelet's configured Seccomp profile location (configured with the --root-dir
flag).
Here is an example that sets the Seccomp profile to the node's container runtime default profile:
...
securityContext:
seccompProfile:
type: RuntimeDefault
Here is an example that sets the Seccomp profile to a pre-configured file at
<kubelet-root-dir>/seccomp/my-profiles/profile-allow.json:
...
securityContext:
seccompProfile:
type: Localhost
localhostProfile: my-profiles/profile-allow.json
To assign SELinux labels to a Container, include the seLinuxOptions field in
the securityContext section of your Pod or Container manifest. The
seLinuxOptions field is an
SELinuxOptions
object. Here's an example that applies an SELinux level:
...
securityContext:
seLinuxOptions:
level: "s0:c123,c456"
The security context for a Pod applies to the Pod's Containers and also to
the Pod's Volumes when applicable. Specifically fsGroup and seLinuxOptions are
applied to Volumes as follows:
fsGroup: Volumes that support ownership management are modified to be owned
and writable by the GID specified in fsGroup. See the
Ownership Management design document
for more details.
seLinuxOptions: Volumes that support SELinux labeling are relabeled to be accessible
by the label specified under seLinuxOptions. Usually you only
need to set the level section. This sets the
Multi-Category Security (MCS)
label given to all Containers in the Pod as well as the Volumes.
Delete the Pod:
kubectl delete pod security-context-demo
kubectl delete pod security-context-demo-2
kubectl delete pod security-context-demo-3
kubectl delete pod security-context-demo-4
A service account provides an identity for processes that run in a Pod.
When you (a human) access the cluster (for example, using kubectl), you are
authenticated by the apiserver as a particular User Account (currently this is
usually admin, unless your cluster administrator has customized your cluster). Processes in containers inside pods can also contact the apiserver.
When they do, they are authenticated as a particular Service Account (for example, default).
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
When you create a pod, if you do not specify a service account, it is
automatically assigned the default service account in the same namespace.
If you get the raw json or yaml for a pod you have created (for example, kubectl get pods/<podname> -o yaml),
you can see the spec.serviceAccountName field has been
automatically set.
You can access the API from inside a pod using automatically mounted service account credentials, as described in Accessing the Cluster. The API permissions of the service account depend on the authorization plugin and policy in use.
In version 1.6+, you can opt out of automounting API credentials for a service account by setting automountServiceAccountToken: false on the service account:
apiVersion: v1
kind: ServiceAccount
metadata:
name: build-robot
automountServiceAccountToken: false
...
In version 1.6+, you can also opt out of automounting API credentials for a particular pod:
apiVersion: v1
kind: Pod
metadata:
name: my-pod
spec:
serviceAccountName: build-robot
automountServiceAccountToken: false
...
The pod spec takes precedence over the service account if both specify a automountServiceAccountToken value.
Every namespace has a default service account resource called default.
You can list this and any other serviceAccount resources in the namespace with this command:
kubectl get serviceaccounts
The output is similar to this:
NAME SECRETS AGE
default 1 1d
You can create additional ServiceAccount objects like this:
kubectl apply -f - <<EOF
apiVersion: v1
kind: ServiceAccount
metadata:
name: build-robot
EOF
The name of a ServiceAccount object must be a valid DNS subdomain name.
If you get a complete dump of the service account object, like this:
kubectl get serviceaccounts/build-robot -o yaml
The output is similar to this:
apiVersion: v1
kind: ServiceAccount
metadata:
creationTimestamp: 2015-06-16T00:12:59Z
name: build-robot
namespace: default
resourceVersion: "272500"
uid: 721ab723-13bc-11e5-aec2-42010af0021e
secrets:
- name: build-robot-token-bvbk5
then you will see that a token has automatically been created and is referenced by the service account.
You may use authorization plugins to set permissions on service accounts.
To use a non-default service account, set the spec.serviceAccountName
field of a pod to the name of the service account you wish to use.
The service account has to exist at the time the pod is created, or it will be rejected.
You cannot update the service account of an already created pod.
You can clean up the service account from this example like this:
kubectl delete serviceaccount/build-robot
Suppose we have an existing service account named "build-robot" as mentioned above, and we create a new secret manually.
kubectl apply -f - <<EOF
apiVersion: v1
kind: Secret
metadata:
name: build-robot-secret
annotations:
kubernetes.io/service-account.name: build-robot
type: kubernetes.io/service-account-token
EOF
Now you can confirm that the newly built secret is populated with an API token for the "build-robot" service account.
Any tokens for non-existent service accounts will be cleaned up by the token controller.
kubectl describe secrets/build-robot-secret
The output is similar to this:
Name: build-robot-secret
Namespace: default
Labels: <none>
Annotations: kubernetes.io/service-account.name: build-robot
kubernetes.io/service-account.uid: da68f9c6-9d26-11e7-b84e-002dc52800da
Type: kubernetes.io/service-account-token
Data
====
ca.crt: 1338 bytes
namespace: 7 bytes
token: ...
token is elided here.
Create an imagePullSecret, as described in Specifying ImagePullSecrets on a Pod.
kubectl create secret docker-registry myregistrykey --docker-server=DUMMY_SERVER \
--docker-username=DUMMY_USERNAME --docker-password=DUMMY_DOCKER_PASSWORD \
--docker-email=DUMMY_DOCKER_EMAIL
Verify it has been created.
kubectl get secrets myregistrykey
The output is similar to this:
NAME TYPE DATA AGE
myregistrykey kubernetes.io/.dockerconfigjson 1 1d
Next, modify the default service account for the namespace to use this secret as an imagePullSecret.
kubectl patch serviceaccount default -p '{"imagePullSecrets": [{"name": "myregistrykey"}]}'
You can instead use kubectl edit, or manually edit the YAML manifests as shown below:
kubectl get serviceaccounts default -o yaml > ./sa.yaml
The output of the sa.yaml file is similar to this:
apiVersion: v1
kind: ServiceAccount
metadata:
creationTimestamp: 2015-08-07T22:02:39Z
name: default
namespace: default
resourceVersion: "243024"
uid: 052fb0f4-3d50-11e5-b066-42010af0d7b6
secrets:
- name: default-token-uudge
Using your editor of choice (for example vi), open the sa.yaml file, delete line with key resourceVersion, add lines with imagePullSecrets: and save.
The output of the sa.yaml file is similar to this:
apiVersion: v1
kind: ServiceAccount
metadata:
creationTimestamp: 2015-08-07T22:02:39Z
name: default
namespace: default
uid: 052fb0f4-3d50-11e5-b066-42010af0d7b6
secrets:
- name: default-token-uudge
imagePullSecrets:
- name: myregistrykey
Finally replace the serviceaccount with the new updated sa.yaml file
kubectl replace serviceaccount default -f ./sa.yaml
Now, when a new Pod is created in the current namespace and using the default ServiceAccount, the new Pod has its spec.imagePullSecrets field set automatically:
kubectl run nginx --image=nginx --restart=Never
kubectl get pod nginx -o=jsonpath='{.spec.imagePullSecrets[0].name}{"\n"}'
The output is:
myregistrykey
Kubernetes v1.20 [stable]
To enable and use token request projection, you must specify each of the following
command line arguments to kube-apiserver:
--service-account-issuer--service-account-key-file--service-account-signing-key-file--api-audiencesThe kubelet can also project a service account token into a Pod. You can specify desired properties of the token, such as the audience and the validity duration. These properties are not configurable on the default service account token. The service account token will also become invalid against the API when the Pod or the ServiceAccount is deleted.
This behavior is configured on a PodSpec using a ProjectedVolume type called ServiceAccountToken. To provide a pod with a token with an audience of "vault" and a validity duration of two hours, you would configure the following in your PodSpec:
apiVersion: v1
kind: Pod
metadata:
name: nginx
spec:
containers:
- image: nginx
name: nginx
volumeMounts:
- mountPath: /var/run/secrets/tokens
name: vault-token
serviceAccountName: build-robot
volumes:
- name: vault-token
projected:
sources:
- serviceAccountToken:
path: vault-token
expirationSeconds: 7200
audience: vault
Create the Pod:
kubectl create -f https://k8s.io/examples/pods/pod-projected-svc-token.yaml
The kubelet will request and store the token on behalf of the pod, make the token available to the pod at a configurable file path, and refresh the token as it approaches expiration. The kubelet proactively rotates the token if it is older than 80% of its total TTL, or if the token is older than 24 hours.
The application is responsible for reloading the token when it rotates. Periodic reloading (e.g. once every 5 minutes) is sufficient for most use cases.
Kubernetes v1.21 [stable]
The Service Account Issuer Discovery feature is enabled when the Service Account Token Projection feature is enabled, as described above.
The issuer URL must comply with the
OIDC Discovery Spec. In
practice, this means it must use the https scheme, and should serve an OpenID
provider configuration at {service-account-issuer}/.well-known/openid-configuration.
If the URL does not comply, the ServiceAccountIssuerDiscovery endpoints will
not be registered, even if the feature is enabled.
The Service Account Issuer Discovery feature enables federation of Kubernetes service account tokens issued by a cluster (the identity provider) with external systems (relying parties).
When enabled, the Kubernetes API server provides an OpenID Provider
Configuration document at /.well-known/openid-configuration and the associated
JSON Web Key Set (JWKS) at /openid/v1/jwks. The OpenID Provider Configuration
is sometimes referred to as the discovery document.
Clusters include a default RBAC ClusterRole called
system:service-account-issuer-discovery. A default RBAC ClusterRoleBinding
assigns this role to the system:serviceaccounts group, which all service
accounts implicitly belong to. This allows pods running on the cluster to access
the service account discovery document via their mounted service account token.
Administrators may, additionally, choose to bind the role to
system:authenticated or system:unauthenticated depending on their security
requirements and which external systems they intend to federate with.
/.well-known/openid-configuration and
/openid/v1/jwks are designed to be OIDC compatible, but not strictly OIDC
compliant. Those documents contain only the parameters necessary to perform
validation of Kubernetes service account tokens.
The JWKS response contains public keys that a relying party can use to validate
the Kubernetes service account tokens. Relying parties first query for the
OpenID Provider Configuration, and use the jwks_uri field in the response to
find the JWKS.
In many cases, Kubernetes API servers are not available on the public internet,
but public endpoints that serve cached responses from the API server can be made
available by users or service providers. In these cases, it is possible to
override the jwks_uri in the OpenID Provider Configuration so that it points
to the public endpoint, rather than the API server's address, by passing the
--service-account-jwks-uri flag to the API server. Like the issuer URL, the
JWKS URI is required to use the https scheme.
See also:
This page shows how to create a Pod that uses a Secret to pull an image from a private Docker registry or repository.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
To do this exercise, you need a Docker ID and password.
On your laptop, you must authenticate with a registry in order to pull a private image:
docker login
When prompted, enter your Docker username and password.
The login process creates or updates a config.json file that holds an authorization token.
View the config.json file:
cat ~/.docker/config.json
The output contains a section similar to this:
{
"auths": {
"https://index.docker.io/v1/": {
"auth": "c3R...zE2"
}
}
}
auth entry but a credsStore entry with the name of the store as value.
A Kubernetes cluster uses the Secret of kubernetes.io/dockerconfigjson type to authenticate with
a container registry to pull a private image.
If you already ran docker login, you can copy that credential into Kubernetes:
kubectl create secret generic regcred \
--from-file=.dockerconfigjson=<path/to/.docker/config.json> \
--type=kubernetes.io/dockerconfigjson
If you need more control (for example, to set a namespace or a label on the new secret) then you can customise the Secret before storing it. Be sure to:
.dockerconfigjsondata[".dockerconfigjson"]type to kubernetes.io/dockerconfigjsonExample:
apiVersion: v1
kind: Secret
metadata:
name: myregistrykey
namespace: awesomeapps
data:
.dockerconfigjson: UmVhbGx5IHJlYWxseSByZWVlZWVlZWVlZWFhYWFhYWFhYWFhYWFhYWFhYWFhYWFhYWFhYWxsbGxsbGxsbGxsbGxsbGxsbGxsbGxsbGxsbGxsbGx5eXl5eXl5eXl5eXl5eXl5eXl5eSBsbGxsbGxsbGxsbGxsbG9vb29vb29vb29vb29vb29vb29vb29vb29vb25ubm5ubm5ubm5ubm5ubm5ubm5ubm5ubmdnZ2dnZ2dnZ2dnZ2dnZ2dnZ2cgYXV0aCBrZXlzCg==
type: kubernetes.io/dockerconfigjson
If you get the error message error: no objects passed to create, it may mean the base64 encoded string is invalid.
If you get an error message like Secret "myregistrykey" is invalid: data[.dockerconfigjson]: invalid value ..., it means
the base64 encoded string in the data was successfully decoded, but could not be parsed as a .docker/config.json file.
Create this Secret, naming it regcred:
kubectl create secret docker-registry regcred --docker-server=<your-registry-server> --docker-username=<your-name> --docker-password=<your-pword> --docker-email=<your-email>
where:
<your-registry-server> is your Private Docker Registry FQDN.
Use https://index.docker.io/v1/ for DockerHub.<your-name> is your Docker username.<your-pword> is your Docker password.<your-email> is your Docker email.You have successfully set your Docker credentials in the cluster as a Secret called regcred.
kubectl is running.
regcredTo understand the contents of the regcred Secret you created, start by viewing the Secret in YAML format:
kubectl get secret regcred --output=yaml
The output is similar to this:
apiVersion: v1
kind: Secret
metadata:
...
name: regcred
...
data:
.dockerconfigjson: eyJodHRwczovL2luZGV4L ... J0QUl6RTIifX0=
type: kubernetes.io/dockerconfigjson
The value of the .dockerconfigjson field is a base64 representation of your Docker credentials.
To understand what is in the .dockerconfigjson field, convert the secret data to a
readable format:
kubectl get secret regcred --output="jsonpath={.data.\.dockerconfigjson}" | base64 --decode
The output is similar to this:
{"auths":{"your.private.registry.example.com":{"username":"janedoe","password":"xxxxxxxxxxx","email":"jdoe@example.com","auth":"c3R...zE2"}}}
To understand what is in the auth field, convert the base64-encoded data to a readable format:
echo "c3R...zE2" | base64 --decode
The output, username and password concatenated with a :, is similar to this:
janedoe:xxxxxxxxxxx
Notice that the Secret data contains the authorization token similar to your local ~/.docker/config.json file.
You have successfully set your Docker credentials as a Secret called regcred in the cluster.
Here is a configuration file for a Pod that needs access to your Docker credentials in regcred:
apiVersion: v1
kind: Pod
metadata:
name: private-reg
spec:
containers:
- name: private-reg-container
image: <your-private-image>
imagePullSecrets:
- name: regcred
Download the above file:
wget -O my-private-reg-pod.yaml https://k8s.io/examples/pods/private-reg-pod.yaml
In file my-private-reg-pod.yaml, replace <your-private-image> with the path to an image in a private registry such as:
your.private.registry.example.com/janedoe/jdoe-private:v1
To pull the image from the private registry, Kubernetes needs credentials.
The imagePullSecrets field in the configuration file specifies that
Kubernetes should get the credentials from a Secret named regcred.
Create a Pod that uses your Secret, and verify that the Pod is running:
kubectl apply -f my-private-reg-pod.yaml
kubectl get pod private-reg
imagePullSecrets field of PodSpec.This page shows how to configure liveness, readiness and startup probes for containers.
The kubelet uses liveness probes to know when to restart a container. For example, liveness probes could catch a deadlock, where an application is running, but unable to make progress. Restarting a container in such a state can help to make the application more available despite bugs.
The kubelet uses readiness probes to know when a container is ready to start accepting traffic. A Pod is considered ready when all of its containers are ready. One use of this signal is to control which Pods are used as backends for Services. When a Pod is not ready, it is removed from Service load balancers.
The kubelet uses startup probes to know when a container application has started. If such a probe is configured, it disables liveness and readiness checks until it succeeds, making sure those probes don't interfere with the application startup. This can be used to adopt liveness checks on slow starting containers, avoiding them getting killed by the kubelet before they are up and running.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Many applications running for long periods of time eventually transition to broken states, and cannot recover except by being restarted. Kubernetes provides liveness probes to detect and remedy such situations.
In this exercise, you create a Pod that runs a container based on the
k8s.gcr.io/busybox image. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
labels:
test: liveness
name: liveness-exec
spec:
containers:
- name: liveness
image: k8s.gcr.io/busybox
args:
- /bin/sh
- -c
- touch /tmp/healthy; sleep 30; rm -rf /tmp/healthy; sleep 600
livenessProbe:
exec:
command:
- cat
- /tmp/healthy
initialDelaySeconds: 5
periodSeconds: 5
In the configuration file, you can see that the Pod has a single Container.
The periodSeconds field specifies that the kubelet should perform a liveness
probe every 5 seconds. The initialDelaySeconds field tells the kubelet that it
should wait 5 seconds before performing the first probe. To perform a probe, the
kubelet executes the command cat /tmp/healthy in the target container. If the
command succeeds, it returns 0, and the kubelet considers the container to be alive and
healthy. If the command returns a non-zero value, the kubelet kills the container
and restarts it.
When the container starts, it executes this command:
/bin/sh -c "touch /tmp/healthy; sleep 30; rm -rf /tmp/healthy; sleep 600"
For the first 30 seconds of the container's life, there is a /tmp/healthy file.
So during the first 30 seconds, the command cat /tmp/healthy returns a success
code. After 30 seconds, cat /tmp/healthy returns a failure code.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/probe/exec-liveness.yaml
Within 30 seconds, view the Pod events:
kubectl describe pod liveness-exec
The output indicates that no liveness probes have failed yet:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
24s 24s 1 {default-scheduler } Normal Scheduled Successfully assigned liveness-exec to worker0
23s 23s 1 {kubelet worker0} spec.containers{liveness} Normal Pulling pulling image "k8s.gcr.io/busybox"
23s 23s 1 {kubelet worker0} spec.containers{liveness} Normal Pulled Successfully pulled image "k8s.gcr.io/busybox"
23s 23s 1 {kubelet worker0} spec.containers{liveness} Normal Created Created container with docker id 86849c15382e; Security:[seccomp=unconfined]
23s 23s 1 {kubelet worker0} spec.containers{liveness} Normal Started Started container with docker id 86849c15382e
After 35 seconds, view the Pod events again:
kubectl describe pod liveness-exec
At the bottom of the output, there are messages indicating that the liveness probes have failed, and the containers have been killed and recreated.
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
37s 37s 1 {default-scheduler } Normal Scheduled Successfully assigned liveness-exec to worker0
36s 36s 1 {kubelet worker0} spec.containers{liveness} Normal Pulling pulling image "k8s.gcr.io/busybox"
36s 36s 1 {kubelet worker0} spec.containers{liveness} Normal Pulled Successfully pulled image "k8s.gcr.io/busybox"
36s 36s 1 {kubelet worker0} spec.containers{liveness} Normal Created Created container with docker id 86849c15382e; Security:[seccomp=unconfined]
36s 36s 1 {kubelet worker0} spec.containers{liveness} Normal Started Started container with docker id 86849c15382e
2s 2s 1 {kubelet worker0} spec.containers{liveness} Warning Unhealthy Liveness probe failed: cat: can't open '/tmp/healthy': No such file or directory
Wait another 30 seconds, and verify that the container has been restarted:
kubectl get pod liveness-exec
The output shows that RESTARTS has been incremented:
NAME READY STATUS RESTARTS AGE
liveness-exec 1/1 Running 1 1m
Another kind of liveness probe uses an HTTP GET request. Here is the configuration
file for a Pod that runs a container based on the k8s.gcr.io/liveness
image.
apiVersion: v1
kind: Pod
metadata:
labels:
test: liveness
name: liveness-http
spec:
containers:
- name: liveness
image: k8s.gcr.io/liveness
args:
- /server
livenessProbe:
httpGet:
path: /healthz
port: 8080
httpHeaders:
- name: Custom-Header
value: Awesome
initialDelaySeconds: 3
periodSeconds: 3
In the configuration file, you can see that the Pod has a single container.
The periodSeconds field specifies that the kubelet should perform a liveness
probe every 3 seconds. The initialDelaySeconds field tells the kubelet that it
should wait 3 seconds before performing the first probe. To perform a probe, the
kubelet sends an HTTP GET request to the server that is running in the container
and listening on port 8080. If the handler for the server's /healthz path
returns a success code, the kubelet considers the container to be alive and
healthy. If the handler returns a failure code, the kubelet kills the container
and restarts it.
Any code greater than or equal to 200 and less than 400 indicates success. Any other code indicates failure.
You can see the source code for the server in server.go.
For the first 10 seconds that the container is alive, the /healthz handler
returns a status of 200. After that, the handler returns a status of 500.
http.HandleFunc("/healthz", func(w http.ResponseWriter, r *http.Request) {
duration := time.Now().Sub(started)
if duration.Seconds() > 10 {
w.WriteHeader(500)
w.Write([]byte(fmt.Sprintf("error: %v", duration.Seconds())))
} else {
w.WriteHeader(200)
w.Write([]byte("ok"))
}
})
The kubelet starts performing health checks 3 seconds after the container starts. So the first couple of health checks will succeed. But after 10 seconds, the health checks will fail, and the kubelet will kill and restart the container.
To try the HTTP liveness check, create a Pod:
kubectl apply -f https://k8s.io/examples/pods/probe/http-liveness.yaml
After 10 seconds, view Pod events to verify that liveness probes have failed and the container has been restarted:
kubectl describe pod liveness-http
In releases prior to v1.13 (including v1.13), if the environment variable
http_proxy (or HTTP_PROXY) is set on the node where a Pod is running,
the HTTP liveness probe uses that proxy.
In releases after v1.13, local HTTP proxy environment variable settings do not
affect the HTTP liveness probe.
A third type of liveness probe uses a TCP socket. With this configuration, the kubelet will attempt to open a socket to your container on the specified port. If it can establish a connection, the container is considered healthy, if it can't it is considered a failure.
apiVersion: v1
kind: Pod
metadata:
name: goproxy
labels:
app: goproxy
spec:
containers:
- name: goproxy
image: k8s.gcr.io/goproxy:0.1
ports:
- containerPort: 8080
readinessProbe:
tcpSocket:
port: 8080
initialDelaySeconds: 5
periodSeconds: 10
livenessProbe:
tcpSocket:
port: 8080
initialDelaySeconds: 15
periodSeconds: 20
As you can see, configuration for a TCP check is quite similar to an HTTP check.
This example uses both readiness and liveness probes. The kubelet will send the
first readiness probe 5 seconds after the container starts. This will attempt to
connect to the goproxy container on port 8080. If the probe succeeds, the Pod
will be marked as ready. The kubelet will continue to run this check every 10
seconds.
In addition to the readiness probe, this configuration includes a liveness probe.
The kubelet will run the first liveness probe 15 seconds after the container
starts. Similar to the readiness probe, this will attempt to connect to the
goproxy container on port 8080. If the liveness probe fails, the container
will be restarted.
To try the TCP liveness check, create a Pod:
kubectl apply -f https://k8s.io/examples/pods/probe/tcp-liveness-readiness.yaml
After 15 seconds, view Pod events to verify that liveness probes:
kubectl describe pod goproxy
You can use a named ContainerPort for HTTP or TCP liveness checks:
ports:
- name: liveness-port
containerPort: 8080
hostPort: 8080
livenessProbe:
httpGet:
path: /healthz
port: liveness-port
Sometimes, you have to deal with legacy applications that might require
an additional startup time on their first initialization.
In such cases, it can be tricky to set up liveness probe parameters without
compromising the fast response to deadlocks that motivated such a probe.
The trick is to set up a startup probe with the same command, HTTP or TCP
check, with a failureThreshold * periodSeconds long enough to cover the
worse case startup time.
So, the previous example would become:
ports:
- name: liveness-port
containerPort: 8080
hostPort: 8080
livenessProbe:
httpGet:
path: /healthz
port: liveness-port
failureThreshold: 1
periodSeconds: 10
startupProbe:
httpGet:
path: /healthz
port: liveness-port
failureThreshold: 30
periodSeconds: 10
Thanks to the startup probe, the application will have a maximum of 5 minutes
(30 * 10 = 300s) to finish its startup.
Once the startup probe has succeeded once, the liveness probe takes over to
provide a fast response to container deadlocks.
If the startup probe never succeeds, the container is killed after 300s and
subject to the pod's restartPolicy.
Sometimes, applications are temporarily unable to serve traffic. For example, an application might need to load large data or configuration files during startup, or depend on external services after startup. In such cases, you don't want to kill the application, but you don't want to send it requests either. Kubernetes provides readiness probes to detect and mitigate these situations. A pod with containers reporting that they are not ready does not receive traffic through Kubernetes Services.
Readiness probes are configured similarly to liveness probes. The only difference
is that you use the readinessProbe field instead of the livenessProbe field.
readinessProbe:
exec:
command:
- cat
- /tmp/healthy
initialDelaySeconds: 5
periodSeconds: 5
Configuration for HTTP and TCP readiness probes also remains identical to liveness probes.
Readiness and liveness probes can be used in parallel for the same container. Using both can ensure that traffic does not reach a container that is not ready for it, and that containers are restarted when they fail.
Probes have a number of fields that you can use to more precisely control the behavior of liveness and readiness checks:
initialDelaySeconds: Number of seconds after the container has started
before liveness or readiness probes are initiated. Defaults to 0 seconds. Minimum value is 0.periodSeconds: How often (in seconds) to perform the probe. Default to 10
seconds. Minimum value is 1.timeoutSeconds: Number of seconds after which the probe times out. Defaults
to 1 second. Minimum value is 1.successThreshold: Minimum consecutive successes for the probe to be
considered successful after having failed. Defaults to 1. Must be 1 for liveness
and startup Probes. Minimum value is 1.failureThreshold: When a probe fails, Kubernetes will
try failureThreshold times before giving up. Giving up in case of liveness probe means restarting the container. In case of readiness probe the Pod will be marked Unready.
Defaults to 3. Minimum value is 1.Before Kubernetes 1.20, the field timeoutSeconds was not respected for exec probes:
probes continued running indefinitely, even past their configured deadline,
until a result was returned.
This defect was corrected in Kubernetes v1.20. You may have been relying on the previous behavior,
even without realizing it, as the default timeout is 1 second.
As a cluster administrator, you can disable the feature gate ExecProbeTimeout (set it to false)
on each kubelet to restore the behavior from older versions, then remove that override
once all the exec probes in the cluster have a timeoutSeconds value set.
If you have pods that are impacted from the default 1 second timeout,
you should update their probe timeout so that you're ready for the
eventual removal of that feature gate.
With the fix of the defect, for exec probes, on Kubernetes 1.20+ with the dockershim container runtime,
the process inside the container may keep running even after probe returned failure because of the timeout.
HTTP probes
have additional fields that can be set on httpGet:
host: Host name to connect to, defaults to the pod IP. You probably want to
set "Host" in httpHeaders instead.scheme: Scheme to use for connecting to the host (HTTP or HTTPS). Defaults to HTTP.path: Path to access on the HTTP server. Defaults to /.httpHeaders: Custom headers to set in the request. HTTP allows repeated headers.port: Name or number of the port to access on the container. Number must be
in the range 1 to 65535.For an HTTP probe, the kubelet sends an HTTP request to the specified path and
port to perform the check. The kubelet sends the probe to the pod's IP address,
unless the address is overridden by the optional host field in httpGet. If
scheme field is set to HTTPS, the kubelet sends an HTTPS request skipping the
certificate verification. In most scenarios, you do not want to set the host field.
Here's one scenario where you would set it. Suppose the container listens on 127.0.0.1
and the Pod's hostNetwork field is true. Then host, under httpGet, should be set
to 127.0.0.1. If your pod relies on virtual hosts, which is probably the more common
case, you should not use host, but rather set the Host header in httpHeaders.
For an HTTP probe, the kubelet sends two request headers in addition to the mandatory Host header:
User-Agent, and Accept. The default values for these headers are kube-probe/1.22
(where 1.22 is the version of the kubelet ), and */* respectively.
You can override the default headers by defining .httpHeaders for the probe; for example
livenessProbe:
httpGet:
httpHeaders:
- name: Accept
value: application/json
startupProbe:
httpGet:
httpHeaders:
- name: User-Agent
value: MyUserAgent
You can also remove these two headers by defining them with an empty value.
livenessProbe:
httpGet:
httpHeaders:
- name: Accept
value: ""
startupProbe:
httpGet:
httpHeaders:
- name: User-Agent
value: ""
For a TCP probe, the kubelet makes the probe connection at the node, not in the pod, which
means that you can not use a service name in the host parameter since the kubelet is unable
to resolve it.
terminationGracePeriodSecondsKubernetes v1.22 [beta]
Prior to release 1.21, the pod-level terminationGracePeriodSeconds was used
for terminating a container that failed its liveness or startup probe. This
coupling was unintended and may have resulted in failed containers taking an
unusually long time to restart when a pod-level terminationGracePeriodSeconds
was set.
In 1.21 and beyond, when the feature gate ProbeTerminationGracePeriod is
enabled, users can specify a probe-level terminationGracePeriodSeconds as
part of the probe specification. When the feature gate is enabled, and both a
pod- and probe-level terminationGracePeriodSeconds are set, the kubelet will
use the probe-level value.
As of Kubernetes 1.22, the ProbeTerminationGracePeriod feature gate is only
available on the API Server. The kubelet always honors the probe-level
terminationGracePeriodSeconds field if it is present on a Pod.
If you have existing Pods where the terminationGracePeriodSeconds field is set and
you no longer wish to use per-probe termination grace periods, you must delete
those existing Pods.
When you (or the control plane, or some other component) create replacement
Pods, and the feature gate ProbeTerminationGracePeriod is disabled, then the
API server ignores the Pod-level terminationGracePeriodSeconds field, even if
a Pod or pod template specifies it.
For example,
spec:
terminationGracePeriodSeconds: 3600 # pod-level
containers:
- name: test
image: ...
ports:
- name: liveness-port
containerPort: 8080
hostPort: 8080
livenessProbe:
httpGet:
path: /healthz
port: liveness-port
failureThreshold: 1
periodSeconds: 60
# Override pod-level terminationGracePeriodSeconds #
terminationGracePeriodSeconds: 60
Probe-level terminationGracePeriodSeconds cannot be set for readiness probes.
It will be rejected by the API server.
You can also read the API references for:
This page shows how to assign a Kubernetes Pod to a particular node in a Kubernetes cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
List the nodes in your cluster, along with their labels:
kubectl get nodes --show-labels
The output is similar to this:
NAME STATUS ROLES AGE VERSION LABELS
worker0 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker0
worker1 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker1
worker2 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker2
Chose one of your nodes, and add a label to it:
kubectl label nodes <your-node-name> disktype=ssd
where <your-node-name> is the name of your chosen node.
Verify that your chosen node has a disktype=ssd label:
kubectl get nodes --show-labels
The output is similar to this:
NAME STATUS ROLES AGE VERSION LABELS
worker0 Ready <none> 1d v1.13.0 ...,disktype=ssd,kubernetes.io/hostname=worker0
worker1 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker1
worker2 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker2
In the preceding output, you can see that the worker0 node has a
disktype=ssd label.
This pod configuration file describes a pod that has a node selector,
disktype: ssd. This means that the pod will get scheduled on a node that has
a disktype=ssd label.
apiVersion: v1
kind: Pod
metadata:
name: nginx
labels:
env: test
spec:
containers:
- name: nginx
image: nginx
imagePullPolicy: IfNotPresent
nodeSelector:
disktype: ssd
Use the configuration file to create a pod that will get scheduled on your chosen node:
kubectl apply -f https://k8s.io/examples/pods/pod-nginx.yaml
Verify that the pod is running on your chosen node:
kubectl get pods --output=wide
The output is similar to this:
NAME READY STATUS RESTARTS AGE IP NODE
nginx 1/1 Running 0 13s 10.200.0.4 worker0
You can also schedule a pod to one specific node via setting nodeName.
apiVersion: v1
kind: Pod
metadata:
name: nginx
spec:
nodeName: foo-node # schedule pod to specific node
containers:
- name: nginx
image: nginx
imagePullPolicy: IfNotPresent
Use the configuration file to create a pod that will get scheduled on foo-node only.
This page shows how to assign a Kubernetes Pod to a particular node using Node Affinity in a Kubernetes cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.10. To check the version, enterkubectl version.
List the nodes in your cluster, along with their labels:
kubectl get nodes --show-labels
The output is similar to this:
NAME STATUS ROLES AGE VERSION LABELS
worker0 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker0
worker1 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker1
worker2 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker2
Chose one of your nodes, and add a label to it:
kubectl label nodes <your-node-name> disktype=ssd
where <your-node-name> is the name of your chosen node.
Verify that your chosen node has a disktype=ssd label:
kubectl get nodes --show-labels
The output is similar to this:
NAME STATUS ROLES AGE VERSION LABELS
worker0 Ready <none> 1d v1.13.0 ...,disktype=ssd,kubernetes.io/hostname=worker0
worker1 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker1
worker2 Ready <none> 1d v1.13.0 ...,kubernetes.io/hostname=worker2
In the preceding output, you can see that the worker0 node has a
disktype=ssd label.
This manifest describes a Pod that has a requiredDuringSchedulingIgnoredDuringExecution node affinity,disktype: ssd.
This means that the pod will get scheduled only on a node that has a disktype=ssd label.
apiVersion: v1
kind: Pod
metadata:
name: nginx
spec:
affinity:
nodeAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- key: disktype
operator: In
values:
- ssd
containers:
- name: nginx
image: nginx
imagePullPolicy: IfNotPresent
Apply the manifest to create a Pod that is scheduled onto your chosen node:
kubectl apply -f https://k8s.io/examples/pods/pod-nginx-required-affinity.yaml
Verify that the pod is running on your chosen node:
kubectl get pods --output=wide
The output is similar to this:
NAME READY STATUS RESTARTS AGE IP NODE
nginx 1/1 Running 0 13s 10.200.0.4 worker0
This manifest describes a Pod that has a preferredDuringSchedulingIgnoredDuringExecution node affinity,disktype: ssd.
This means that the pod will prefer a node that has a disktype=ssd label.
apiVersion: v1
kind: Pod
metadata:
name: nginx
spec:
affinity:
nodeAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 1
preference:
matchExpressions:
- key: disktype
operator: In
values:
- ssd
containers:
- name: nginx
image: nginx
imagePullPolicy: IfNotPresent
Apply the manifest to create a Pod that is scheduled onto your chosen node:
kubectl apply -f https://k8s.io/examples/pods/pod-nginx-preferred-affinity.yaml
Verify that the pod is running on your chosen node:
kubectl get pods --output=wide
The output is similar to this:
NAME READY STATUS RESTARTS AGE IP NODE
nginx 1/1 Running 0 13s 10.200.0.4 worker0
Learn more about Node Affinity.
This page shows how to use an Init Container to initialize a Pod before an application Container runs.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise you create a Pod that has one application Container and one Init Container. The init container runs to completion before the application container starts.
Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: init-demo
spec:
containers:
- name: nginx
image: nginx
ports:
- containerPort: 80
volumeMounts:
- name: workdir
mountPath: /usr/share/nginx/html
# These containers are run during pod initialization
initContainers:
- name: install
image: busybox
command:
- wget
- "-O"
- "/work-dir/index.html"
- http://info.cern.ch
volumeMounts:
- name: workdir
mountPath: "/work-dir"
dnsPolicy: Default
volumes:
- name: workdir
emptyDir: {}
In the configuration file, you can see that the Pod has a Volume that the init container and the application container share.
The init container mounts the
shared Volume at /work-dir, and the application container mounts the shared
Volume at /usr/share/nginx/html. The init container runs the following command
and then terminates:
wget -O /work-dir/index.html http://info.cern.ch
Notice that the init container writes the index.html file in the root directory
of the nginx server.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/init-containers.yaml
Verify that the nginx container is running:
kubectl get pod init-demo
The output shows that the nginx container is running:
NAME READY STATUS RESTARTS AGE
init-demo 1/1 Running 0 1m
Get a shell into the nginx container running in the init-demo Pod:
kubectl exec -it init-demo -- /bin/bash
In your shell, send a GET request to the nginx server:
root@nginx:~# apt-get update
root@nginx:~# apt-get install curl
root@nginx:~# curl localhost
The output shows that nginx is serving the web page that was written by the init container:
<html><head></head><body><header>
<title>http://info.cern.ch</title>
</header>
<h1>http://info.cern.ch - home of the first website</h1>
...
<li><a href="http://info.cern.ch/hypertext/WWW/TheProject.html">Browse the first website</a></li>
...
This page shows how to attach handlers to Container lifecycle events. Kubernetes supports the postStart and preStop events. Kubernetes sends the postStart event immediately after a Container is started, and it sends the preStop event immediately before the Container is terminated. A Container may specify one handler per event.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise, you create a Pod that has one Container. The Container has handlers for the postStart and preStop events.
Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: lifecycle-demo
spec:
containers:
- name: lifecycle-demo-container
image: nginx
lifecycle:
postStart:
exec:
command: ["/bin/sh", "-c", "echo Hello from the postStart handler > /usr/share/message"]
preStop:
exec:
command: ["/bin/sh","-c","nginx -s quit; while killall -0 nginx; do sleep 1; done"]
In the configuration file, you can see that the postStart command writes a message
file to the Container's /usr/share directory. The preStop command shuts down
nginx gracefully. This is helpful if the Container is being terminated because of a failure.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/lifecycle-events.yaml
Verify that the Container in the Pod is running:
kubectl get pod lifecycle-demo
Get a shell into the Container running in your Pod:
kubectl exec -it lifecycle-demo -- /bin/bash
In your shell, verify that the postStart handler created the message file:
root@lifecycle-demo:/# cat /usr/share/message
The output shows the text written by the postStart handler:
Hello from the postStart handler
Kubernetes sends the postStart event immediately after the Container is created. There is no guarantee, however, that the postStart handler is called before the Container's entrypoint is called. The postStart handler runs asynchronously relative to the Container's code, but Kubernetes' management of the container blocks until the postStart handler completes. The Container's status is not set to RUNNING until the postStart handler completes.
Kubernetes sends the preStop event immediately before the Container is terminated. Kubernetes' management of the Container blocks until the preStop handler completes, unless the Pod's grace period expires. For more details, see Pod Lifecycle.
ConfigMaps allow you to decouple configuration artifacts from image content to keep containerized applications portable. This page provides a series of usage examples demonstrating how to create ConfigMaps and configure Pods using data stored in ConfigMaps.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
You can use either kubectl create configmap or a ConfigMap generator in kustomization.yaml to create a ConfigMap. Note that kubectl starts to support kustomization.yaml since 1.14.
Use the kubectl create configmap command to create ConfigMaps from directories, files, or literal values:
kubectl create configmap <map-name> <data-source>
where <map-name> is the name you want to assign to the ConfigMap and <data-source> is the directory, file, or literal value to draw the data from. The name of a ConfigMap object must be a valid DNS subdomain name.
When you are creating a ConfigMap based on a file, the key in the <data-source> defaults to the basename of the file, and the value defaults to the file content.
You can use kubectl describe or
kubectl get to retrieve information
about a ConfigMap.
You can use kubectl create configmap to create a ConfigMap from multiple files in the same directory. When you are creating a ConfigMap based on a directory, kubectl identifies files whose basename is a valid key in the directory and packages each of those files into the new ConfigMap. Any directory entries except regular files are ignored (e.g. subdirectories, symlinks, devices, pipes, etc).
For example:
# Create the local directory
mkdir -p configure-pod-container/configmap/
# Download the sample files into `configure-pod-container/configmap/` directory
wget https://kubernetes.io/examples/configmap/game.properties -O configure-pod-container/configmap/game.properties
wget https://kubernetes.io/examples/configmap/ui.properties -O configure-pod-container/configmap/ui.properties
# Create the configmap
kubectl create configmap game-config --from-file=configure-pod-container/configmap/
The above command packages each file, in this case, game.properties and ui.properties in the configure-pod-container/configmap/ directory into the game-config ConfigMap. You can display details of the ConfigMap using the following command:
kubectl describe configmaps game-config
The output is similar to this:
Name: game-config
Namespace: default
Labels: <none>
Annotations: <none>
Data
====
game.properties:
----
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
ui.properties:
----
color.good=purple
color.bad=yellow
allow.textmode=true
how.nice.to.look=fairlyNice
The game.properties and ui.properties files in the configure-pod-container/configmap/ directory are represented in the data section of the ConfigMap.
kubectl get configmaps game-config -o yaml
The output is similar to this:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2016-02-18T18:52:05Z
name: game-config
namespace: default
resourceVersion: "516"
uid: b4952dc3-d670-11e5-8cd0-68f728db1985
data:
game.properties: |
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
ui.properties: |
color.good=purple
color.bad=yellow
allow.textmode=true
how.nice.to.look=fairlyNice
You can use kubectl create configmap to create a ConfigMap from an individual file, or from multiple files.
For example,
kubectl create configmap game-config-2 --from-file=configure-pod-container/configmap/game.properties
would produce the following ConfigMap:
kubectl describe configmaps game-config-2
where the output is similar to this:
Name: game-config-2
Namespace: default
Labels: <none>
Annotations: <none>
Data
====
game.properties:
----
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
You can pass in the --from-file argument multiple times to create a ConfigMap from multiple data sources.
kubectl create configmap game-config-2 --from-file=configure-pod-container/configmap/game.properties --from-file=configure-pod-container/configmap/ui.properties
You can display details of the game-config-2 ConfigMap using the following command:
kubectl describe configmaps game-config-2
The output is similar to this:
Name: game-config-2
Namespace: default
Labels: <none>
Annotations: <none>
Data
====
game.properties:
----
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
ui.properties:
----
color.good=purple
color.bad=yellow
allow.textmode=true
how.nice.to.look=fairlyNice
When kubectl creates a ConfigMap from inputs that are not ASCII or UTF-8, the tool puts these into the binaryData field of the ConfigMap, and not in data. Both text and binary data sources can be combined in one ConfigMap.
If you want to view the binaryData keys (and their values) in a ConfigMap, you can run kubectl get configmap -o jsonpath='{.binaryData}' <name>.
Use the option --from-env-file to create a ConfigMap from an env-file, for example:
# Env-files contain a list of environment variables.
# These syntax rules apply:
# Each line in an env file has to be in VAR=VAL format.
# Lines beginning with # (i.e. comments) are ignored.
# Blank lines are ignored.
# There is no special handling of quotation marks (i.e. they will be part of the ConfigMap value)).
# Download the sample files into `configure-pod-container/configmap/` directory
wget https://kubernetes.io/examples/configmap/game-env-file.properties -O configure-pod-container/configmap/game-env-file.properties
# The env-file `game-env-file.properties` looks like below
cat configure-pod-container/configmap/game-env-file.properties
enemies=aliens
lives=3
allowed="true"
# This comment and the empty line above it are ignored
kubectl create configmap game-config-env-file \
--from-env-file=configure-pod-container/configmap/game-env-file.properties
would produce the following ConfigMap:
kubectl get configmap game-config-env-file -o yaml
where the output is similar to this:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2017-12-27T18:36:28Z
name: game-config-env-file
namespace: default
resourceVersion: "809965"
uid: d9d1ca5b-eb34-11e7-887b-42010a8002b8
data:
allowed: '"true"'
enemies: aliens
lives: "3"
--from-env-file multiple times to create a ConfigMap from multiple data sources, only the last env-file is used.
The behavior of passing --from-env-file multiple times is demonstrated by:
# Download the sample files into `configure-pod-container/configmap/` directory
wget https://kubernetes.io/examples/configmap/ui-env-file.properties -O configure-pod-container/configmap/ui-env-file.properties
# Create the configmap
kubectl create configmap config-multi-env-files \
--from-env-file=configure-pod-container/configmap/game-env-file.properties \
--from-env-file=configure-pod-container/configmap/ui-env-file.properties
would produce the following ConfigMap:
kubectl get configmap config-multi-env-files -o yaml
where the output is similar to this:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2017-12-27T18:38:34Z
name: config-multi-env-files
namespace: default
resourceVersion: "810136"
uid: 252c4572-eb35-11e7-887b-42010a8002b8
data:
color: purple
how: fairlyNice
textmode: "true"
You can define a key other than the file name to use in the data section of your ConfigMap when using the --from-file argument:
kubectl create configmap game-config-3 --from-file=<my-key-name>=<path-to-file>
where <my-key-name> is the key you want to use in the ConfigMap and <path-to-file> is the location of the data source file you want the key to represent.
For example:
kubectl create configmap game-config-3 --from-file=game-special-key=configure-pod-container/configmap/game.properties
would produce the following ConfigMap:
kubectl get configmaps game-config-3 -o yaml
where the output is similar to this:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2016-02-18T18:54:22Z
name: game-config-3
namespace: default
resourceVersion: "530"
uid: 05f8da22-d671-11e5-8cd0-68f728db1985
data:
game-special-key: |
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
You can use kubectl create configmap with the --from-literal argument to define a literal value from the command line:
kubectl create configmap special-config --from-literal=special.how=very --from-literal=special.type=charm
You can pass in multiple key-value pairs. Each pair provided on the command line is represented as a separate entry in the data section of the ConfigMap.
kubectl get configmaps special-config -o yaml
The output is similar to this:
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2016-02-18T19:14:38Z
name: special-config
namespace: default
resourceVersion: "651"
uid: dadce046-d673-11e5-8cd0-68f728db1985
data:
special.how: very
special.type: charm
kubectl supports kustomization.yaml since 1.14.
You can also create a ConfigMap from generators and then apply it to create the object on
the Apiserver. The generators
should be specified in a kustomization.yaml inside a directory.
For example, to generate a ConfigMap from files configure-pod-container/configmap/game.properties
# Create a kustomization.yaml file with ConfigMapGenerator
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: game-config-4
files:
- configure-pod-container/configmap/game.properties
EOF
Apply the kustomization directory to create the ConfigMap object.
kubectl apply -k .
configmap/game-config-4-m9dm2f92bt created
You can check that the ConfigMap was created like this:
kubectl get configmap
NAME DATA AGE
game-config-4-m9dm2f92bt 1 37s
kubectl describe configmaps/game-config-4-m9dm2f92bt
Name: game-config-4-m9dm2f92bt
Namespace: default
Labels: <none>
Annotations: kubectl.kubernetes.io/last-applied-configuration:
{"apiVersion":"v1","data":{"game.properties":"enemies=aliens\nlives=3\nenemies.cheat=true\nenemies.cheat.level=noGoodRotten\nsecret.code.p...
Data
====
game.properties:
----
enemies=aliens
lives=3
enemies.cheat=true
enemies.cheat.level=noGoodRotten
secret.code.passphrase=UUDDLRLRBABAS
secret.code.allowed=true
secret.code.lives=30
Events: <none>
Note that the generated ConfigMap name has a suffix appended by hashing the contents. This ensures that a new ConfigMap is generated each time the content is modified.
You can define a key other than the file name to use in the ConfigMap generator.
For example, to generate a ConfigMap from files configure-pod-container/configmap/game.properties
with the key game-special-key
# Create a kustomization.yaml file with ConfigMapGenerator
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: game-config-5
files:
- game-special-key=configure-pod-container/configmap/game.properties
EOF
Apply the kustomization directory to create the ConfigMap object.
kubectl apply -k .
configmap/game-config-5-m67dt67794 created
To generate a ConfigMap from literals special.type=charm and special.how=very,
you can specify the ConfigMap generator in kustomization.yaml as
# Create a kustomization.yaml file with ConfigMapGenerator
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: special-config-2
literals:
- special.how=very
- special.type=charm
EOF
Apply the kustomization directory to create the ConfigMap object.
kubectl apply -k .
configmap/special-config-2-c92b5mmcf2 created
Define an environment variable as a key-value pair in a ConfigMap:
kubectl create configmap special-config --from-literal=special.how=very
Assign the special.how value defined in the ConfigMap to the SPECIAL_LEVEL_KEY environment variable in the Pod specification.
apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/sh", "-c", "env" ]
env:
# Define the environment variable
- name: SPECIAL_LEVEL_KEY
valueFrom:
configMapKeyRef:
# The ConfigMap containing the value you want to assign to SPECIAL_LEVEL_KEY
name: special-config
# Specify the key associated with the value
key: special.how
restartPolicy: Never
Create the Pod:
kubectl create -f https://kubernetes.io/examples/pods/pod-single-configmap-env-variable.yaml
Now, the Pod's output includes environment variable SPECIAL_LEVEL_KEY=very.
As with the previous example, create the ConfigMaps first.
apiVersion: v1
kind: ConfigMap
metadata:
name: special-config
namespace: default
data:
special.how: very
---
apiVersion: v1
kind: ConfigMap
metadata:
name: env-config
namespace: default
data:
log_level: INFO
Create the ConfigMap:
kubectl create -f https://kubernetes.io/examples/configmap/configmaps.yaml
Define the environment variables in the Pod specification.
apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/sh", "-c", "env" ]
env:
- name: SPECIAL_LEVEL_KEY
valueFrom:
configMapKeyRef:
name: special-config
key: special.how
- name: LOG_LEVEL
valueFrom:
configMapKeyRef:
name: env-config
key: log_level
restartPolicy: Never
Create the Pod:
kubectl create -f https://kubernetes.io/examples/pods/pod-multiple-configmap-env-variable.yaml
Now, the Pod's output includes environment variables SPECIAL_LEVEL_KEY=very and LOG_LEVEL=INFO.
Create a ConfigMap containing multiple key-value pairs.
apiVersion: v1
kind: ConfigMap
metadata:
name: special-config
namespace: default
data:
SPECIAL_LEVEL: very
SPECIAL_TYPE: charm
Create the ConfigMap:
kubectl create -f https://kubernetes.io/examples/configmap/configmap-multikeys.yaml
envFrom to define all of the ConfigMap's data as container environment variables. The key from the ConfigMap becomes the environment variable name in the Pod.apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/sh", "-c", "env" ]
envFrom:
- configMapRef:
name: special-config
restartPolicy: Never
Create the Pod:
kubectl create -f https://kubernetes.io/examples/pods/pod-configmap-envFrom.yaml
Now, the Pod's output includes environment variables SPECIAL_LEVEL=very and SPECIAL_TYPE=charm.
You can use ConfigMap-defined environment variables in the command and args of a container using the $(VAR_NAME) Kubernetes substitution syntax.
For example, the following Pod specification
apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/echo", "$(SPECIAL_LEVEL_KEY) $(SPECIAL_TYPE_KEY)" ]
env:
- name: SPECIAL_LEVEL_KEY
valueFrom:
configMapKeyRef:
name: special-config
key: SPECIAL_LEVEL
- name: SPECIAL_TYPE_KEY
valueFrom:
configMapKeyRef:
name: special-config
key: SPECIAL_TYPE
restartPolicy: Never
created by running
kubectl create -f https://kubernetes.io/examples/pods/pod-configmap-env-var-valueFrom.yaml
produces the following output in the test-container container:
very charm
As explained in Create ConfigMaps from files, when you create a ConfigMap using --from-file, the filename becomes a key stored in the data section of the ConfigMap. The file contents become the key's value.
The examples in this section refer to a ConfigMap named special-config, shown below.
apiVersion: v1
kind: ConfigMap
metadata:
name: special-config
namespace: default
data:
SPECIAL_LEVEL: very
SPECIAL_TYPE: charm
Create the ConfigMap:
kubectl create -f https://kubernetes.io/examples/configmap/configmap-multikeys.yaml
Add the ConfigMap name under the volumes section of the Pod specification.
This adds the ConfigMap data to the directory specified as volumeMounts.mountPath (in this case, /etc/config).
The command section lists directory files with names that match the keys in ConfigMap.
apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/sh", "-c", "ls /etc/config/" ]
volumeMounts:
- name: config-volume
mountPath: /etc/config
volumes:
- name: config-volume
configMap:
# Provide the name of the ConfigMap containing the files you want
# to add to the container
name: special-config
restartPolicy: Never
Create the Pod:
kubectl create -f https://kubernetes.io/examples/pods/pod-configmap-volume.yaml
When the pod runs, the command ls /etc/config/ produces the output below:
SPECIAL_LEVEL
SPECIAL_TYPE
/etc/config/ directory, they will be deleted.
Use the path field to specify the desired file path for specific ConfigMap items.
In this case, the SPECIAL_LEVEL item will be mounted in the config-volume volume at /etc/config/keys.
apiVersion: v1
kind: Pod
metadata:
name: dapi-test-pod
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "/bin/sh","-c","cat /etc/config/keys" ]
volumeMounts:
- name: config-volume
mountPath: /etc/config
volumes:
- name: config-volume
configMap:
name: special-config
items:
- key: SPECIAL_LEVEL
path: keys
restartPolicy: Never
Create the Pod:
kubectl create -f https://kubernetes.io/examples/pods/pod-configmap-volume-specific-key.yaml
When the pod runs, the command cat /etc/config/keys produces the output below:
very
/etc/config/ directory will be deleted.
You can project keys to specific paths and specific permissions on a per-file basis. The Secrets user guide explains the syntax.
A ConfigMap reference may be marked "optional". If the ConfigMap is non-existent, the mounted volume will be empty. If the ConfigMap exists, but the referenced key is non-existent the path will be absent beneath the mount point.
When a mounted ConfigMap is updated, the projected content is eventually updated too. This applies in the case where an optionally referenced ConfigMap comes into existence after a pod has started.
Kubelet checks whether the mounted ConfigMap is fresh on every periodic sync. However, it uses its local TTL-based cache for getting the current value of the ConfigMap. As a result, the total delay from the moment when the ConfigMap is updated to the moment when new keys are projected to the pod can be as long as kubelet sync period (1 minute by default) + TTL of ConfigMaps cache (1 minute by default) in kubelet. You can trigger an immediate refresh by updating one of the pod's annotations.
The ConfigMap API resource stores configuration data as key-value pairs. The data can be consumed in pods or provide the configurations for system components such as controllers. ConfigMap is similar to Secrets, but provides a means of working with strings that don't contain sensitive information. Users and system components alike can store configuration data in ConfigMap.
/etc directory and its contents. For example, if you create a Kubernetes Volume from a ConfigMap, each data item in the ConfigMap is represented by an individual file in the volume.
The ConfigMap's data field contains the configuration data. As shown in the example below, this can be simple -- like individual properties defined using --from-literal -- or complex -- like configuration files or JSON blobs defined using --from-file.
apiVersion: v1
kind: ConfigMap
metadata:
creationTimestamp: 2016-02-18T19:14:38Z
name: example-config
namespace: default
data:
# example of a simple property defined using --from-literal
example.property.1: hello
example.property.2: world
# example of a complex property defined using --from-file
example.property.file: |-
property.1=value-1
property.2=value-2
property.3=value-3
You must create a ConfigMap before referencing it in a Pod specification (unless you mark the ConfigMap as "optional"). If you reference a ConfigMap that doesn't exist, the Pod won't start. Likewise, references to keys that don't exist in the ConfigMap will prevent the pod from starting.
If you use envFrom to define environment variables from ConfigMaps, keys that are considered invalid will be skipped. The pod will be allowed to start, but the invalid names will be recorded in the event log (InvalidVariableNames). The log message lists each skipped key. For example:
kubectl get events
The output is similar to this:
LASTSEEN FIRSTSEEN COUNT NAME KIND SUBOBJECT TYPE REASON SOURCE MESSAGE
0s 0s 1 dapi-test-pod Pod Warning InvalidEnvironmentVariableNames {kubelet, 127.0.0.1} Keys [1badkey, 2alsobad] from the EnvFrom configMap default/myconfig were skipped since they are considered invalid environment variable names.
ConfigMaps reside in a specific Namespace. A ConfigMap can only be referenced by pods residing in the same namespace.
You can't use ConfigMaps for static pods, because the Kubelet does not support this.
Kubernetes v1.17 [stable]
This page shows how to configure process namespace sharing for a pod. When process namespace sharing is enabled, processes in a container are visible to all other containers in that pod.
You can use this feature to configure cooperating containers, such as a log handler sidecar container, or to troubleshoot container images that don't include debugging utilities like a shell.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.10. To check the version, enterkubectl version.
Process Namespace Sharing is enabled using the shareProcessNamespace field of
v1.PodSpec. For example:
Create the pod nginx on your cluster:
kubectl apply -f https://k8s.io/examples/pods/share-process-namespace.yaml
Attach to the shell container and run ps:
kubectl attach -it nginx -c shell
If you don't see a command prompt, try pressing enter.
/ # ps ax
PID USER TIME COMMAND
1 root 0:00 /pause
8 root 0:00 nginx: master process nginx -g daemon off;
14 101 0:00 nginx: worker process
15 root 0:00 sh
21 root 0:00 ps ax
You can signal processes in other containers. For example, send SIGHUP to
nginx to restart the worker process. This requires the SYS_PTRACE capability.
/ # kill -HUP 8
/ # ps ax
PID USER TIME COMMAND
1 root 0:00 /pause
8 root 0:00 nginx: master process nginx -g daemon off;
15 root 0:00 sh
22 101 0:00 nginx: worker process
23 root 0:00 ps ax
It's even possible to access another container image using the
/proc/$pid/root link.
/ # head /proc/8/root/etc/nginx/nginx.conf
user nginx;
worker_processes 1;
error_log /var/log/nginx/error.log warn;
pid /var/run/nginx.pid;
events {
worker_connections 1024;
Pods share many resources so it makes sense they would also share a process namespace. Some container images may expect to be isolated from other containers, though, so it's important to understand these differences:
The container process no longer has PID 1. Some container images refuse
to start without PID 1 (for example, containers using systemd) or run
commands like kill -HUP 1 to signal the container process. In pods with a
shared process namespace, kill -HUP 1 will signal the pod sandbox.
(/pause in the above example.)
Processes are visible to other containers in the pod. This includes all
information visible in /proc, such as passwords that were passed as arguments
or environment variables. These are protected only by regular Unix permissions.
Container filesystems are visible to other containers in the pod through the
/proc/$pid/root link. This makes debugging easier, but it also means
that filesystem secrets are protected only by filesystem permissions.
Static Pods are managed directly by the kubelet daemon on a specific node, without the API server observing them. Unlike Pods that are managed by the control plane (for example, a Deployment); instead, the kubelet watches each static Pod (and restarts it if it fails).
Static Pods are always bound to one Kubelet on a specific node.
The kubelet automatically tries to create a mirror Pod on the Kubernetes API server for each static Pod. This means that the Pods running on a node are visible on the API server, but cannot be controlled from there. The Pod names will be suffixed with the node hostname with a leading hyphen.
spec of a static Pod cannot refer to other API objects
(e.g., ServiceAccount,
ConfigMap,
Secret, etc).
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
This page assumes you're using Docker to run Pods, and that your nodes are running the Fedora operating system. Instructions for other distributions or Kubernetes installations may vary.
You can configure a static Pod with either a file system hosted configuration file or a web hosted configuration file.
Manifests are standard Pod definitions in JSON or YAML format in a specific directory. Use the staticPodPath: <the directory> field in the
kubelet configuration file,
which periodically scans the directory and creates/deletes static Pods as YAML/JSON files appear/disappear there.
Note that the kubelet will ignore files starting with dots when scanning the specified directory.
For example, this is how to start a simple web server as a static Pod:
Choose a node where you want to run the static Pod. In this example, it's my-node1.
ssh my-node1
Choose a directory, say /etc/kubelet.d and place a web server Pod definition there, for example /etc/kubelet.d/static-web.yaml:
# Run this command on the node where kubelet is running
mkdir /etc/kubelet.d/
cat <<EOF >/etc/kubelet.d/static-web.yaml
apiVersion: v1
kind: Pod
metadata:
name: static-web
labels:
role: myrole
spec:
containers:
- name: web
image: nginx
ports:
- name: web
containerPort: 80
protocol: TCP
EOF
Configure your kubelet on the node to use this directory by running it with --pod-manifest-path=/etc/kubelet.d/ argument. On Fedora edit /etc/kubernetes/kubelet to include this line:
KUBELET_ARGS="--cluster-dns=10.254.0.10 --cluster-domain=kube.local --pod-manifest-path=/etc/kubelet.d/"
or add the staticPodPath: <the directory> field in the
kubelet configuration file.
Restart the kubelet. On Fedora, you would run:
# Run this command on the node where the kubelet is running
systemctl restart kubelet
Kubelet periodically downloads a file specified by --manifest-url=<URL> argument
and interprets it as a JSON/YAML file that contains Pod definitions.
Similar to how filesystem-hosted manifests work, the kubelet
refetches the manifest on a schedule. If there are changes to the list of static
Pods, the kubelet applies them.
To use this approach:
Create a YAML file and store it on a web server so that you can pass the URL of that file to the kubelet.
apiVersion: v1
kind: Pod
metadata:
name: static-web
labels:
role: myrole
spec:
containers:
- name: web
image: nginx
ports:
- name: web
containerPort: 80
protocol: TCP
Configure the kubelet on your selected node to use this web manifest by running it with --manifest-url=<manifest-url>. On Fedora, edit /etc/kubernetes/kubelet to include this line:
KUBELET_ARGS="--cluster-dns=10.254.0.10 --cluster-domain=kube.local --manifest-url=<manifest-url>"
Restart the kubelet. On Fedora, you would run:
# Run this command on the node where the kubelet is running
systemctl restart kubelet
When the kubelet starts, it automatically starts all defined static Pods. As you have defined a static Pod and restarted the kubelet, the new static Pod should already be running.
You can view running containers (including static Pods) by running (on the node):
# Run this command on the node where the kubelet is running
docker ps
The output might be something like:
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
f6d05272b57e nginx:latest "nginx" 8 minutes ago Up 8 minutes k8s_web.6f802af4_static-web-fk-node1_default_67e24ed9466ba55986d120c867395f3c_378e5f3c
You can see the mirror Pod on the API server:
kubectl get pods
NAME READY STATUS RESTARTS AGE
static-web-my-node1 1/1 Running 0 2m
Labels from the static Pod are propagated into the mirror Pod. You can use those labels as normal via selectors, etc.
If you try to use kubectl to delete the mirror Pod from the API server,
the kubelet doesn't remove the static Pod:
kubectl delete pod static-web-my-node1
pod "static-web-my-node1" deleted
You can see that the Pod is still running:
kubectl get pods
NAME READY STATUS RESTARTS AGE
static-web-my-node1 1/1 Running 0 12s
Back on your node where the kubelet is running, you can try to stop the Docker container manually. You'll see that, after a time, the kubelet will notice and will restart the Pod automatically:
# Run these commands on the node where the kubelet is running
docker stop f6d05272b57e # replace with the ID of your container
sleep 20
docker ps
CONTAINER ID IMAGE COMMAND CREATED ...
5b920cbaf8b1 nginx:latest "nginx -g 'daemon of 2 seconds ago ...
The running kubelet periodically scans the configured directory (/etc/kubelet.d in our example) for changes and adds/removes Pods as files appear/disappear in this directory.
# This assumes you are using filesystem-hosted static Pod configuration
# Run these commands on the node where the kubelet is running
#
mv /etc/kubelet.d/static-web.yaml /tmp
sleep 20
docker ps
# You see that no nginx container is running
mv /tmp/static-web.yaml /etc/kubelet.d/
sleep 20
docker ps
CONTAINER ID IMAGE COMMAND CREATED ...
e7a62e3427f1 nginx:latest "nginx -g 'daemon of 27 seconds ago
What's Kompose? It's a conversion tool for all things compose (namely Docker Compose) to container orchestrators (Kubernetes or OpenShift).
More information can be found on the Kompose website at http://kompose.io.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
We have multiple ways to install Kompose. Our preferred method is downloading the binary from the latest GitHub release.
Kompose is released via GitHub on a three-week cycle, you can see all current releases on the GitHub release page.
# Linux
curl -L https://github.com/kubernetes/kompose/releases/download/v1.24.0/kompose-linux-amd64 -o kompose
# macOS
curl -L https://github.com/kubernetes/kompose/releases/download/v1.24.0/kompose-darwin-amd64 -o kompose
# Windows
curl -L https://github.com/kubernetes/kompose/releases/download/v1.24.0/kompose-windows-amd64.exe -o kompose.exe
chmod +x kompose
sudo mv ./kompose /usr/local/bin/kompose
Alternatively, you can download the tarball.
Installing using go get pulls from the master branch with the latest development changes.
go get -u github.com/kubernetes/kompose
Kompose is in EPEL CentOS repository.
If you don't have EPEL repository already installed and enabled you can do it by running sudo yum install epel-release
If you have EPEL enabled in your system, you can install Kompose like any other package.
sudo yum -y install kompose
Kompose is in Fedora 24, 25 and 26 repositories. You can install it like any other package.
sudo dnf -y install kompose
On macOS you can install latest release via Homebrew:
brew install kompose
In a few steps, we'll take you from Docker Compose to Kubernetes. All
you need is an existing docker-compose.yml file.
Go to the directory containing your docker-compose.yml file. If you don't have one, test using this one.
version: "2"
services:
redis-master:
image: k8s.gcr.io/redis:e2e
ports:
- "6379"
redis-slave:
image: gcr.io/google_samples/gb-redisslave:v3
ports:
- "6379"
environment:
- GET_HOSTS_FROM=dns
frontend:
image: gcr.io/google-samples/gb-frontend:v4
ports:
- "80:80"
environment:
- GET_HOSTS_FROM=dns
labels:
kompose.service.type: LoadBalancer
To convert the docker-compose.yml file to files that you can use with
kubectl, run kompose convert and then kubectl apply -f <output file>.
kompose convert
The output is similar to:
INFO Kubernetes file "frontend-service.yaml" created
INFO Kubernetes file "frontend-service.yaml" created
INFO Kubernetes file "frontend-service.yaml" created
INFO Kubernetes file "redis-master-service.yaml" created
INFO Kubernetes file "redis-master-service.yaml" created
INFO Kubernetes file "redis-master-service.yaml" created
INFO Kubernetes file "redis-slave-service.yaml" created
INFO Kubernetes file "redis-slave-service.yaml" created
INFO Kubernetes file "redis-slave-service.yaml" created
INFO Kubernetes file "frontend-deployment.yaml" created
INFO Kubernetes file "frontend-deployment.yaml" created
INFO Kubernetes file "frontend-deployment.yaml" created
INFO Kubernetes file "redis-master-deployment.yaml" created
INFO Kubernetes file "redis-master-deployment.yaml" created
INFO Kubernetes file "redis-master-deployment.yaml" created
INFO Kubernetes file "redis-slave-deployment.yaml" created
INFO Kubernetes file "redis-slave-deployment.yaml" created
INFO Kubernetes file "redis-slave-deployment.yaml" created
kubectl apply -f frontend-service.yaml,redis-master-service.yaml,redis-slave-service.yaml,frontend-deployment.yaml,
The output is similar to:
redis-master-deployment.yaml,redis-slave-deployment.yaml
service/frontend created
service/redis-master created
service/redis-slave created
deployment.apps/frontend created
deployment.apps/redis-master created
deployment.apps/redis-slave created
Your deployments are running in Kubernetes.
Access your application.
If you're already using minikube for your development process:
minikube service frontend
Otherwise, let's look up what IP your service is using!
kubectl describe svc frontend
Name: frontend
Namespace: default
Labels: service=frontend
Selector: service=frontend
Type: LoadBalancer
IP: 10.0.0.183
LoadBalancer Ingress: 192.0.2.89
Port: 80 80/TCP
NodePort: 80 31144/TCP
Endpoints: 172.17.0.4:80
Session Affinity: None
No events.
If you're using a cloud provider, your IP will be listed next to LoadBalancer Ingress.
curl http://192.0.2.89
Kompose has support for two providers: OpenShift and Kubernetes.
You can choose a targeted provider using global option --provider. If no provider is specified, Kubernetes is set by default.
kompose convertKompose supports conversion of V1, V2, and V3 Docker Compose files into Kubernetes and OpenShift objects.
kompose convert examplekompose --file docker-voting.yml convert
WARN Unsupported key networks - ignoring
WARN Unsupported key build - ignoring
INFO Kubernetes file "worker-svc.yaml" created
INFO Kubernetes file "db-svc.yaml" created
INFO Kubernetes file "redis-svc.yaml" created
INFO Kubernetes file "result-svc.yaml" created
INFO Kubernetes file "vote-svc.yaml" created
INFO Kubernetes file "redis-deployment.yaml" created
INFO Kubernetes file "result-deployment.yaml" created
INFO Kubernetes file "vote-deployment.yaml" created
INFO Kubernetes file "worker-deployment.yaml" created
INFO Kubernetes file "db-deployment.yaml" created
ls
db-deployment.yaml docker-compose.yml docker-gitlab.yml redis-deployment.yaml result-deployment.yaml vote-deployment.yaml worker-deployment.yaml
db-svc.yaml docker-voting.yml redis-svc.yaml result-svc.yaml vote-svc.yaml worker-svc.yaml
You can also provide multiple docker-compose files at the same time:
kompose -f docker-compose.yml -f docker-guestbook.yml convert
INFO Kubernetes file "frontend-service.yaml" created
INFO Kubernetes file "mlbparks-service.yaml" created
INFO Kubernetes file "mongodb-service.yaml" created
INFO Kubernetes file "redis-master-service.yaml" created
INFO Kubernetes file "redis-slave-service.yaml" created
INFO Kubernetes file "frontend-deployment.yaml" created
INFO Kubernetes file "mlbparks-deployment.yaml" created
INFO Kubernetes file "mongodb-deployment.yaml" created
INFO Kubernetes file "mongodb-claim0-persistentvolumeclaim.yaml" created
INFO Kubernetes file "redis-master-deployment.yaml" created
INFO Kubernetes file "redis-slave-deployment.yaml" created
ls
mlbparks-deployment.yaml mongodb-service.yaml redis-slave-service.jsonmlbparks-service.yaml
frontend-deployment.yaml mongodb-claim0-persistentvolumeclaim.yaml redis-master-service.yaml
frontend-service.yaml mongodb-deployment.yaml redis-slave-deployment.yaml
redis-master-deployment.yaml
When multiple docker-compose files are provided the configuration is merged. Any configuration that is common will be over ridden by subsequent file.
kompose convert examplekompose --provider openshift --file docker-voting.yml convert
WARN [worker] Service cannot be created because of missing port.
INFO OpenShift file "vote-service.yaml" created
INFO OpenShift file "db-service.yaml" created
INFO OpenShift file "redis-service.yaml" created
INFO OpenShift file "result-service.yaml" created
INFO OpenShift file "vote-deploymentconfig.yaml" created
INFO OpenShift file "vote-imagestream.yaml" created
INFO OpenShift file "worker-deploymentconfig.yaml" created
INFO OpenShift file "worker-imagestream.yaml" created
INFO OpenShift file "db-deploymentconfig.yaml" created
INFO OpenShift file "db-imagestream.yaml" created
INFO OpenShift file "redis-deploymentconfig.yaml" created
INFO OpenShift file "redis-imagestream.yaml" created
INFO OpenShift file "result-deploymentconfig.yaml" created
INFO OpenShift file "result-imagestream.yaml" created
It also supports creating buildconfig for build directive in a service. By default, it uses the remote repo for the current git branch as the source repo, and the current branch as the source branch for the build. You can specify a different source repo and branch using --build-repo and --build-branch options respectively.
kompose --provider openshift --file buildconfig/docker-compose.yml convert
WARN [foo] Service cannot be created because of missing port.
INFO OpenShift Buildconfig using git@github.com:rtnpro/kompose.git::master as source.
INFO OpenShift file "foo-deploymentconfig.yaml" created
INFO OpenShift file "foo-imagestream.yaml" created
INFO OpenShift file "foo-buildconfig.yaml" created
oc create -f, you need to ensure that you push the imagestream artifact before the buildconfig artifact, to workaround this OpenShift issue: https://github.com/openshift/origin/issues/4518 .
kompose upKompose supports a straightforward way to deploy your "composed" application to Kubernetes or OpenShift via kompose up.
kompose up examplekompose --file ./examples/docker-guestbook.yml up
We are going to create Kubernetes deployments and services for your Dockerized application.
If you need different kind of resources, use the 'kompose convert' and 'kubectl apply -f' commands instead.
INFO Successfully created service: redis-master
INFO Successfully created service: redis-slave
INFO Successfully created service: frontend
INFO Successfully created deployment: redis-master
INFO Successfully created deployment: redis-slave
INFO Successfully created deployment: frontend
Your application has been deployed to Kubernetes. You can run 'kubectl get deployment,svc,pods' for details.
kubectl get deployment,svc,pods
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
deployment.extensions/frontend 1 1 1 1 4m
deployment.extensions/redis-master 1 1 1 1 4m
deployment.extensions/redis-slave 1 1 1 1 4m
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/frontend ClusterIP 10.0.174.12 <none> 80/TCP 4m
service/kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 13d
service/redis-master ClusterIP 10.0.202.43 <none> 6379/TCP 4m
service/redis-slave ClusterIP 10.0.1.85 <none> 6379/TCP 4m
NAME READY STATUS RESTARTS AGE
pod/frontend-2768218532-cs5t5 1/1 Running 0 4m
pod/redis-master-1432129712-63jn8 1/1 Running 0 4m
pod/redis-slave-2504961300-nve7b 1/1 Running 0 4m
kompose convert and kubectl apply -f commands instead.kompose up examplekompose --file ./examples/docker-guestbook.yml --provider openshift up
We are going to create OpenShift DeploymentConfigs and Services for your Dockerized application.
If you need different kind of resources, use the 'kompose convert' and 'oc create -f' commands instead.
INFO Successfully created service: redis-slave
INFO Successfully created service: frontend
INFO Successfully created service: redis-master
INFO Successfully created deployment: redis-slave
INFO Successfully created ImageStream: redis-slave
INFO Successfully created deployment: frontend
INFO Successfully created ImageStream: frontend
INFO Successfully created deployment: redis-master
INFO Successfully created ImageStream: redis-master
Your application has been deployed to OpenShift. You can run 'oc get dc,svc,is' for details.
oc get dc,svc,is
NAME REVISION DESIRED CURRENT TRIGGERED BY
dc/frontend 0 1 0 config,image(frontend:v4)
dc/redis-master 0 1 0 config,image(redis-master:e2e)
dc/redis-slave 0 1 0 config,image(redis-slave:v1)
NAME CLUSTER-IP EXTERNAL-IP PORT(S) AGE
svc/frontend 172.30.46.64 <none> 80/TCP 8s
svc/redis-master 172.30.144.56 <none> 6379/TCP 8s
svc/redis-slave 172.30.75.245 <none> 6379/TCP 8s
NAME DOCKER REPO TAGS UPDATED
is/frontend 172.30.12.200:5000/fff/frontend
is/redis-master 172.30.12.200:5000/fff/redis-master
is/redis-slave 172.30.12.200:5000/fff/redis-slave v1
oc context (oc login).
kompose downOnce you have deployed "composed" application to Kubernetes, kompose down will help you to take the application out by deleting its deployments and services. If you need to remove other resources, use the 'kubectl' command.
kompose --file docker-guestbook.yml down
INFO Successfully deleted service: redis-master
INFO Successfully deleted deployment: redis-master
INFO Successfully deleted service: redis-slave
INFO Successfully deleted deployment: redis-slave
INFO Successfully deleted service: frontend
INFO Successfully deleted deployment: frontend
kubectl context.
Kompose supports both building and pushing Docker images. When using the build key within your Docker Compose file, your image will:
image key specified within your file.docker/config)Using an example Docker Compose file:
version: "2"
services:
foo:
build: "./build"
image: docker.io/foo/bar
Using kompose up with a build key:
kompose up
INFO Build key detected. Attempting to build and push image 'docker.io/foo/bar'
INFO Building image 'docker.io/foo/bar' from directory 'build'
INFO Image 'docker.io/foo/bar' from directory 'build' built successfully
INFO Pushing image 'foo/bar:latest' to registry 'docker.io'
INFO Attempting authentication credentials 'https://index.docker.io/v1/
INFO Successfully pushed image 'foo/bar:latest' to registry 'docker.io'
INFO We are going to create Kubernetes Deployments, Services and PersistentVolumeClaims for your Dockerized application. If you need different kind of resources, use the 'kompose convert' and 'kubectl apply -f' commands instead.
INFO Deploying application in "default" namespace
INFO Successfully created Service: foo
INFO Successfully created Deployment: foo
Your application has been deployed to Kubernetes. You can run 'kubectl get deployment,svc,pods,pvc' for details.
In order to disable the functionality, or choose to use BuildConfig generation (with OpenShift) --build (local|build-config|none) can be passed.
# Disable building/pushing Docker images
kompose up --build none
# Generate Build Config artifacts for OpenShift
kompose up --provider openshift --build build-config
The default kompose transformation will generate Kubernetes Deployments and Services, in yaml format. You have alternative option to generate json with -j. Also, you can alternatively generate Replication Controllers objects, Daemon Sets, or Helm charts.
kompose convert -j
INFO Kubernetes file "redis-svc.json" created
INFO Kubernetes file "web-svc.json" created
INFO Kubernetes file "redis-deployment.json" created
INFO Kubernetes file "web-deployment.json" created
The *-deployment.json files contain the Deployment objects.
kompose convert --replication-controller
INFO Kubernetes file "redis-svc.yaml" created
INFO Kubernetes file "web-svc.yaml" created
INFO Kubernetes file "redis-replicationcontroller.yaml" created
INFO Kubernetes file "web-replicationcontroller.yaml" created
The *-replicationcontroller.yaml files contain the Replication Controller objects. If you want to specify replicas (default is 1), use --replicas flag: kompose convert --replication-controller --replicas 3
kompose convert --daemon-set
INFO Kubernetes file "redis-svc.yaml" created
INFO Kubernetes file "web-svc.yaml" created
INFO Kubernetes file "redis-daemonset.yaml" created
INFO Kubernetes file "web-daemonset.yaml" created
The *-daemonset.yaml files contain the DaemonSet objects
If you want to generate a Chart to be used with Helm run:
kompose convert -c
INFO Kubernetes file "web-svc.yaml" created
INFO Kubernetes file "redis-svc.yaml" created
INFO Kubernetes file "web-deployment.yaml" created
INFO Kubernetes file "redis-deployment.yaml" created
chart created in "./docker-compose/"
tree docker-compose/
docker-compose
├── Chart.yaml
├── README.md
└── templates
├── redis-deployment.yaml
├── redis-svc.yaml
├── web-deployment.yaml
└── web-svc.yaml
The chart structure is aimed at providing a skeleton for building your Helm charts.
kompose supports Kompose-specific labels within the docker-compose.yml file in order to explicitly define a service's behavior upon conversion.
kompose.service.type defines the type of service to be created.For example:
version: "2"
services:
nginx:
image: nginx
dockerfile: foobar
build: ./foobar
cap_add:
- ALL
container_name: foobar
labels:
kompose.service.type: nodeport
kompose.service.expose defines if the service needs to be made accessible from outside the cluster or not. If the value is set to "true", the provider sets the endpoint automatically, and for any other value, the value is set as the hostname. If multiple ports are defined in a service, the first one is chosen to be the exposed.
For example:
version: "2"
services:
web:
image: tuna/docker-counter23
ports:
- "5000:5000"
links:
- redis
labels:
kompose.service.expose: "counter.example.com"
redis:
image: redis:3.0
ports:
- "6379"
The currently supported options are:
| Key | Value |
|---|---|
| kompose.service.type | nodeport / clusterip / loadbalancer |
| kompose.service.expose | true / hostname |
kompose.service.type label should be defined with ports only, otherwise kompose will fail.
If you want to create normal pods without controllers you can use restart construct of docker-compose to define that. Follow table below to see what happens on the restart value.
docker-compose restart |
object created | Pod restartPolicy |
|---|---|---|
"" |
controller object | Always |
always |
controller object | Always |
on-failure |
Pod | OnFailure |
no |
Pod | Never |
deployment or replicationcontroller.
For example, the pival service will become pod down here. This container calculated value of pi.
version: '2'
services:
pival:
image: perl
command: ["perl", "-Mbignum=bpi", "-wle", "print bpi(2000)"]
restart: "on-failure"
If the Docker Compose file has a volume specified for a service, the Deployment (Kubernetes) or DeploymentConfig (OpenShift) strategy is changed to "Recreate" instead of "RollingUpdate" (default). This is done to avoid multiple instances of a service from accessing a volume at the same time.
If the Docker Compose file has service name with _ in it (eg.web_service), then it will be replaced by - and the service name will be renamed accordingly (eg.web-service). Kompose does this because "Kubernetes" doesn't allow _ in object name.
Please note that changing service name might break some docker-compose files.
Kompose supports Docker Compose versions: 1, 2 and 3. We have limited support on versions 2.1 and 3.2 due to their experimental nature.
A full list on compatibility between all three versions is listed in our conversion document including a list of all incompatible Docker Compose keys.
As of v1.22, Kubernetes provides a built-in admission controller to enforce the Pod Security Standards. You can configure this admission controller to set cluster-wide defaults and exemptions.
Your Kubernetes server must be version v1.22.
To check the version, enter kubectl version.
PodSecurity feature gate.apiVersion: apiserver.config.k8s.io/v1
kind: AdmissionConfiguration
plugins:
- name: PodSecurity
configuration:
apiVersion: pod-security.admission.config.k8s.io/v1alpha1
kind: PodSecurityConfiguration
# Defaults applied when a mode label is not set.
#
# Level label values must be one of:
# - "privileged" (default)
# - "baseline"
# - "restricted"
#
# Version label values must be one of:
# - "latest" (default)
# - specific version like "v1.22"
defaults:
enforce: "privileged"
enforce-version: "latest"
audit: "privileged"
audit-version: "latest"
warn: "privileged"
warn-version: "latest"
exemptions:
# Array of authenticated usernames to exempt.
usernames: []
# Array of runtime class names to exempt.
runtimeClassNames: []
# Array of namespaces to exempt.
namespaces: []
Namespaces can be labeled to enforce the Pod Security Standards.
Your Kubernetes server must be version v1.22.
To check the version, enter kubectl version.
PodSecurity feature gate.baseline Pod Security Standard with namespace labelsThis manifest defines a Namespace my-baseline-namespace that:
baseline policy requirements.restricted policy requirements.baseline and restricted policies to v1.22.apiVersion: v1
kind: Namespace
metadata:
name: my-baseline-namespace
labels:
pod-security.kubernetes.io/enforce: baseline
pod-security.kubernetes.io/enforce-version: v1.22
# We are setting these to our _desired_ `enforce` level.
pod-security.kubernetes.io/audit: restricted
pod-security.kubernetes.io/audit-version: v1.22
pod-security.kubernetes.io/warn: restricted
pod-security.kubernetes.io/warn-version: v1.22
kubectl labelenforce policy (or version) label is added or changed, the admission plugin will test
each pod in the namespace against the new policy. Violations are returned to the user as warnings.
It is helpful to apply the --dry-run flag when initially evaluating security profile changes for
namespaces. The Pod Security Standard checks will still be run in dry run mode, giving you
information about how the new policy would treat existing pods, without actually updating a policy.
kubectl label --dry-run=server --overwrite ns --all \
pod-security.kubernetes.io/enforce=baseline
If you're just getting started with the Pod Security Standards, a suitable first step would be to
configure all namespaces with audit annotations for a stricter level such as baseline:
kubectl label --overwrite ns --all \
pod-security.kubernetes.io/audit=baseline \
pod-security.kubernetes.io/warn=baseline
Note that this is not setting an enforce level, so that namespaces that haven't been explicitly evaluated can be distinguished. You can list namespaces without an explicitly set enforce level using this command:
kubectl get namespaces --selector='!pod-security.kubernetes.io/enforce'
You can update a specific namespace as well. This command adds the enforce=restricted
policy to my-existing-namespace, pinning the restricted policy version to v1.22.
kubectl label --overwrite ns my-existing-namespace \
pod-security.kubernetes.io/enforce=restricted \
pod-security.kubernetes.io/enforce-version=v1.22
This page describes the process of migrating from PodSecurityPolicies to the built-in PodSecurity
admission controller. This can be done effectively using a combination of dry-run and audit and
warn modes, although this becomes harder if mutating PSPs are used.
Your Kubernetes server must be version v1.22.
To check the version, enter kubectl version.
PodSecurity feature gate.warn and audit mode. This will give you an idea of how
your Pods will respond to the new policies, without breaking existing workloads. Iterate on your
Pods' configuration until
they are in compliance with the selected profiles.enforce mode.PodSecurityPolicy in the --enable-admission-plugins flag.Kubernetes objects can be created, updated, and deleted by storing multiple
object configuration files in a directory and using kubectl apply to
recursively create and update those objects as needed. This method
retains writes made to live objects without merging the changes
back into the object configuration files. kubectl diff also gives you a
preview of what changes apply will make.
Install kubectl.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The kubectl tool supports three kinds of object management:
See Kubernetes Object Management for a discussion of the advantages and disadvantage of each kind of object management.
Declarative object configuration requires a firm understanding of the Kubernetes object definitions and configuration. Read and complete the following documents if you have not already:
Following are definitions for terms used in this document:
kubectl apply. Configuration files are typically stored in source control, such as Git.kubectl apply to write the changes.Use kubectl apply to create all objects, except those that already exist,
defined by configuration files in a specified directory:
kubectl apply -f <directory>/
This sets the kubectl.kubernetes.io/last-applied-configuration: '{...}'
annotation on each object. The annotation contains the contents of the object
configuration file that was used to create the object.
-R flag to recursively process directories.
Here's an example of an object configuration file:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
minReadySeconds: 5
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Run kubectl diff to print the object that will be created:
kubectl diff -f https://k8s.io/examples/application/simple_deployment.yaml
diff uses server-side dry-run,
which needs to be enabled on kube-apiserver.
Since diff performs a server-side apply request in dry-run mode,
it requires granting PATCH, CREATE, and UPDATE permissions.
See Dry-Run Authorization
for details.
Create the object using kubectl apply:
kubectl apply -f https://k8s.io/examples/application/simple_deployment.yaml
Print the live configuration using kubectl get:
kubectl get -f https://k8s.io/examples/application/simple_deployment.yaml -o yaml
The output shows that the kubectl.kubernetes.io/last-applied-configuration annotation
was written to the live configuration, and it matches the configuration file:
kind: Deployment
metadata:
annotations:
# ...
# This is the json representation of simple_deployment.yaml
# It was written by kubectl apply when the object was created
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"minReadySeconds":5,"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.14.2","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
# ...
minReadySeconds: 5
selector:
matchLabels:
# ...
app: nginx
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.14.2
# ...
name: nginx
ports:
- containerPort: 80
# ...
# ...
# ...
# ...
You can also use kubectl apply to update all objects defined in a directory, even
if those objects already exist. This approach accomplishes the following:
kubectl diff -f <directory>/
kubectl apply -f <directory>/
-R flag to recursively process directories.
Here's an example configuration file:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
minReadySeconds: 5
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Create the object using kubectl apply:
kubectl apply -f https://k8s.io/examples/application/simple_deployment.yaml
Print the live configuration using kubectl get:
kubectl get -f https://k8s.io/examples/application/simple_deployment.yaml -o yaml
The output shows that the kubectl.kubernetes.io/last-applied-configuration annotation
was written to the live configuration, and it matches the configuration file:
kind: Deployment
metadata:
annotations:
# ...
# This is the json representation of simple_deployment.yaml
# It was written by kubectl apply when the object was created
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"minReadySeconds":5,"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.14.2","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
# ...
minReadySeconds: 5
selector:
matchLabels:
# ...
app: nginx
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.14.2
# ...
name: nginx
ports:
- containerPort: 80
# ...
# ...
# ...
# ...
Directly update the replicas field in the live configuration by using kubectl scale.
This does not use kubectl apply:
kubectl scale deployment/nginx-deployment --replicas=2
Print the live configuration using kubectl get:
kubectl get deployment nginx-deployment -o yaml
The output shows that the replicas field has been set to 2, and the last-applied-configuration
annotation does not contain a replicas field:
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
# ...
# note that the annotation does not contain replicas
# because it was not updated through apply
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"minReadySeconds":5,"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.14.2","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
replicas: 2 # written by scale
# ...
minReadySeconds: 5
selector:
matchLabels:
# ...
app: nginx
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.14.2
# ...
name: nginx
ports:
- containerPort: 80
# ...
Update the simple_deployment.yaml configuration file to change the image from
nginx:1.14.2 to nginx:1.16.1, and delete the minReadySeconds field:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.16.1 # update the image
ports:
- containerPort: 80
Apply the changes made to the configuration file:
kubectl diff -f https://k8s.io/examples/application/update_deployment.yaml
kubectl apply -f https://k8s.io/examples/application/update_deployment.yaml
Print the live configuration using kubectl get:
kubectl get -f https://k8s.io/examples/application/update_deployment.yaml -o yaml
The output shows the following changes to the live configuration:
replicas field retains the value of 2 set by kubectl scale.
This is possible because it is omitted from the configuration file.image field has been updated to nginx:1.16.1 from nginx:1.14.2.last-applied-configuration annotation has been updated with the new image.minReadySeconds field has been cleared.last-applied-configuration annotation no longer contains the minReadySeconds field.apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
# ...
# The annotation contains the updated image to nginx 1.11.9,
# but does not contain the updated replicas to 2
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.16.1","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
replicas: 2 # Set by `kubectl scale`. Ignored by `kubectl apply`.
# minReadySeconds cleared by `kubectl apply`
# ...
selector:
matchLabels:
# ...
app: nginx
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.16.1 # Set by `kubectl apply`
# ...
name: nginx
ports:
- containerPort: 80
# ...
# ...
# ...
# ...
kubectl apply with the imperative object configuration commands
create and replace is not supported. This is because create
and replace do not retain the kubectl.kubernetes.io/last-applied-configuration
that kubectl apply uses to compute updates.
There are two approaches to delete objects managed by kubectl apply.
kubectl delete -f <filename>Manually deleting objects using the imperative command is the recommended approach, as it is more explicit about what is being deleted, and less likely to result in the user deleting something unintentionally:
kubectl delete -f <filename>
kubectl apply -f <directory/> --prune -l your=labelOnly use this if you know what you are doing.
kubectl apply --prune is in alpha, and backwards incompatible
changes might be introduced in subsequent releases.
As an alternative to kubectl delete, you can use kubectl apply to identify objects to be deleted after their
configuration files have been removed from the directory. Apply with --prune
queries the API server for all objects matching a set of labels, and attempts
to match the returned live object configurations against the object
configuration files. If an object matches the query, and it does not have a
configuration file in the directory, and it has a last-applied-configuration annotation,
it is deleted.
kubectl apply -f <directory/> --prune -l <labels>
-l <labels> and
do not appear in the subdirectory.
You can use kubectl get with -o yaml to view the configuration of a live object:
kubectl get -f <filename|url> -o yaml
When kubectl apply updates the live configuration for an object,
it does so by sending a patch request to the API server. The
patch defines updates scoped to specific fields of the live object
configuration. The kubectl apply command calculates this patch request
using the configuration file, the live configuration, and the
last-applied-configuration annotation stored in the live configuration.
The kubectl apply command writes the contents of the configuration file to the
kubectl.kubernetes.io/last-applied-configuration annotation. This
is used to identify fields that have been removed from the configuration
file and need to be cleared from the live configuration. Here are the steps used
to calculate which fields should be deleted or set:
last-applied-configuration and missing from the configuration file.Here's an example. Suppose this is the configuration file for a Deployment object:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.16.1 # update the image
ports:
- containerPort: 80
Also, suppose this is the live configuration for the same Deployment object:
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
# ...
# note that the annotation does not contain replicas
# because it was not updated through apply
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"minReadySeconds":5,"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.14.2","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
replicas: 2 # written by scale
# ...
minReadySeconds: 5
selector:
matchLabels:
# ...
app: nginx
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.14.2
# ...
name: nginx
ports:
- containerPort: 80
# ...
Here are the merge calculations that would be performed by kubectl apply:
last-applied-configuration and comparing them to values in the
configuration file.
Clear fields explicitly set to null in the local object configuration file
regardless of whether they appear in the last-applied-configuration.
In this example, minReadySeconds appears in the
last-applied-configuration annotation, but does not appear in the configuration file.
Action: Clear minReadySeconds from the live configuration.image in the configuration file does not match
the value in the live configuration. Action: Set the value of image in the live configuration.last-applied-configuration annotation to match the value
of the configuration file.Here is the live configuration that is the result of the merge:
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
# ...
# The annotation contains the updated image to nginx 1.11.9,
# but does not contain the updated replicas to 2
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"apps/v1","kind":"Deployment",
"metadata":{"annotations":{},"name":"nginx-deployment","namespace":"default"},
"spec":{"selector":{"matchLabels":{"app":nginx}},"template":{"metadata":{"labels":{"app":"nginx"}},
"spec":{"containers":[{"image":"nginx:1.16.1","name":"nginx",
"ports":[{"containerPort":80}]}]}}}}
# ...
spec:
selector:
matchLabels:
# ...
app: nginx
replicas: 2 # Set by `kubectl scale`. Ignored by `kubectl apply`.
# minReadySeconds cleared by `kubectl apply`
# ...
template:
metadata:
# ...
labels:
app: nginx
spec:
containers:
- image: nginx:1.16.1 # Set by `kubectl apply`
# ...
name: nginx
ports:
- containerPort: 80
# ...
# ...
# ...
# ...
How a particular field in a configuration file is merged with the live configuration depends on the type of the field. There are several types of fields:
primitive: A field of type string, integer, or boolean.
For example, image and replicas are primitive fields. Action: Replace.
map, also called object: A field of type map or a complex type that contains subfields. For example, labels,
annotations,spec and metadata are all maps. Action: Merge elements or subfields.
list: A field containing a list of items that can be either primitive types or maps.
For example, containers, ports, and args are lists. Action: Varies.
When kubectl apply updates a map or list field, it typically does
not replace the entire field, but instead updates the individual subelements.
For instance, when merging the spec on a Deployment, the entire spec is
not replaced. Instead the subfields of spec, such as replicas, are compared
and merged.
Primitive fields are replaced or cleared.
- is used for "not applicable" because the value is not used.
| Field in object configuration file | Field in live object configuration | Field in last-applied-configuration | Action |
|---|---|---|---|
| Yes | Yes | - | Set live to configuration file value. |
| Yes | No | - | Set live to local configuration. |
| No | - | Yes | Clear from live configuration. |
| No | - | No | Do nothing. Keep live value. |
Fields that represent maps are merged by comparing each of the subfields or elements of the map:
- is used for "not applicable" because the value is not used.
| Key in object configuration file | Key in live object configuration | Field in last-applied-configuration | Action |
|---|---|---|---|
| Yes | Yes | - | Compare sub fields values. |
| Yes | No | - | Set live to local configuration. |
| No | - | Yes | Delete from live configuration. |
| No | - | No | Do nothing. Keep live value. |
Merging changes to a list uses one of three strategies:
The choice of strategy is made on a per-field basis.
Treat the list the same as a primitive field. Replace or delete the entire list. This preserves ordering.
Example: Use kubectl apply to update the args field of a Container in a Pod. This sets
the value of args in the live configuration to the value in the configuration file.
Any args elements that had previously been added to the live configuration are lost.
The order of the args elements defined in the configuration file is
retained in the live configuration.
# last-applied-configuration value
args: ["a", "b"]
# configuration file value
args: ["a", "c"]
# live configuration
args: ["a", "b", "d"]
# result after merge
args: ["a", "c"]
Explanation: The merge used the configuration file value as the new list value.
Treat the list as a map, and treat a specific field of each element as a key. Add, delete, or update individual elements. This does not preserve ordering.
This merge strategy uses a special tag on each field called a patchMergeKey. The
patchMergeKey is defined for each field in the Kubernetes source code:
types.go
When merging a list of maps, the field specified as the patchMergeKey for a given element
is used like a map key for that element.
Example: Use kubectl apply to update the containers field of a PodSpec.
This merges the list as though it was a map where each element is keyed
by name.
# last-applied-configuration value
containers:
- name: nginx
image: nginx:1.16
- name: nginx-helper-a # key: nginx-helper-a; will be deleted in result
image: helper:1.3
- name: nginx-helper-b # key: nginx-helper-b; will be retained
image: helper:1.3
# configuration file value
containers:
- name: nginx
image: nginx:1.16
- name: nginx-helper-b
image: helper:1.3
- name: nginx-helper-c # key: nginx-helper-c; will be added in result
image: helper:1.3
# live configuration
containers:
- name: nginx
image: nginx:1.16
- name: nginx-helper-a
image: helper:1.3
- name: nginx-helper-b
image: helper:1.3
args: ["run"] # Field will be retained
- name: nginx-helper-d # key: nginx-helper-d; will be retained
image: helper:1.3
# result after merge
containers:
- name: nginx
image: nginx:1.16
# Element nginx-helper-a was deleted
- name: nginx-helper-b
image: helper:1.3
args: ["run"] # Field was retained
- name: nginx-helper-c # Element was added
image: helper:1.3
- name: nginx-helper-d # Element was ignored
image: helper:1.3
Explanation:
args
in the live configuration. kubectl apply was able to identify
that "nginx-helper-b" in the live configuration was the same
"nginx-helper-b" as in the configuration file, even though their fields
had different values (no args in the configuration file). This is
because the patchMergeKey field value (name) was identical in both.As of Kubernetes 1.5, merging lists of primitive elements is not supported.
patchStrategy tag in types.go
If no patchStrategy is specified for a field of type list, then
the list is replaced.
The API server sets certain fields to default values in the live configuration if they are not specified when the object is created.
Here's a configuration file for a Deployment. The file does not specify strategy:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
minReadySeconds: 5
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Create the object using kubectl apply:
kubectl apply -f https://k8s.io/examples/application/simple_deployment.yaml
Print the live configuration using kubectl get:
kubectl get -f https://k8s.io/examples/application/simple_deployment.yaml -o yaml
The output shows that the API server set several fields to default values in the live configuration. These fields were not specified in the configuration file.
apiVersion: apps/v1
kind: Deployment
# ...
spec:
selector:
matchLabels:
app: nginx
minReadySeconds: 5
replicas: 1 # defaulted by apiserver
strategy:
rollingUpdate: # defaulted by apiserver - derived from strategy.type
maxSurge: 1
maxUnavailable: 1
type: RollingUpdate # defaulted by apiserver
template:
metadata:
creationTimestamp: null
labels:
app: nginx
spec:
containers:
- image: nginx:1.14.2
imagePullPolicy: IfNotPresent # defaulted by apiserver
name: nginx
ports:
- containerPort: 80
protocol: TCP # defaulted by apiserver
resources: {} # defaulted by apiserver
terminationMessagePath: /dev/termination-log # defaulted by apiserver
dnsPolicy: ClusterFirst # defaulted by apiserver
restartPolicy: Always # defaulted by apiserver
securityContext: {} # defaulted by apiserver
terminationGracePeriodSeconds: 30 # defaulted by apiserver
# ...
In a patch request, defaulted fields are not re-defaulted unless they are explicitly cleared as part of a patch request. This can cause unexpected behavior for fields that are defaulted based on the values of other fields. When the other fields are later changed, the values defaulted from them will not be updated unless they are explicitly cleared.
For this reason, it is recommended that certain fields defaulted by the server are explicitly defined in the configuration file, even if the desired values match the server defaults. This makes it easier to recognize conflicting values that will not be re-defaulted by the server.
Example:
# last-applied-configuration
spec:
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
# configuration file
spec:
strategy:
type: Recreate # updated value
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
# live configuration
spec:
strategy:
type: RollingUpdate # defaulted value
rollingUpdate: # defaulted value derived from type
maxSurge : 1
maxUnavailable: 1
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
# result after merge - ERROR!
spec:
strategy:
type: Recreate # updated value: incompatible with rollingUpdate
rollingUpdate: # defaulted value: incompatible with "type: Recreate"
maxSurge : 1
maxUnavailable: 1
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Explanation:
strategy.type.strategy.type to RollingUpdate and defaults the
strategy.rollingUpdate values.strategy.type to Recreate. The strategy.rollingUpdate
values remain at their defaulted values, though the server expects them to be cleared.
If the strategy.rollingUpdate values had been defined initially in the configuration file,
it would have been more clear that they needed to be deleted.strategy.rollingUpdate is not cleared. The strategy.rollingupdate
field cannot be defined with a strategy.type of Recreate.Recommendation: These fields should be explicitly defined in the object configuration file:
Fields that do not appear in the configuration file can be cleared by
setting their values to null and then applying the configuration file.
For fields defaulted by the server, this triggers re-defaulting
the values.
These are the only methods you should use to change an individual object field:
kubectl apply.kubectl scale.Add the field to the configuration file. For the field, discontinue direct updates to
the live configuration that do not go through kubectl apply.
As of Kubernetes 1.5, changing ownership of a field from a configuration file to an imperative writer requires manual steps:
kubectl.kubernetes.io/last-applied-configuration annotation on the live object.Kubernetes objects should be managed using only one method at a time. Switching from one method to another is possible, but is a manual process.
Migrating from imperative command management to declarative object configuration involves several manual steps:
Export the live object to a local configuration file:
kubectl get <kind>/<name> -o yaml > <kind>_<name>.yaml
Manually remove the status field from the configuration file.
kubectl apply does not update the status field
even if it is present in the configuration file.
Set the kubectl.kubernetes.io/last-applied-configuration annotation on the object:
kubectl replace --save-config -f <kind>_<name>.yaml
Change processes to use kubectl apply for managing the object exclusively.
Set the kubectl.kubernetes.io/last-applied-configuration annotation on the object:
kubectl replace --save-config -f <kind>_<name>.yaml
Change processes to use kubectl apply for managing the object exclusively.
The recommended approach is to define a single, immutable PodTemplate label used only by the controller selector with no other semantic meaning.
Example:
selector:
matchLabels:
controller-selector: "apps/v1/deployment/nginx"
template:
metadata:
labels:
controller-selector: "apps/v1/deployment/nginx"
Kustomize is a standalone tool to customize Kubernetes objects through a kustomization file.
Since 1.14, Kubectl also supports the management of Kubernetes objects using a kustomization file. To view Resources found in a directory containing a kustomization file, run the following command:
kubectl kustomize <kustomization_directory>
To apply those Resources, run kubectl apply with --kustomize or -k flag:
kubectl apply -k <kustomization_directory>
Install kubectl.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Kustomize is a tool for customizing Kubernetes configurations. It has the following features to manage application configuration files:
ConfigMaps and Secrets hold configuration or sensitive data that are used by other Kubernetes objects, such as Pods. The source of truth of ConfigMaps or Secrets are usually external to a cluster, such as a .properties file or an SSH keyfile.
Kustomize has secretGenerator and configMapGenerator, which generate Secret and ConfigMap from files or literals.
To generate a ConfigMap from a file, add an entry to the files list in configMapGenerator. Here is an example of generating a ConfigMap with a data item from a .properties file:
# Create a application.properties file
cat <<EOF >application.properties
FOO=Bar
EOF
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: example-configmap-1
files:
- application.properties
EOF
The generated ConfigMap can be examined with the following command:
kubectl kustomize ./
The generated ConfigMap is:
apiVersion: v1
data:
application.properties: |
FOO=Bar
kind: ConfigMap
metadata:
name: example-configmap-1-8mbdf7882g
To generate a ConfigMap from an env file, add an entry to the envs list in configMapGenerator. Here is an example of generating a ConfigMap with a data item from a .env file:
# Create a .env file
cat <<EOF >.env
FOO=Bar
EOF
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: example-configmap-1
envs:
- .env
EOF
The generated ConfigMap can be examined with the following command:
kubectl kustomize ./
The generated ConfigMap is:
apiVersion: v1
data:
FOO=Bar
kind: ConfigMap
metadata:
name: example-configmap-1-8mbdf7882g
.env file becomes a separate key in the ConfigMap that you generate. This is different from the previous example which embeds a file named .properties (and all its entries) as the value for a single key.
ConfigMaps can also be generated from literal key-value pairs. To generate a ConfigMap from a literal key-value pair, add an entry to the literals list in configMapGenerator. Here is an example of generating a ConfigMap with a data item from a key-value pair:
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: example-configmap-2
literals:
- FOO=Bar
EOF
The generated ConfigMap can be checked by the following command:
kubectl kustomize ./
The generated ConfigMap is:
apiVersion: v1
data:
FOO: Bar
kind: ConfigMap
metadata:
name: example-configmap-2-g2hdhfc6tk
To use a generated ConfigMap in a Deployment, reference it by the name of the configMapGenerator. Kustomize will automatically replace this name with the generated name.
This is an example deployment that uses a generated ConfigMap:
# Create a application.properties file
cat <<EOF >application.properties
FOO=Bar
EOF
cat <<EOF >deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-app
labels:
app: my-app
spec:
selector:
matchLabels:
app: my-app
template:
metadata:
labels:
app: my-app
spec:
containers:
- name: app
image: my-app
volumeMounts:
- name: config
mountPath: /config
volumes:
- name: config
configMap:
name: example-configmap-1
EOF
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
configMapGenerator:
- name: example-configmap-1
files:
- application.properties
EOF
Generate the ConfigMap and Deployment:
kubectl kustomize ./
The generated Deployment will refer to the generated ConfigMap by name:
apiVersion: v1
data:
application.properties: |
FOO=Bar
kind: ConfigMap
metadata:
name: example-configmap-1-g4hk9g2ff8
---
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: my-app
name: my-app
spec:
selector:
matchLabels:
app: my-app
template:
metadata:
labels:
app: my-app
spec:
containers:
- image: my-app
name: app
volumeMounts:
- mountPath: /config
name: config
volumes:
- configMap:
name: example-configmap-1-g4hk9g2ff8
name: config
You can generate Secrets from files or literal key-value pairs. To generate a Secret from a file, add an entry to the files list in secretGenerator. Here is an example of generating a Secret with a data item from a file:
# Create a password.txt file
cat <<EOF >./password.txt
username=admin
password=secret
EOF
cat <<EOF >./kustomization.yaml
secretGenerator:
- name: example-secret-1
files:
- password.txt
EOF
The generated Secret is as follows:
apiVersion: v1
data:
password.txt: dXNlcm5hbWU9YWRtaW4KcGFzc3dvcmQ9c2VjcmV0Cg==
kind: Secret
metadata:
name: example-secret-1-t2kt65hgtb
type: Opaque
To generate a Secret from a literal key-value pair, add an entry to literals list in secretGenerator. Here is an example of generating a Secret with a data item from a key-value pair:
cat <<EOF >./kustomization.yaml
secretGenerator:
- name: example-secret-2
literals:
- username=admin
- password=secret
EOF
The generated Secret is as follows:
apiVersion: v1
data:
password: c2VjcmV0
username: YWRtaW4=
kind: Secret
metadata:
name: example-secret-2-t52t6g96d8
type: Opaque
Like ConfigMaps, generated Secrets can be used in Deployments by refering to the name of the secretGenerator:
# Create a password.txt file
cat <<EOF >./password.txt
username=admin
password=secret
EOF
cat <<EOF >deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-app
labels:
app: my-app
spec:
selector:
matchLabels:
app: my-app
template:
metadata:
labels:
app: my-app
spec:
containers:
- name: app
image: my-app
volumeMounts:
- name: password
mountPath: /secrets
volumes:
- name: password
secret:
secretName: example-secret-1
EOF
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
secretGenerator:
- name: example-secret-1
files:
- password.txt
EOF
The generated ConfigMaps and Secrets have a content hash suffix appended. This ensures that a new ConfigMap or Secret is generated when the contents are changed. To disable the behavior of appending a suffix, one can use generatorOptions. Besides that, it is also possible to specify cross-cutting options for generated ConfigMaps and Secrets.
cat <<EOF >./kustomization.yaml
configMapGenerator:
- name: example-configmap-3
literals:
- FOO=Bar
generatorOptions:
disableNameSuffixHash: true
labels:
type: generated
annotations:
note: generated
EOF
Runkubectl kustomize ./ to view the generated ConfigMap:
apiVersion: v1
data:
FOO: Bar
kind: ConfigMap
metadata:
annotations:
note: generated
labels:
type: generated
name: example-configmap-3
It is quite common to set cross-cutting fields for all Kubernetes resources in a project. Some use cases for setting cross-cutting fields:
Here is an example:
# Create a deployment.yaml
cat <<EOF >./deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
labels:
app: nginx
spec:
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx
EOF
cat <<EOF >./kustomization.yaml
namespace: my-namespace
namePrefix: dev-
nameSuffix: "-001"
commonLabels:
app: bingo
commonAnnotations:
oncallPager: 800-555-1212
resources:
- deployment.yaml
EOF
Run kubectl kustomize ./ to view those fields are all set in the Deployment Resource:
apiVersion: apps/v1
kind: Deployment
metadata:
annotations:
oncallPager: 800-555-1212
labels:
app: bingo
name: dev-nginx-deployment-001
namespace: my-namespace
spec:
selector:
matchLabels:
app: bingo
template:
metadata:
annotations:
oncallPager: 800-555-1212
labels:
app: bingo
spec:
containers:
- image: nginx
name: nginx
It is common to compose a set of Resources in a project and manage them inside the same file or directory. Kustomize offers composing Resources from different files and applying patches or other customization to them.
Kustomize supports composition of different resources. The resources field, in the kustomization.yaml file, defines the list of resources to include in a configuration. Set the path to a resource's configuration file in the resources list.
Here is an example of an NGINX application comprised of a Deployment and a Service:
# Create a deployment.yaml file
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
ports:
- containerPort: 80
EOF
# Create a service.yaml file
cat <<EOF > service.yaml
apiVersion: v1
kind: Service
metadata:
name: my-nginx
labels:
run: my-nginx
spec:
ports:
- port: 80
protocol: TCP
selector:
run: my-nginx
EOF
# Create a kustomization.yaml composing them
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
- service.yaml
EOF
The Resources from kubectl kustomize ./ contain both the Deployment and the Service objects.
Patches can be used to apply different customizations to Resources. Kustomize supports different patching
mechanisms through patchesStrategicMerge and patchesJson6902. patchesStrategicMerge is a list of file paths. Each file should be resolved to a strategic merge patch. The names inside the patches must match Resource names that are already loaded. Small patches that do one thing are recommended. For example, create one patch for increasing the deployment replica number and another patch for setting the memory limit.
# Create a deployment.yaml file
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
ports:
- containerPort: 80
EOF
# Create a patch increase_replicas.yaml
cat <<EOF > increase_replicas.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
replicas: 3
EOF
# Create another patch set_memory.yaml
cat <<EOF > set_memory.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
template:
spec:
containers:
- name: my-nginx
resources:
limits:
memory: 512Mi
EOF
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
patchesStrategicMerge:
- increase_replicas.yaml
- set_memory.yaml
EOF
Run kubectl kustomize ./ to view the Deployment:
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
replicas: 3
selector:
matchLabels:
run: my-nginx
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- image: nginx
name: my-nginx
ports:
- containerPort: 80
resources:
limits:
memory: 512Mi
Not all Resources or fields support strategic merge patches. To support modifying arbitrary fields in arbitrary Resources,
Kustomize offers applying JSON patch through patchesJson6902.
To find the correct Resource for a Json patch, the group, version, kind and name of that Resource need to be
specified in kustomization.yaml. For example, increasing the replica number of a Deployment object can also be done
through patchesJson6902.
# Create a deployment.yaml file
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
ports:
- containerPort: 80
EOF
# Create a json patch
cat <<EOF > patch.yaml
- op: replace
path: /spec/replicas
value: 3
EOF
# Create a kustomization.yaml
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
patchesJson6902:
- target:
group: apps
version: v1
kind: Deployment
name: my-nginx
path: patch.yaml
EOF
Run kubectl kustomize ./ to see the replicas field is updated:
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
replicas: 3
selector:
matchLabels:
run: my-nginx
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- image: nginx
name: my-nginx
ports:
- containerPort: 80
In addition to patches, Kustomize also offers customizing container images or injecting field values from other objects into containers
without creating patches. For example, you can change the image used inside containers by specifying the new image in images field in kustomization.yaml.
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
ports:
- containerPort: 80
EOF
cat <<EOF >./kustomization.yaml
resources:
- deployment.yaml
images:
- name: nginx
newName: my.image.registry/nginx
newTag: 1.4.0
EOF
Run kubectl kustomize ./ to see that the image being used is updated:
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
replicas: 2
selector:
matchLabels:
run: my-nginx
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- image: my.image.registry/nginx:1.4.0
name: my-nginx
ports:
- containerPort: 80
Sometimes, the application running in a Pod may need to use configuration values from other objects. For example,
a Pod from a Deployment object need to read the corresponding Service name from Env or as a command argument.
Since the Service name may change as namePrefix or nameSuffix is added in the kustomization.yaml file. It is
not recommended to hard code the Service name in the command argument. For this usage, Kustomize can inject the Service name into containers through vars.
# Create a deployment.yaml file
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
command: ["start", "--host", "$(MY_SERVICE_NAME)"]
EOF
# Create a service.yaml file
cat <<EOF > service.yaml
apiVersion: v1
kind: Service
metadata:
name: my-nginx
labels:
run: my-nginx
spec:
ports:
- port: 80
protocol: TCP
selector:
run: my-nginx
EOF
cat <<EOF >./kustomization.yaml
namePrefix: dev-
nameSuffix: "-001"
resources:
- deployment.yaml
- service.yaml
vars:
- name: MY_SERVICE_NAME
objref:
kind: Service
name: my-nginx
apiVersion: v1
EOF
Run kubectl kustomize ./ to see that the Service name injected into containers is dev-my-nginx-001:
apiVersion: apps/v1
kind: Deployment
metadata:
name: dev-my-nginx-001
spec:
replicas: 2
selector:
matchLabels:
run: my-nginx
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- command:
- start
- --host
- dev-my-nginx-001
image: nginx
name: my-nginx
Kustomize has the concepts of bases and overlays. A base is a directory with a kustomization.yaml, which contains a
set of resources and associated customization. A base could be either a local directory or a directory from a remote repo,
as long as a kustomization.yaml is present inside. An overlay is a directory with a kustomization.yaml that refers to other
kustomization directories as its bases. A base has no knowledge of an overlay and can be used in multiple overlays.
An overlay may have multiple bases and it composes all resources
from bases and may also have customization on top of them.
Here is an example of a base:
# Create a directory to hold the base
mkdir base
# Create a base/deployment.yaml
cat <<EOF > base/deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
EOF
# Create a base/service.yaml file
cat <<EOF > base/service.yaml
apiVersion: v1
kind: Service
metadata:
name: my-nginx
labels:
run: my-nginx
spec:
ports:
- port: 80
protocol: TCP
selector:
run: my-nginx
EOF
# Create a base/kustomization.yaml
cat <<EOF > base/kustomization.yaml
resources:
- deployment.yaml
- service.yaml
EOF
This base can be used in multiple overlays. You can add different namePrefix or other cross-cutting fields
in different overlays. Here are two overlays using the same base.
mkdir dev
cat <<EOF > dev/kustomization.yaml
bases:
- ../base
namePrefix: dev-
EOF
mkdir prod
cat <<EOF > prod/kustomization.yaml
bases:
- ../base
namePrefix: prod-
EOF
Use --kustomize or -k in kubectl commands to recognize Resources managed by kustomization.yaml.
Note that -k should point to a kustomization directory, such as
kubectl apply -k <kustomization directory>/
Given the following kustomization.yaml,
# Create a deployment.yaml file
cat <<EOF > deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: my-nginx
spec:
selector:
matchLabels:
run: my-nginx
replicas: 2
template:
metadata:
labels:
run: my-nginx
spec:
containers:
- name: my-nginx
image: nginx
ports:
- containerPort: 80
EOF
# Create a kustomization.yaml
cat <<EOF >./kustomization.yaml
namePrefix: dev-
commonLabels:
app: my-nginx
resources:
- deployment.yaml
EOF
Run the following command to apply the Deployment object dev-my-nginx:
> kubectl apply -k ./
deployment.apps/dev-my-nginx created
Run one of the following commands to view the Deployment object dev-my-nginx:
kubectl get -k ./
kubectl describe -k ./
Run the following command to compare the Deployment object dev-my-nginx against the state that the cluster would be in if the manifest was applied:
kubectl diff -k ./
Run the following command to delete the Deployment object dev-my-nginx:
> kubectl delete -k ./
deployment.apps "dev-my-nginx" deleted
| Field | Type | Explanation |
|---|---|---|
| namespace | string | add namespace to all resources |
| namePrefix | string | value of this field is prepended to the names of all resources |
| nameSuffix | string | value of this field is appended to the names of all resources |
| commonLabels | map[string]string | labels to add to all resources and selectors |
| commonAnnotations | map[string]string | annotations to add to all resources |
| resources | []string | each entry in this list must resolve to an existing resource configuration file |
| configMapGenerator | []ConfigMapArgs | Each entry in this list generates a ConfigMap |
| secretGenerator | []SecretArgs | Each entry in this list generates a Secret |
| generatorOptions | GeneratorOptions | Modify behaviors of all ConfigMap and Secret generator |
| bases | []string | Each entry in this list should resolve to a directory containing a kustomization.yaml file |
| patchesStrategicMerge | []string | Each entry in this list should resolve a strategic merge patch of a Kubernetes object |
| patchesJson6902 | []Patch | Each entry in this list should resolve to a Kubernetes object and a Json Patch |
| vars | []Var | Each entry is to capture text from one resource's field |
| images | []Image | Each entry is to modify the name, tags and/or digest for one image without creating patches |
| configurations | []string | Each entry in this list should resolve to a file containing Kustomize transformer configurations |
| crds | []string | Each entry in this list should resolve to an OpenAPI definition file for Kubernetes types |
Kubernetes objects can quickly be created, updated, and deleted directly using
imperative commands built into the kubectl command-line tool. This document
explains how those commands are organized and how to use them to manage live objects.
Install kubectl.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The kubectl tool supports three kinds of object management:
See Kubernetes Object Management for a discussion of the advantages and disadvantage of each kind of object management.
The kubectl tool supports verb-driven commands for creating some of the most common
object types. The commands are named to be recognizable to users unfamiliar with
the Kubernetes object types.
run: Create a new Pod to run a Container.expose: Create a new Service object to load balance traffic across Pods.autoscale: Create a new Autoscaler object to automatically horizontally scale a controller, such as a Deployment.The kubectl tool also supports creation commands driven by object type.
These commands support more object types and are more explicit about
their intent, but require users to know the type of objects they intend
to create.
create <objecttype> [<subtype>] <instancename>Some objects types have subtypes that you can specify in the create command.
For example, the Service object has several subtypes including ClusterIP,
LoadBalancer, and NodePort. Here's an example that creates a Service with
subtype NodePort:
kubectl create service nodeport <myservicename>
In the preceding example, the create service nodeport command is called
a subcommand of the create service command.
You can use the -h flag to find the arguments and flags supported by
a subcommand:
kubectl create service nodeport -h
The kubectl command supports verb-driven commands for some common update operations.
These commands are named to enable users unfamiliar with Kubernetes
objects to perform updates without knowing the specific fields
that must be set:
scale: Horizontally scale a controller to add or remove Pods by updating the replica count of the controller.annotate: Add or remove an annotation from an object.label: Add or remove a label from an object.The kubectl command also supports update commands driven by an aspect of the object.
Setting this aspect may set different fields for different object types:
set <field>: Set an aspect of an object.The kubectl tool supports these additional ways to update a live object directly,
however they require a better understanding of the Kubernetes object schema.
edit: Directly edit the raw configuration of a live object by opening its configuration in an editor.patch: Directly modify specific fields of a live object by using a patch string.
For more details on patch strings, see the patch section in
API Conventions.You can use the delete command to delete an object from a cluster:
delete <type>/<name>kubectl delete for both imperative commands and imperative object
configuration. The difference is in the arguments passed to the command. To use
kubectl delete as an imperative command, pass the object to be deleted as
an argument. Here's an example that passes a Deployment object named nginx:
kubectl delete deployment/nginx
There are several commands for printing information about an object:
get: Prints basic information about matching objects. Use get -h to see a list of options.describe: Prints aggregated detailed information about matching objects.logs: Prints the stdout and stderr for a container running in a Pod.set commands to modify objects before creationThere are some object fields that don't have a flag you can use
in a create command. In some of those cases, you can use a combination of
set and create to specify a value for the field before object
creation. This is done by piping the output of the create command to the
set command, and then back to the create command. Here's an example:
kubectl create service clusterip my-svc --clusterip="None" -o yaml --dry-run=client | kubectl set selector --local -f - 'environment=qa' -o yaml | kubectl create -f -
kubectl create service -o yaml --dry-run=client command creates the configuration for the Service, but prints it to stdout as YAML instead of sending it to the Kubernetes API server.kubectl set selector --local -f - -o yaml command reads the configuration from stdin, and writes the updated configuration to stdout as YAML.kubectl create -f - command creates the object using the configuration provided via stdin.--edit to modify objects before creationYou can use kubectl create --edit to make arbitrary changes to an object
before it is created. Here's an example:
kubectl create service clusterip my-svc --clusterip="None" -o yaml --dry-run=client > /tmp/srv.yaml
kubectl create --edit -f /tmp/srv.yaml
kubectl create service command creates the configuration for the Service and saves it to /tmp/srv.yaml.kubectl create --edit command opens the configuration file for editing before it creates the object.Kubernetes objects can be created, updated, and deleted by using the kubectl
command-line tool along with an object configuration file written in YAML or JSON.
This document explains how to define and manage objects using configuration files.
Install kubectl.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The kubectl tool supports three kinds of object management:
See Kubernetes Object Management for a discussion of the advantages and disadvantage of each kind of object management.
You can use kubectl create -f to create an object from a configuration file.
Refer to the kubernetes API reference
for details.
kubectl create -f <filename|url>replace command drops all
parts of the spec not specified in the configuration file. This
should not be used with objects whose specs are partially managed
by the cluster, such as Services of type LoadBalancer, where
the externalIPs field is managed independently from the configuration
file. Independently managed fields must be copied to the configuration
file to prevent replace from dropping them.
You can use kubectl replace -f to update a live object according to a
configuration file.
kubectl replace -f <filename|url>You can use kubectl delete -f to delete an object that is described in a
configuration file.
kubectl delete -f <filename|url>If configuration file has specified the generateName field in the metadata
section instead of the name field, you cannot delete the object using
kubectl delete -f <filename|url>.
You will have to use other flags for deleting the object. For example:
kubectl delete <type> <name>
kubectl delete <type> -l <label>
You can use kubectl get -f to view information about an object that is
described in a configuration file.
kubectl get -f <filename|url> -o yamlThe -o yaml flag specifies that the full object configuration is printed.
Use kubectl get -h to see a list of options.
The create, replace, and delete commands work well when each object's
configuration is fully defined and recorded in its configuration
file. However when a live object is updated, and the updates are not merged
into its configuration file, the updates will be lost the next time a replace
is executed. This can happen if a controller, such as
a HorizontalPodAutoscaler, makes updates directly to a live object. Here's
an example:
If you need to support multiple writers to the same object, you can use
kubectl apply to manage the object.
Suppose you have the URL of an object configuration file. You can use
kubectl create --edit to make changes to the configuration before the
object is created. This is particularly useful for tutorials and tasks
that point to a configuration file that could be modified by the reader.
kubectl create -f <url> --edit
Migrating from imperative commands to imperative object configuration involves several manual steps.
Export the live object to a local object configuration file:
kubectl get <kind>/<name> -o yaml > <kind>_<name>.yaml
Manually remove the status field from the object configuration file.
For subsequent object management, use replace exclusively.
kubectl replace -f <kind>_<name>.yaml
The recommended approach is to define a single, immutable PodTemplate label used only by the controller selector with no other semantic meaning.
Example label:
selector:
matchLabels:
controller-selector: "apps/v1/deployment/nginx"
template:
metadata:
labels:
controller-selector: "apps/v1/deployment/nginx"
This task shows how to use kubectl patch to update an API object in place. The exercises
in this task demonstrate a strategic merge patch and a JSON merge patch.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Here's the configuration file for a Deployment that has two replicas. Each replica is a Pod that has one container:
apiVersion: apps/v1
kind: Deployment
metadata:
name: patch-demo
spec:
replicas: 2
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: patch-demo-ctr
image: nginx
tolerations:
- effect: NoSchedule
key: dedicated
value: test-team
Create the Deployment:
kubectl apply -f https://k8s.io/examples/application/deployment-patch.yaml
View the Pods associated with your Deployment:
kubectl get pods
The output shows that the Deployment has two Pods. The 1/1 indicates that
each Pod has one container:
NAME READY STATUS RESTARTS AGE
patch-demo-28633765-670qr 1/1 Running 0 23s
patch-demo-28633765-j5qs3 1/1 Running 0 23s
Make a note of the names of the running Pods. Later, you will see that these Pods get terminated and replaced by new ones.
At this point, each Pod has one Container that runs the nginx image. Now suppose you want each Pod to have two containers: one that runs nginx and one that runs redis.
Create a file named patch-file.yaml that has this content:
spec:
template:
spec:
containers:
- name: patch-demo-ctr-2
image: redis
Patch your Deployment:
kubectl patch deployment patch-demo --patch "$(cat patch-file.yaml)"
kubectl patch deployment patch-demo --patch $(Get-Content patch-file.yaml -Raw)
View the patched Deployment:
kubectl get deployment patch-demo --output yaml
The output shows that the PodSpec in the Deployment has two Containers:
containers:
- image: redis
imagePullPolicy: Always
name: patch-demo-ctr-2
...
- image: nginx
imagePullPolicy: Always
name: patch-demo-ctr
...
View the Pods associated with your patched Deployment:
kubectl get pods
The output shows that the running Pods have different names from the Pods that
were running previously. The Deployment terminated the old Pods and created two
new Pods that comply with the updated Deployment spec. The 2/2 indicates that
each Pod has two Containers:
NAME READY STATUS RESTARTS AGE
patch-demo-1081991389-2wrn5 2/2 Running 0 1m
patch-demo-1081991389-jmg7b 2/2 Running 0 1m
Take a closer look at one of the patch-demo Pods:
kubectl get pod <your-pod-name> --output yaml
The output shows that the Pod has two Containers: one running nginx and one running redis:
containers:
- image: redis
...
- image: nginx
...
The patch you did in the preceding exercise is called a strategic merge patch.
Notice that the patch did not replace the containers list. Instead it added a new
Container to the list. In other words, the list in the patch was merged with the
existing list. This is not always what happens when you use a strategic merge patch on a list.
In some cases, the list is replaced, not merged.
With a strategic merge patch, a list is either replaced or merged depending on its
patch strategy. The patch strategy is specified by the value of the patchStrategy key
in a field tag in the Kubernetes source code. For example, the Containers field of PodSpec
struct has a patchStrategy of merge:
type PodSpec struct {
...
Containers []Container `json:"containers" patchStrategy:"merge" patchMergeKey:"name" ...`
You can also see the patch strategy in the OpenApi spec:
"io.k8s.api.core.v1.PodSpec": {
...
"containers": {
"description": "List of containers belonging to the pod. ...
},
"x-kubernetes-patch-merge-key": "name",
"x-kubernetes-patch-strategy": "merge"
},
And you can see the patch strategy in the Kubernetes API documentation.
Create a file named patch-file-tolerations.yaml that has this content:
spec:
template:
spec:
tolerations:
- effect: NoSchedule
key: disktype
value: ssd
Patch your Deployment:
kubectl patch deployment patch-demo --patch "$(cat patch-file-tolerations.yaml)"
View the patched Deployment:
kubectl get deployment patch-demo --output yaml
The output shows that the PodSpec in the Deployment has only one Toleration:
tolerations:
- effect: NoSchedule
key: disktype
value: ssd
Notice that the tolerations list in the PodSpec was replaced, not merged. This is because
the Tolerations field of PodSpec does not have a patchStrategy key in its field tag. So the
strategic merge patch uses the default patch strategy, which is replace.
type PodSpec struct {
...
Tolerations []Toleration `json:"tolerations,omitempty" protobuf:"bytes,22,opt,name=tolerations"`
A strategic merge patch is different from a JSON merge patch. With a JSON merge patch, if you want to update a list, you have to specify the entire new list. And the new list completely replaces the existing list.
The kubectl patch command has a type parameter that you can set to one of these values:
| Parameter value | Merge type |
|---|---|
| json | JSON Patch, RFC 6902 |
| merge | JSON Merge Patch, RFC 7386 |
| strategic | Strategic merge patch |
For a comparison of JSON patch and JSON merge patch, see JSON Patch and JSON Merge Patch.
The default value for the type parameter is strategic. So in the preceding exercise, you
did a strategic merge patch.
Next, do a JSON merge patch on your same Deployment. Create a file named patch-file-2.yaml
that has this content:
spec:
template:
spec:
containers:
- name: patch-demo-ctr-3
image: gcr.io/google-samples/node-hello:1.0
In your patch command, set type to merge:
kubectl patch deployment patch-demo --type merge --patch "$(cat patch-file-2.yaml)"
View the patched Deployment:
kubectl get deployment patch-demo --output yaml
The containers list that you specified in the patch has only one Container.
The output shows that your list of one Container replaced the existing containers list.
spec:
containers:
- image: gcr.io/google-samples/node-hello:1.0
...
name: patch-demo-ctr-3
List the running Pods:
kubectl get pods
In the output, you can see that the existing Pods were terminated, and new Pods
were created. The 1/1 indicates that each new Pod is running only one Container.
NAME READY STATUS RESTARTS AGE
patch-demo-1307768864-69308 1/1 Running 0 1m
patch-demo-1307768864-c86dc 1/1 Running 0 1m
Here's the configuration file for a Deployment that uses the RollingUpdate strategy:
apiVersion: apps/v1
kind: Deployment
metadata:
name: retainkeys-demo
spec:
selector:
matchLabels:
app: nginx
strategy:
rollingUpdate:
maxSurge: 30%
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: retainkeys-demo-ctr
image: nginx
Create the deployment:
kubectl apply -f https://k8s.io/examples/application/deployment-retainkeys.yaml
At this point, the deployment is created and is using the RollingUpdate strategy.
Create a file named patch-file-no-retainkeys.yaml that has this content:
spec:
strategy:
type: Recreate
Patch your Deployment:
kubectl patch deployment retainkeys-demo --type merge --patch "$(cat patch-file-no-retainkeys.yaml)"
kubectl patch deployment retainkeys-demo --type merge --patch $(Get-Content patch-file-no-retainkeys.yaml -Raw)
In the output, you can see that it is not possible to set type as Recreate when a value is defined for spec.strategy.rollingUpdate:
The Deployment "retainkeys-demo" is invalid: spec.strategy.rollingUpdate: Forbidden: may not be specified when strategy `type` is 'Recreate'
The way to remove the value for spec.strategy.rollingUpdate when updating the value for type is to use the retainKeys strategy for the strategic merge.
Create another file named patch-file-retainkeys.yaml that has this content:
spec:
strategy:
$retainKeys:
- type
type: Recreate
With this patch, we indicate that we want to retain only the type key of the strategy object. Thus, the rollingUpdate will be removed during the patch operation.
Patch your Deployment again with this new patch:
kubectl patch deployment retainkeys-demo --type merge --patch "$(cat patch-file-retainkeys.yaml)"
kubectl patch deployment retainkeys-demo --type merge --patch $(Get-Content patch-file-retainkeys.yaml -Raw)
Examine the content of the Deployment:
kubectl get deployment retainkeys-demo --output yaml
The output shows that the strategy object in the Deployment does not contain the rollingUpdate key anymore:
spec:
strategy:
type: Recreate
template:
The patch you did in the preceding exercise is called a strategic merge patch with retainKeys strategy. This method introduces a new directive $retainKeys that has the following strategies:
$retainKeys list.$retainKeys list must be a superset or the same as the fields present in the patch.The retainKeys strategy does not work for all objects. It only works when the value of the patchStrategy key in a field tag in the Kubernetes source code contains retainKeys. For example, the Strategy field of the DeploymentSpec struct has a patchStrategy of retainKeys:
type DeploymentSpec struct {
...
// +patchStrategy=retainKeys
Strategy DeploymentStrategy `json:"strategy,omitempty" patchStrategy:"retainKeys" ...`
You can also see the retainKeys strategy in the OpenApi spec:
"io.k8s.api.apps.v1.DeploymentSpec": {
...
"strategy": {
"$ref": "#/definitions/io.k8s.api.apps.v1.DeploymentStrategy",
"description": "The deployment strategy to use to replace existing pods with new ones.",
"x-kubernetes-patch-strategy": "retainKeys"
},
And you can see the retainKeys strategy in the
Kubernetes API documentation.
The kubectl patch command takes YAML or JSON. It can take the patch as a file or
directly on the command line.
Create a file named patch-file.json that has this content:
{
"spec": {
"template": {
"spec": {
"containers": [
{
"name": "patch-demo-ctr-2",
"image": "redis"
}
]
}
}
}
}
The following commands are equivalent:
kubectl patch deployment patch-demo --patch "$(cat patch-file.yaml)"
kubectl patch deployment patch-demo --patch 'spec:\n template:\n spec:\n containers:\n - name: patch-demo-ctr-2\n image: redis'
kubectl patch deployment patch-demo --patch "$(cat patch-file.json)"
kubectl patch deployment patch-demo --patch '{"spec": {"template": {"spec": {"containers": [{"name": "patch-demo-ctr-2","image": "redis"}]}}}}'
In this exercise, you used kubectl patch to change the live configuration
of a Deployment object. You did not change the configuration file that you originally used to
create the Deployment object. Other commands for updating API objects include
kubectl annotate,
kubectl edit,
kubectl replace,
kubectl scale,
and
kubectl apply.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
A Secret can contain user credentials required by pods to access a database.
For example, a database connection string consists of a username and password.
You can store the username in a file ./username.txt and the password in a
file ./password.txt on your local machine.
echo -n 'admin' > ./username.txt
echo -n '1f2d1e2e67df' > ./password.txt
In these commands, the -n flag ensures that the generated files do not have
an extra newline character at the end of the text. This is important because
when kubectl reads a file and encodes the content into a base64 string, the
extra newline character gets encoded too.
The kubectl create secret command packages these files into a Secret and creates
the object on the API server.
kubectl create secret generic db-user-pass \
--from-file=./username.txt \
--from-file=./password.txt
The output is similar to:
secret/db-user-pass created
The default key name is the filename. You can optionally set the key name using
--from-file=[key=]source. For example:
kubectl create secret generic db-user-pass \
--from-file=username=./username.txt \
--from-file=password=./password.txt
You do not need to escape special characters in password strings that you include in a file.
You can also provide Secret data using the --from-literal=<key>=<value> tag.
This tag can be specified more than once to provide multiple key-value pairs.
Note that special characters such as $, \, *, =, and ! will be
interpreted by your shell
and require escaping.
In most shells, the easiest way to escape the password is to surround it with
single quotes ('). For example, if your password is S!B\*d$zDsb=,
run the following command:
kubectl create secret generic db-user-pass \
--from-literal=username=devuser \
--from-literal=password='S!B\*d$zDsb='
Check that the Secret was created:
kubectl get secrets
The output is similar to:
NAME TYPE DATA AGE
db-user-pass Opaque 2 51s
You can view a description of the Secret:
kubectl describe secrets/db-user-pass
The output is similar to:
Name: db-user-pass
Namespace: default
Labels: <none>
Annotations: <none>
Type: Opaque
Data
====
password: 12 bytes
username: 5 bytes
The commands kubectl get and kubectl describe avoid showing the contents
of a Secret by default. This is to protect the Secret from being exposed
accidentally, or from being stored in a terminal log.
To view the contents of the Secret you created, run the following command:
kubectl get secret db-user-pass -o jsonpath='{.data}'
The output is similar to:
{"password":"MWYyZDFlMmU2N2Rm","username":"YWRtaW4="}
Now you can decode the password data:
echo 'MWYyZDFlMmU2N2Rm' | base64 --decode
The output is similar to:
1f2d1e2e67df
Delete the Secret you created:
kubectl delete secret db-user-pass
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
You can create a Secret in a file first, in JSON or YAML format, and then
create that object. The
Secret
resource contains two maps: data and stringData.
The data field is used to store arbitrary data, encoded using base64. The
stringData field is provided for convenience, and it allows you to provide
Secret data as unencoded strings.
The keys of data and stringData must consist of alphanumeric characters,
-, _ or ..
For example, to store two strings in a Secret using the data field, convert
the strings to base64 as follows:
echo -n 'admin' | base64
The output is similar to:
YWRtaW4=
echo -n '1f2d1e2e67df' | base64
The output is similar to:
MWYyZDFlMmU2N2Rm
Write a Secret config file that looks like this:
apiVersion: v1
kind: Secret
metadata:
name: mysecret
type: Opaque
data:
username: YWRtaW4=
password: MWYyZDFlMmU2N2Rm
Note that the name of a Secret object must be a valid DNS subdomain name.
base64 utility on Darwin/macOS, users should avoid using the -b
option to split long lines. Conversely, Linux users should add the option
-w 0 to base64 commands or the pipeline base64 | tr -d '\n' if the -w
option is not available.
For certain scenarios, you may wish to use the stringData field instead. This
field allows you to put a non-base64 encoded string directly into the Secret,
and the string will be encoded for you when the Secret is created or updated.
A practical example of this might be where you are deploying an application that uses a Secret to store a configuration file, and you want to populate parts of that configuration file during your deployment process.
For example, if your application uses the following configuration file:
apiUrl: "https://my.api.com/api/v1"
username: "<user>"
password: "<password>"
You could store this in a Secret using the following definition:
apiVersion: v1
kind: Secret
metadata:
name: mysecret
type: Opaque
stringData:
config.yaml: |
apiUrl: "https://my.api.com/api/v1"
username: <user>
password: <password>
Now create the Secret using kubectl apply:
kubectl apply -f ./secret.yaml
The output is similar to:
secret/mysecret created
The stringData field is a write-only convenience field. It is never output when
retrieving Secrets. For example, if you run the following command:
kubectl get secret mysecret -o yaml
The output is similar to:
apiVersion: v1
data:
config.yaml: YXBpVXJsOiAiaHR0cHM6Ly9teS5hcGkuY29tL2FwaS92MSIKdXNlcm5hbWU6IHt7dXNlcm5hbWV9fQpwYXNzd29yZDoge3twYXNzd29yZH19
kind: Secret
metadata:
creationTimestamp: 2018-11-15T20:40:59Z
name: mysecret
namespace: default
resourceVersion: "7225"
uid: c280ad2e-e916-11e8-98f2-025000000001
type: Opaque
The commands kubectl get and kubectl describe avoid showing the contents of a Secret by
default. This is to protect the Secret from being exposed accidentally to an onlooker,
or from being stored in a terminal log.
To check the actual content of the encoded data, please refer to
decoding secret.
If a field, such as username, is specified in both data and stringData,
the value from stringData is used. For example, the following Secret definition:
apiVersion: v1
kind: Secret
metadata:
name: mysecret
type: Opaque
data:
username: YWRtaW4=
stringData:
username: administrator
Results in the following Secret:
apiVersion: v1
data:
username: YWRtaW5pc3RyYXRvcg==
kind: Secret
metadata:
creationTimestamp: 2018-11-15T20:46:46Z
name: mysecret
namespace: default
resourceVersion: "7579"
uid: 91460ecb-e917-11e8-98f2-025000000001
type: Opaque
Where YWRtaW5pc3RyYXRvcg== decodes to administrator.
To delete the Secret you have created:
kubectl delete secret mysecret
kubectl commandSince Kubernetes v1.14, kubectl supports
managing objects using Kustomize.
Kustomize provides resource Generators to create Secrets and ConfigMaps. The
Kustomize generators should be specified in a kustomization.yaml file inside
a directory. After generating the Secret, you can create the Secret on the API
server with kubectl apply.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
You can generate a Secret by defining a secretGenerator in a
kustomization.yaml file that references other existing files.
For example, the following kustomization file references the
./username.txt and the ./password.txt files:
secretGenerator:
- name: db-user-pass
files:
- username.txt
- password.txt
You can also define the secretGenerator in the kustomization.yaml
file by providing some literals.
For example, the following kustomization.yaml file contains two literals
for username and password respectively:
secretGenerator:
- name: db-user-pass
literals:
- username=admin
- password=1f2d1e2e67df
You can also define the secretGenerator in the kustomization.yaml
file by providing .env files.
For example, the following kustomization.yaml file pulls in data from
.env.secret file:
secretGenerator:
- name: db-user-pass
envs:
- .env.secret
Note that in all cases, you don't need to base64 encode the values.
Apply the directory containing the kustomization.yaml to create the Secret.
kubectl apply -k .
The output is similar to:
secret/db-user-pass-96mffmfh4k created
Note that when a Secret is generated, the Secret name is created by hashing the Secret data and appending the hash value to the name. This ensures that a new Secret is generated each time the data is modified.
You can check that the secret was created:
kubectl get secrets
The output is similar to:
NAME TYPE DATA AGE
db-user-pass-96mffmfh4k Opaque 2 51s
You can view a description of the secret:
kubectl describe secrets/db-user-pass-96mffmfh4k
The output is similar to:
Name: db-user-pass-96mffmfh4k
Namespace: default
Labels: <none>
Annotations: <none>
Type: Opaque
Data
====
password.txt: 12 bytes
username.txt: 5 bytes
The commands kubectl get and kubectl describe avoid showing the contents of a Secret by
default. This is to protect the Secret from being exposed accidentally to an onlooker,
or from being stored in a terminal log.
To check the actual content of the encoded data, please refer to
decoding secret.
To delete the Secret you have created:
kubectl delete secret db-user-pass-96mffmfh4k
kubectl commandThis page shows how to define commands and arguments when you run a container in a Pod.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
When you create a Pod, you can define a command and arguments for the
containers that run in the Pod. To define a command, include the command
field in the configuration file. To define arguments for the command, include
the args field in the configuration file. The command and arguments that
you define cannot be changed after the Pod is created.
The command and arguments that you define in the configuration file override the default command and arguments provided by the container image. If you define args, but do not define a command, the default command is used with your new arguments.
command field corresponds to entrypoint in some container
runtimes. Refer to the Notes below.
In this exercise, you create a Pod that runs one container. The configuration file for the Pod defines a command and two arguments:
apiVersion: v1
kind: Pod
metadata:
name: command-demo
labels:
purpose: demonstrate-command
spec:
containers:
- name: command-demo-container
image: debian
command: ["printenv"]
args: ["HOSTNAME", "KUBERNETES_PORT"]
restartPolicy: OnFailure
Create a Pod based on the YAML configuration file:
kubectl apply -f https://k8s.io/examples/pods/commands.yaml
List the running Pods:
kubectl get pods
The output shows that the container that ran in the command-demo Pod has completed.
To see the output of the command that ran in the container, view the logs from the Pod:
kubectl logs command-demo
The output shows the values of the HOSTNAME and KUBERNETES_PORT environment variables:
command-demo
tcp://10.3.240.1:443
In the preceding example, you defined the arguments directly by providing strings. As an alternative to providing strings directly, you can define arguments by using environment variables:
env:
- name: MESSAGE
value: "hello world"
command: ["/bin/echo"]
args: ["$(MESSAGE)"]
This means you can define an argument for a Pod using any of the techniques available for defining environment variables, including ConfigMaps and Secrets.
"$(VAR)". This is
required for the variable to be expanded in the command or args field.
In some cases, you need your command to run in a shell. For example, your command might consist of several commands piped together, or it might be a shell script. To run your command in a shell, wrap it like this:
command: ["/bin/sh"]
args: ["-c", "while true; do echo hello; sleep 10;done"]
This table summarizes the field names used by Docker and Kubernetes.
| Description | Docker field name | Kubernetes field name |
|---|---|---|
| The command run by the container | Entrypoint | command |
| The arguments passed to the command | Cmd | args |
When you override the default Entrypoint and Cmd, these rules apply:
If you do not supply command or args for a Container, the defaults defined
in the Docker image are used.
If you supply a command but no args for a Container, only the supplied
command is used. The default EntryPoint and the default Cmd defined in the Docker
image are ignored.
If you supply only args for a Container, the default Entrypoint defined in
the Docker image is run with the args that you supplied.
If you supply a command and args, the default Entrypoint and the default
Cmd defined in the Docker image are ignored. Your command is run with your
args.
Here are some examples:
| Image Entrypoint | Image Cmd | Container command | Container args | Command run |
|---|---|---|---|---|
[/ep-1] |
[foo bar] |
<not set> | <not set> | [ep-1 foo bar] |
[/ep-1] |
[foo bar] |
[/ep-2] |
<not set> | [ep-2] |
[/ep-1] |
[foo bar] |
<not set> | [zoo boo] |
[ep-1 zoo boo] |
[/ep-1] |
[foo bar] |
[/ep-2] |
[zoo boo] |
[ep-2 zoo boo] |
This page shows how to define dependent environment variables for a container in a Kubernetes Pod.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
When you create a Pod, you can set dependent environment variables for the containers that run in the Pod. To set dependent environment variables, you can use $(VAR_NAME) in the value of env in the configuration file.
In this exercise, you create a Pod that runs one container. The configuration file for the Pod defines an dependent environment variable with common usage defined. Here is the configuration manifest for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: dependent-envars-demo
spec:
containers:
- name: dependent-envars-demo
args:
- while true; do echo -en '\n'; printf UNCHANGED_REFERENCE=$UNCHANGED_REFERENCE'\n'; printf SERVICE_ADDRESS=$SERVICE_ADDRESS'\n';printf ESCAPED_REFERENCE=$ESCAPED_REFERENCE'\n'; sleep 30; done;
command:
- sh
- -c
image: busybox
env:
- name: SERVICE_PORT
value: "80"
- name: SERVICE_IP
value: "172.17.0.1"
- name: UNCHANGED_REFERENCE
value: "$(PROTOCOL)://$(SERVICE_IP):$(SERVICE_PORT)"
- name: PROTOCOL
value: "https"
- name: SERVICE_ADDRESS
value: "$(PROTOCOL)://$(SERVICE_IP):$(SERVICE_PORT)"
- name: ESCAPED_REFERENCE
value: "$$(PROTOCOL)://$(SERVICE_IP):$(SERVICE_PORT)"
Create a Pod based on that manifest:
kubectl apply -f https://k8s.io/examples/pods/inject/dependent-envars.yaml
pod/dependent-envars-demo created
List the running Pods:
kubectl get pods dependent-envars-demo
NAME READY STATUS RESTARTS AGE
dependent-envars-demo 1/1 Running 0 9s
Check the logs for the container running in your Pod:
kubectl logs pod/dependent-envars-demo
UNCHANGED_REFERENCE=$(PROTOCOL)://172.17.0.1:80
SERVICE_ADDRESS=https://172.17.0.1:80
ESCAPED_REFERENCE=$(PROTOCOL)://172.17.0.1:80
As shown above, you have defined the correct dependency reference of SERVICE_ADDRESS, bad dependency reference of UNCHANGED_REFERENCE and skip dependent references of ESCAPED_REFERENCE.
When an environment variable is already defined when being referenced,
the reference can be correctly resolved, such as in the SERVICE_ADDRESS case.
When the environment variable is undefined or only includes some variables, the undefined environment variable is treated as a normal string, such as UNCHANGED_REFERENCE. Note that incorrectly parsed environment variables, in general, will not block the container from starting.
The $(VAR_NAME) syntax can be escaped with a double $, ie: $$(VAR_NAME).
Escaped references are never expanded, regardless of whether the referenced variable
is defined or not. This can be seen from the ESCAPED_REFERENCE case above.
This page shows how to define environment variables for a container in a Kubernetes Pod.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
When you create a Pod, you can set environment variables for the containers
that run in the Pod. To set environment variables, include the env or
envFrom field in the configuration file.
In this exercise, you create a Pod that runs one container. The configuration
file for the Pod defines an environment variable with name DEMO_GREETING and
value "Hello from the environment". Here is the configuration manifest for the
Pod:
apiVersion: v1
kind: Pod
metadata:
name: envar-demo
labels:
purpose: demonstrate-envars
spec:
containers:
- name: envar-demo-container
image: gcr.io/google-samples/node-hello:1.0
env:
- name: DEMO_GREETING
value: "Hello from the environment"
- name: DEMO_FAREWELL
value: "Such a sweet sorrow"
Create a Pod based on that manifest:
kubectl apply -f https://k8s.io/examples/pods/inject/envars.yaml
List the running Pods:
kubectl get pods -l purpose=demonstrate-envars
The output is similar to:
NAME READY STATUS RESTARTS AGE
envar-demo 1/1 Running 0 9s
List the Pod's container environment variables:
kubectl exec envar-demo -- printenv
The output is similar to this:
NODE_VERSION=4.4.2
EXAMPLE_SERVICE_PORT_8080_TCP_ADDR=10.3.245.237
HOSTNAME=envar-demo
...
DEMO_GREETING=Hello from the environment
DEMO_FAREWELL=Such a sweet sorrow
env or envFrom field
override any environment variables specified in the container image.
Environment variables that you define in a Pod's configuration can be used
elsewhere in the configuration, for example in commands and arguments that
you set for the Pod's containers.
In the example configuration below, the GREETING, HONORIFIC, and
NAME environment variables are set to Warm greetings to, The Most Honorable, and Kubernetes, respectively. Those environment variables
are then used in the CLI arguments passed to the env-print-demo
container.
apiVersion: v1
kind: Pod
metadata:
name: print-greeting
spec:
containers:
- name: env-print-demo
image: bash
env:
- name: GREETING
value: "Warm greetings to"
- name: HONORIFIC
value: "The Most Honorable"
- name: NAME
value: "Kubernetes"
command: ["echo"]
args: ["$(GREETING) $(HONORIFIC) $(NAME)"]
Upon creation, the command echo Warm greetings to The Most Honorable Kubernetes is run on the container.
This page shows how a Pod can use environment variables to expose information about itself to Containers running in the Pod. Environment variables can expose Pod fields and Container fields.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
There are two ways to expose Pod and Container fields to a running Container:
Together, these two ways of exposing Pod and Container fields are called the Downward API.
In this exercise, you create a Pod that has one Container. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: dapi-envars-fieldref
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox
command: [ "sh", "-c"]
args:
- while true; do
echo -en '\n';
printenv MY_NODE_NAME MY_POD_NAME MY_POD_NAMESPACE;
printenv MY_POD_IP MY_POD_SERVICE_ACCOUNT;
sleep 10;
done;
env:
- name: MY_NODE_NAME
valueFrom:
fieldRef:
fieldPath: spec.nodeName
- name: MY_POD_NAME
valueFrom:
fieldRef:
fieldPath: metadata.name
- name: MY_POD_NAMESPACE
valueFrom:
fieldRef:
fieldPath: metadata.namespace
- name: MY_POD_IP
valueFrom:
fieldRef:
fieldPath: status.podIP
- name: MY_POD_SERVICE_ACCOUNT
valueFrom:
fieldRef:
fieldPath: spec.serviceAccountName
restartPolicy: Never
In the configuration file, you can see five environment variables. The env
field is an array of
EnvVars.
The first element in the array specifies that the MY_NODE_NAME environment
variable gets its value from the Pod's spec.nodeName field. Similarly, the
other environment variables get their names from Pod fields.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/inject/dapi-envars-pod.yaml
Verify that the Container in the Pod is running:
kubectl get pods
View the Container's logs:
kubectl logs dapi-envars-fieldref
The output shows the values of selected environment variables:
minikube
dapi-envars-fieldref
default
172.17.0.4
default
To see why these values are in the log, look at the command and args fields
in the configuration file. When the Container starts, it writes the values of
five environment variables to stdout. It repeats this every ten seconds.
Next, get a shell into the Container that is running in your Pod:
kubectl exec -it dapi-envars-fieldref -- sh
In your shell, view the environment variables:
/# printenv
The output shows that certain environment variables have been assigned the values of Pod fields:
MY_POD_SERVICE_ACCOUNT=default
...
MY_POD_NAMESPACE=default
MY_POD_IP=172.17.0.4
...
MY_NODE_NAME=minikube
...
MY_POD_NAME=dapi-envars-fieldref
In the preceding exercise, you used Pod fields as the values for environment variables. In this next exercise, you use Container fields as the values for environment variables. Here is the configuration file for a Pod that has one container:
apiVersion: v1
kind: Pod
metadata:
name: dapi-envars-resourcefieldref
spec:
containers:
- name: test-container
image: k8s.gcr.io/busybox:1.24
command: [ "sh", "-c"]
args:
- while true; do
echo -en '\n';
printenv MY_CPU_REQUEST MY_CPU_LIMIT;
printenv MY_MEM_REQUEST MY_MEM_LIMIT;
sleep 10;
done;
resources:
requests:
memory: "32Mi"
cpu: "125m"
limits:
memory: "64Mi"
cpu: "250m"
env:
- name: MY_CPU_REQUEST
valueFrom:
resourceFieldRef:
containerName: test-container
resource: requests.cpu
- name: MY_CPU_LIMIT
valueFrom:
resourceFieldRef:
containerName: test-container
resource: limits.cpu
- name: MY_MEM_REQUEST
valueFrom:
resourceFieldRef:
containerName: test-container
resource: requests.memory
- name: MY_MEM_LIMIT
valueFrom:
resourceFieldRef:
containerName: test-container
resource: limits.memory
restartPolicy: Never
In the configuration file, you can see four environment variables. The env
field is an array of
EnvVars.
The first element in the array specifies that the MY_CPU_REQUEST environment
variable gets its value from the requests.cpu field of a Container named
test-container. Similarly, the other environment variables get their values
from Container fields.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/inject/dapi-envars-container.yaml
Verify that the Container in the Pod is running:
kubectl get pods
View the Container's logs:
kubectl logs dapi-envars-resourcefieldref
The output shows the values of selected environment variables:
1
1
33554432
67108864
This page shows how a Pod can use a DownwardAPIVolumeFile to expose information about itself to Containers running in the Pod. A DownwardAPIVolumeFile can expose Pod fields and Container fields.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
There are two ways to expose Pod and Container fields to a running Container:
Together, these two ways of exposing Pod and Container fields are called the Downward API.
In this exercise, you create a Pod that has one Container. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: kubernetes-downwardapi-volume-example
labels:
zone: us-est-coast
cluster: test-cluster1
rack: rack-22
annotations:
build: two
builder: john-doe
spec:
containers:
- name: client-container
image: k8s.gcr.io/busybox
command: ["sh", "-c"]
args:
- while true; do
if [[ -e /etc/podinfo/labels ]]; then
echo -en '\n\n'; cat /etc/podinfo/labels; fi;
if [[ -e /etc/podinfo/annotations ]]; then
echo -en '\n\n'; cat /etc/podinfo/annotations; fi;
sleep 5;
done;
volumeMounts:
- name: podinfo
mountPath: /etc/podinfo
volumes:
- name: podinfo
downwardAPI:
items:
- path: "labels"
fieldRef:
fieldPath: metadata.labels
- path: "annotations"
fieldRef:
fieldPath: metadata.annotations
In the configuration file, you can see that the Pod has a downwardAPI Volume,
and the Container mounts the Volume at /etc/podinfo.
Look at the items array under downwardAPI. Each element of the array is a
DownwardAPIVolumeFile.
The first element specifies that the value of the Pod's
metadata.labels field should be stored in a file named labels.
The second element specifies that the value of the Pod's annotations
field should be stored in a file named annotations.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/inject/dapi-volume.yaml
Verify that the Container in the Pod is running:
kubectl get pods
View the Container's logs:
kubectl logs kubernetes-downwardapi-volume-example
The output shows the contents of the labels file and the annotations file:
cluster="test-cluster1"
rack="rack-22"
zone="us-est-coast"
build="two"
builder="john-doe"
Get a shell into the Container that is running in your Pod:
kubectl exec -it kubernetes-downwardapi-volume-example -- sh
In your shell, view the labels file:
/# cat /etc/podinfo/labels
The output shows that all of the Pod's labels have been written
to the labels file:
cluster="test-cluster1"
rack="rack-22"
zone="us-est-coast"
Similarly, view the annotations file:
/# cat /etc/podinfo/annotations
View the files in the /etc/podinfo directory:
/# ls -laR /etc/podinfo
In the output, you can see that the labels and annotations files
are in a temporary subdirectory: in this example,
..2982_06_02_21_47_53.299460680. In the /etc/podinfo directory, ..data is
a symbolic link to the temporary subdirectory. Also in the /etc/podinfo directory,
labels and annotations are symbolic links.
drwxr-xr-x ... Feb 6 21:47 ..2982_06_02_21_47_53.299460680
lrwxrwxrwx ... Feb 6 21:47 ..data -> ..2982_06_02_21_47_53.299460680
lrwxrwxrwx ... Feb 6 21:47 annotations -> ..data/annotations
lrwxrwxrwx ... Feb 6 21:47 labels -> ..data/labels
/etc/..2982_06_02_21_47_53.299460680:
total 8
-rw-r--r-- ... Feb 6 21:47 annotations
-rw-r--r-- ... Feb 6 21:47 labels
Using symbolic links enables dynamic atomic refresh of the metadata; updates are
written to a new temporary directory, and the ..data symlink is updated
atomically using
rename(2).
Exit the shell:
/# exit
The preceding exercise, you stored Pod fields in a DownwardAPIVolumeFile. In this next exercise, you store Container fields. Here is the configuration file for a Pod that has one Container:
apiVersion: v1
kind: Pod
metadata:
name: kubernetes-downwardapi-volume-example-2
spec:
containers:
- name: client-container
image: k8s.gcr.io/busybox:1.24
command: ["sh", "-c"]
args:
- while true; do
echo -en '\n';
if [[ -e /etc/podinfo/cpu_limit ]]; then
echo -en '\n'; cat /etc/podinfo/cpu_limit; fi;
if [[ -e /etc/podinfo/cpu_request ]]; then
echo -en '\n'; cat /etc/podinfo/cpu_request; fi;
if [[ -e /etc/podinfo/mem_limit ]]; then
echo -en '\n'; cat /etc/podinfo/mem_limit; fi;
if [[ -e /etc/podinfo/mem_request ]]; then
echo -en '\n'; cat /etc/podinfo/mem_request; fi;
sleep 5;
done;
resources:
requests:
memory: "32Mi"
cpu: "125m"
limits:
memory: "64Mi"
cpu: "250m"
volumeMounts:
- name: podinfo
mountPath: /etc/podinfo
volumes:
- name: podinfo
downwardAPI:
items:
- path: "cpu_limit"
resourceFieldRef:
containerName: client-container
resource: limits.cpu
divisor: 1m
- path: "cpu_request"
resourceFieldRef:
containerName: client-container
resource: requests.cpu
divisor: 1m
- path: "mem_limit"
resourceFieldRef:
containerName: client-container
resource: limits.memory
divisor: 1Mi
- path: "mem_request"
resourceFieldRef:
containerName: client-container
resource: requests.memory
divisor: 1Mi
In the configuration file, you can see that the Pod has a downwardAPI Volume,
and the Container mounts the Volume at /etc/podinfo.
Look at the items array under downwardAPI. Each element of the array is a
DownwardAPIVolumeFile.
The first element specifies that in the Container named client-container,
the value of the limits.cpu field in the format specified by 1m should be
stored in a file named cpu_limit. The divisor field is optional and has the
default value of 1 which means cores for cpu and bytes for memory.
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/inject/dapi-volume-resources.yaml
Get a shell into the Container that is running in your Pod:
kubectl exec -it kubernetes-downwardapi-volume-example-2 -- sh
In your shell, view the cpu_limit file:
/# cat /etc/podinfo/cpu_limit
You can use similar commands to view the cpu_request, mem_limit and
mem_request files.
The following information is available to containers through environment
variables and downwardAPI volumes:
fieldRef:
metadata.name - the pod's namemetadata.namespace - the pod's namespacemetadata.uid - the pod's UIDmetadata.labels['<KEY>'] - the value of the pod's label <KEY> (for example, metadata.labels['mylabel'])metadata.annotations['<KEY>'] - the value of the pod's annotation <KEY> (for example, metadata.annotations['myannotation'])resourceFieldRef:
DownwardAPIHugePages feature gate is enabled)DownwardAPIHugePages feature gate is enabled)In addition, the following information is available through
downwardAPI volume fieldRef:
metadata.labels - all of the pod's labels, formatted as label-key="escaped-label-value" with one label per linemetadata.annotations - all of the pod's annotations, formatted as annotation-key="escaped-annotation-value" with one annotation per lineThe following information is available through environment variables:
status.podIP - the pod's IP addressspec.serviceAccountName - the pod's service account name, available since v1.4.0-alpha.3spec.nodeName - the node's name, available since v1.4.0-alpha.3status.hostIP - the node's IP, available since v1.7.0-alpha.1You can project keys to specific paths and specific permissions on a per-file basis. For more information, see Secrets.
It is sometimes useful for a Container to have information about itself, without being overly coupled to Kubernetes. The Downward API allows containers to consume information about themselves or the cluster without using the Kubernetes client or API server.
An example is an existing application that assumes a particular well-known environment variable holds a unique identifier. One possibility is to wrap the application, but that is tedious and error prone, and it violates the goal of low coupling. A better option would be to use the Pod's name as an identifier, and inject the Pod's name into the well-known environment variable.
This page shows how to securely inject sensitive data, such as passwords and encryption keys, into Pods.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Suppose you want to have two pieces of secret data: a username my-app and a password
39528$vdg7Jb. First, use a base64 encoding tool to convert your username and password to a base64 representation. Here's an example using the commonly available base64 program:
echo -n 'my-app' | base64
echo -n '39528$vdg7Jb' | base64
The output shows that the base-64 representation of your username is bXktYXBw,
and the base-64 representation of your password is Mzk1MjgkdmRnN0pi.
Here is a configuration file you can use to create a Secret that holds your username and password:
apiVersion: v1
kind: Secret
metadata:
name: test-secret
data:
username: bXktYXBw
password: Mzk1MjgkdmRnN0pi
Create the Secret
kubectl apply -f https://k8s.io/examples/pods/inject/secret.yaml
View information about the Secret:
kubectl get secret test-secret
Output:
NAME TYPE DATA AGE
test-secret Opaque 2 1m
View more detailed information about the Secret:
kubectl describe secret test-secret
Output:
Name: test-secret
Namespace: default
Labels: <none>
Annotations: <none>
Type: Opaque
Data
====
password: 13 bytes
username: 7 bytes
If you want to skip the Base64 encoding step, you can create the
same Secret using the kubectl create secret command. For example:
kubectl create secret generic test-secret --from-literal='username=my-app' --from-literal='password=39528$vdg7Jb'
This is more convenient. The detailed approach shown earlier runs through each step explicitly to demonstrate what is happening.
Here is a configuration file you can use to create a Pod:
apiVersion: v1
kind: Pod
metadata:
name: secret-test-pod
spec:
containers:
- name: test-container
image: nginx
volumeMounts:
# name must match the volume name below
- name: secret-volume
mountPath: /etc/secret-volume
# The secret data is exposed to Containers in the Pod through a Volume.
volumes:
- name: secret-volume
secret:
secretName: test-secret
Create the Pod:
kubectl apply -f https://k8s.io/examples/pods/inject/secret-pod.yaml
Verify that your Pod is running:
kubectl get pod secret-test-pod
Output:
NAME READY STATUS RESTARTS AGE
secret-test-pod 1/1 Running 0 42m
Get a shell into the Container that is running in your Pod:
kubectl exec -i -t secret-test-pod -- /bin/bash
The secret data is exposed to the Container through a Volume mounted under
/etc/secret-volume.
In your shell, list the files in the /etc/secret-volume directory:
# Run this in the shell inside the container
ls /etc/secret-volume
The output shows two files, one for each piece of secret data:
password username
In your shell, display the contents of the username and password files:
# Run this in the shell inside the container
echo "$( cat /etc/secret-volume/username )"
echo "$( cat /etc/secret-volume/password )"
The output is your username and password:
my-app
39528$vdg7Jb
Define an environment variable as a key-value pair in a Secret:
kubectl create secret generic backend-user --from-literal=backend-username='backend-admin'
Assign the backend-username value defined in the Secret to the SECRET_USERNAME environment variable in the Pod specification.
apiVersion: v1
kind: Pod
metadata:
name: env-single-secret
spec:
containers:
- name: envars-test-container
image: nginx
env:
- name: SECRET_USERNAME
valueFrom:
secretKeyRef:
name: backend-user
key: backend-username
Create the Pod:
kubectl create -f https://k8s.io/examples/pods/inject/pod-single-secret-env-variable.yaml
In your shell, display the content of SECRET_USERNAME container environment variable
kubectl exec -i -t env-single-secret -- /bin/sh -c 'echo $SECRET_USERNAME'
The output is
backend-admin
As with the previous example, create the Secrets first.
kubectl create secret generic backend-user --from-literal=backend-username='backend-admin'
kubectl create secret generic db-user --from-literal=db-username='db-admin'
Define the environment variables in the Pod specification.
apiVersion: v1
kind: Pod
metadata:
name: envvars-multiple-secrets
spec:
containers:
- name: envars-test-container
image: nginx
env:
- name: BACKEND_USERNAME
valueFrom:
secretKeyRef:
name: backend-user
key: backend-username
- name: DB_USERNAME
valueFrom:
secretKeyRef:
name: db-user
key: db-username
Create the Pod:
kubectl create -f https://k8s.io/examples/pods/inject/pod-multiple-secret-env-variable.yaml
In your shell, display the container environment variables
kubectl exec -i -t envvars-multiple-secrets -- /bin/sh -c 'env | grep _USERNAME'
The output is
DB_USERNAME=db-admin
BACKEND_USERNAME=backend-admin
Create a Secret containing multiple key-value pairs
kubectl create secret generic test-secret --from-literal=username='my-app' --from-literal=password='39528$vdg7Jb'
Use envFrom to define all of the Secret's data as container environment variables. The key from the Secret becomes the environment variable name in the Pod.
apiVersion: v1
kind: Pod
metadata:
name: envfrom-secret
spec:
containers:
- name: envars-test-container
image: nginx
envFrom:
- secretRef:
name: test-secret
Create the Pod:
kubectl create -f https://k8s.io/examples/pods/inject/pod-secret-envFrom.yaml
In your shell, display username and password container environment variables
kubectl exec -i -t envfrom-secret -- /bin/sh -c 'echo "username: $username\npassword: $password\n"'
The output is
username: my-app
password: 39528$vdg7Jb
This page shows how to run an application using a Kubernetes Deployment object.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.9. To check the version, enterkubectl version.
You can run an application by creating a Kubernetes Deployment object, and you can describe a Deployment in a YAML file. For example, this YAML file describes a Deployment that runs the nginx:1.14.2 Docker image:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
replicas: 2 # tells deployment to run 2 pods matching the template
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Create a Deployment based on the YAML file:
kubectl apply -f https://k8s.io/examples/application/deployment.yaml
Display information about the Deployment:
kubectl describe deployment nginx-deployment
The output is similar to this:
Name: nginx-deployment
Namespace: default
CreationTimestamp: Tue, 30 Aug 2016 18:11:37 -0700
Labels: app=nginx
Annotations: deployment.kubernetes.io/revision=1
Selector: app=nginx
Replicas: 2 desired | 2 updated | 2 total | 2 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 1 max unavailable, 1 max surge
Pod Template:
Labels: app=nginx
Containers:
nginx:
Image: nginx:1.14.2
Port: 80/TCP
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Available True MinimumReplicasAvailable
Progressing True NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet: nginx-deployment-1771418926 (2/2 replicas created)
No events.
List the Pods created by the deployment:
kubectl get pods -l app=nginx
The output is similar to this:
NAME READY STATUS RESTARTS AGE
nginx-deployment-1771418926-7o5ns 1/1 Running 0 16h
nginx-deployment-1771418926-r18az 1/1 Running 0 16h
Display information about a Pod:
kubectl describe pod <pod-name>
where <pod-name> is the name of one of your Pods.
You can update the deployment by applying a new YAML file. This YAML file specifies that the deployment should be updated to use nginx 1.16.1.
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
replicas: 2
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.16.1 # Update the version of nginx from 1.14.2 to 1.16.1
ports:
- containerPort: 80
Apply the new YAML file:
kubectl apply -f https://k8s.io/examples/application/deployment-update.yaml
Watch the deployment create pods with new names and delete the old pods:
kubectl get pods -l app=nginx
You can increase the number of Pods in your Deployment by applying a new YAML
file. This YAML file sets replicas to 4, which specifies that the Deployment
should have four Pods:
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
replicas: 4 # Update the replicas from 2 to 4
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
Apply the new YAML file:
kubectl apply -f https://k8s.io/examples/application/deployment-scale.yaml
Verify that the Deployment has four Pods:
kubectl get pods -l app=nginx
The output is similar to this:
NAME READY STATUS RESTARTS AGE
nginx-deployment-148880595-4zdqq 1/1 Running 0 25s
nginx-deployment-148880595-6zgi1 1/1 Running 0 25s
nginx-deployment-148880595-fxcez 1/1 Running 0 2m
nginx-deployment-148880595-rwovn 1/1 Running 0 2m
Delete the deployment by name:
kubectl delete deployment nginx-deployment
The preferred way to create a replicated application is to use a Deployment, which in turn uses a ReplicaSet. Before the Deployment and ReplicaSet were added to Kubernetes, replicated applications were configured using a ReplicationController.
This page shows you how to run a single-instance stateful application in Kubernetes using a PersistentVolume and a Deployment. The application is MySQL.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
You need to either have a dynamic PersistentVolume provisioner with a default StorageClass, or statically provision PersistentVolumes yourself to satisfy the PersistentVolumeClaims used here.
You can run a stateful application by creating a Kubernetes Deployment and connecting it to an existing PersistentVolume using a PersistentVolumeClaim. For example, this YAML file describes a Deployment that runs MySQL and references the PersistentVolumeClaim. The file defines a volume mount for /var/lib/mysql, and then creates a PersistentVolumeClaim that looks for a 20G volume. This claim is satisfied by any existing volume that meets the requirements, or by a dynamic provisioner.
Note: The password is defined in the config yaml, and this is insecure. See Kubernetes Secrets for a secure solution.
apiVersion: v1
kind: Service
metadata:
name: mysql
spec:
ports:
- port: 3306
selector:
app: mysql
clusterIP: None
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: mysql
spec:
selector:
matchLabels:
app: mysql
strategy:
type: Recreate
template:
metadata:
labels:
app: mysql
spec:
containers:
- image: mysql:5.6
name: mysql
env:
# Use secret in real usage
- name: MYSQL_ROOT_PASSWORD
value: password
ports:
- containerPort: 3306
name: mysql
volumeMounts:
- name: mysql-persistent-storage
mountPath: /var/lib/mysql
volumes:
- name: mysql-persistent-storage
persistentVolumeClaim:
claimName: mysql-pv-claim
apiVersion: v1
kind: PersistentVolume
metadata:
name: mysql-pv-volume
labels:
type: local
spec:
storageClassName: manual
capacity:
storage: 20Gi
accessModes:
- ReadWriteOnce
hostPath:
path: "/mnt/data"
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: mysql-pv-claim
spec:
storageClassName: manual
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 20Gi
Deploy the PV and PVC of the YAML file:
kubectl apply -f https://k8s.io/examples/application/mysql/mysql-pv.yaml
Deploy the contents of the YAML file:
kubectl apply -f https://k8s.io/examples/application/mysql/mysql-deployment.yaml
Display information about the Deployment:
kubectl describe deployment mysql
The output is similar to this:
Name: mysql
Namespace: default
CreationTimestamp: Tue, 01 Nov 2016 11:18:45 -0700
Labels: app=mysql
Annotations: deployment.kubernetes.io/revision=1
Selector: app=mysql
Replicas: 1 desired | 1 updated | 1 total | 0 available | 1 unavailable
StrategyType: Recreate
MinReadySeconds: 0
Pod Template:
Labels: app=mysql
Containers:
mysql:
Image: mysql:5.6
Port: 3306/TCP
Environment:
MYSQL_ROOT_PASSWORD: password
Mounts:
/var/lib/mysql from mysql-persistent-storage (rw)
Volumes:
mysql-persistent-storage:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: mysql-pv-claim
ReadOnly: false
Conditions:
Type Status Reason
---- ------ ------
Available False MinimumReplicasUnavailable
Progressing True ReplicaSetUpdated
OldReplicaSets: <none>
NewReplicaSet: mysql-63082529 (1/1 replicas created)
Events:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
33s 33s 1 {deployment-controller } Normal ScalingReplicaSet Scaled up replica set mysql-63082529 to 1
List the pods created by the Deployment:
kubectl get pods -l app=mysql
The output is similar to this:
NAME READY STATUS RESTARTS AGE
mysql-63082529-2z3ki 1/1 Running 0 3m
Inspect the PersistentVolumeClaim:
kubectl describe pvc mysql-pv-claim
The output is similar to this:
Name: mysql-pv-claim
Namespace: default
StorageClass:
Status: Bound
Volume: mysql-pv-volume
Labels: <none>
Annotations: pv.kubernetes.io/bind-completed=yes
pv.kubernetes.io/bound-by-controller=yes
Capacity: 20Gi
Access Modes: RWO
Events: <none>
The preceding YAML file creates a service that
allows other Pods in the cluster to access the database. The Service option
clusterIP: None lets the Service DNS name resolve directly to the
Pod's IP address. This is optimal when you have only one Pod
behind a Service and you don't intend to increase the number of Pods.
Run a MySQL client to connect to the server:
kubectl run -it --rm --image=mysql:5.6 --restart=Never mysql-client -- mysql -h mysql -ppassword
This command creates a new Pod in the cluster running a MySQL client and connects it to the server through the Service. If it connects, you know your stateful MySQL database is up and running.
Waiting for pod default/mysql-client-274442439-zyp6i to be running, status is Pending, pod ready: false
If you don't see a command prompt, try pressing enter.
mysql>
The image or any other part of the Deployment can be updated as usual
with the kubectl apply command. Here are some precautions that are
specific to stateful apps:
strategy: type: Recreate in the Deployment configuration
YAML file. This instructs Kubernetes to not use rolling
updates. Rolling updates will not work, as you cannot have more than
one Pod running at a time. The Recreate strategy will stop the
first pod before creating a new one with the updated configuration.Delete the deployed objects by name:
kubectl delete deployment,svc mysql
kubectl delete pvc mysql-pv-claim
kubectl delete pv mysql-pv-volume
If you manually provisioned a PersistentVolume, you also need to manually delete it, as well as release the underlying resource. If you used a dynamic provisioner, it automatically deletes the PersistentVolume when it sees that you deleted the PersistentVolumeClaim. Some dynamic provisioners (such as those for EBS and PD) also release the underlying resource upon deleting the PersistentVolume.
Learn more about Deployment objects.
Learn more about Deploying applications
This page shows how to run a replicated stateful application using a StatefulSet controller. This application is a replicated MySQL database. The example topology has a single primary server and multiple replicas, using asynchronous row-based replication.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
You need to either have a dynamic PersistentVolume provisioner with a default StorageClass, or statically provision PersistentVolumes yourself to satisfy the PersistentVolumeClaims used here.
The example MySQL deployment consists of a ConfigMap, two Services, and a StatefulSet.
Create the ConfigMap from the following YAML configuration file:
apiVersion: v1
kind: ConfigMap
metadata:
name: mysql
labels:
app: mysql
data:
primary.cnf: |
# Apply this config only on the primary.
[mysqld]
log-bin
replica.cnf: |
# Apply this config only on replicas.
[mysqld]
super-read-only
kubectl apply -f https://k8s.io/examples/application/mysql/mysql-configmap.yaml
This ConfigMap provides my.cnf overrides that let you independently control
configuration on the primary MySQL server and replicas.
In this case, you want the primary server to be able to serve replication logs to replicas
and you want replicas to reject any writes that don't come via replication.
There's nothing special about the ConfigMap itself that causes different portions to apply to different Pods. Each Pod decides which portion to look at as it's initializing, based on information provided by the StatefulSet controller.
Create the Services from the following YAML configuration file:
# Headless service for stable DNS entries of StatefulSet members.
apiVersion: v1
kind: Service
metadata:
name: mysql
labels:
app: mysql
spec:
ports:
- name: mysql
port: 3306
clusterIP: None
selector:
app: mysql
---
# Client service for connecting to any MySQL instance for reads.
# For writes, you must instead connect to the primary: mysql-0.mysql.
apiVersion: v1
kind: Service
metadata:
name: mysql-read
labels:
app: mysql
spec:
ports:
- name: mysql
port: 3306
selector:
app: mysql
kubectl apply -f https://k8s.io/examples/application/mysql/mysql-services.yaml
The Headless Service provides a home for the DNS entries that the StatefulSet
controller creates for each Pod that's part of the set.
Because the Headless Service is named mysql, the Pods are accessible by
resolving <pod-name>.mysql from within any other Pod in the same Kubernetes
cluster and namespace.
The Client Service, called mysql-read, is a normal Service with its own
cluster IP that distributes connections across all MySQL Pods that report
being Ready. The set of potential endpoints includes the primary MySQL server and all
replicas.
Note that only read queries can use the load-balanced Client Service. Because there is only one primary MySQL server, clients should connect directly to the primary MySQL Pod (through its DNS entry within the Headless Service) to execute writes.
Finally, create the StatefulSet from the following YAML configuration file:
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: mysql
spec:
selector:
matchLabels:
app: mysql
serviceName: mysql
replicas: 3
template:
metadata:
labels:
app: mysql
spec:
initContainers:
- name: init-mysql
image: mysql:5.7
command:
- bash
- "-c"
- |
set -ex
# Generate mysql server-id from pod ordinal index.
[[ `hostname` =~ -([0-9]+)$ ]] || exit 1
ordinal=${BASH_REMATCH[1]}
echo [mysqld] > /mnt/conf.d/server-id.cnf
# Add an offset to avoid reserved server-id=0 value.
echo server-id=$((100 + $ordinal)) >> /mnt/conf.d/server-id.cnf
# Copy appropriate conf.d files from config-map to emptyDir.
if [[ $ordinal -eq 0 ]]; then
cp /mnt/config-map/primary.cnf /mnt/conf.d/
else
cp /mnt/config-map/replica.cnf /mnt/conf.d/
fi
volumeMounts:
- name: conf
mountPath: /mnt/conf.d
- name: config-map
mountPath: /mnt/config-map
- name: clone-mysql
image: gcr.io/google-samples/xtrabackup:1.0
command:
- bash
- "-c"
- |
set -ex
# Skip the clone if data already exists.
[[ -d /var/lib/mysql/mysql ]] && exit 0
# Skip the clone on primary (ordinal index 0).
[[ `hostname` =~ -([0-9]+)$ ]] || exit 1
ordinal=${BASH_REMATCH[1]}
[[ $ordinal -eq 0 ]] && exit 0
# Clone data from previous peer.
ncat --recv-only mysql-$(($ordinal-1)).mysql 3307 | xbstream -x -C /var/lib/mysql
# Prepare the backup.
xtrabackup --prepare --target-dir=/var/lib/mysql
volumeMounts:
- name: data
mountPath: /var/lib/mysql
subPath: mysql
- name: conf
mountPath: /etc/mysql/conf.d
containers:
- name: mysql
image: mysql:5.7
env:
- name: MYSQL_ALLOW_EMPTY_PASSWORD
value: "1"
ports:
- name: mysql
containerPort: 3306
volumeMounts:
- name: data
mountPath: /var/lib/mysql
subPath: mysql
- name: conf
mountPath: /etc/mysql/conf.d
resources:
requests:
cpu: 500m
memory: 1Gi
livenessProbe:
exec:
command: ["mysqladmin", "ping"]
initialDelaySeconds: 30
periodSeconds: 10
timeoutSeconds: 5
readinessProbe:
exec:
# Check we can execute queries over TCP (skip-networking is off).
command: ["mysql", "-h", "127.0.0.1", "-e", "SELECT 1"]
initialDelaySeconds: 5
periodSeconds: 2
timeoutSeconds: 1
- name: xtrabackup
image: gcr.io/google-samples/xtrabackup:1.0
ports:
- name: xtrabackup
containerPort: 3307
command:
- bash
- "-c"
- |
set -ex
cd /var/lib/mysql
# Determine binlog position of cloned data, if any.
if [[ -f xtrabackup_slave_info && "x$(<xtrabackup_slave_info)" != "x" ]]; then
# XtraBackup already generated a partial "CHANGE MASTER TO" query
# because we're cloning from an existing replica. (Need to remove the tailing semicolon!)
cat xtrabackup_slave_info | sed -E 's/;$//g' > change_master_to.sql.in
# Ignore xtrabackup_binlog_info in this case (it's useless).
rm -f xtrabackup_slave_info xtrabackup_binlog_info
elif [[ -f xtrabackup_binlog_info ]]; then
# We're cloning directly from primary. Parse binlog position.
[[ `cat xtrabackup_binlog_info` =~ ^(.*?)[[:space:]]+(.*?)$ ]] || exit 1
rm -f xtrabackup_binlog_info xtrabackup_slave_info
echo "CHANGE MASTER TO MASTER_LOG_FILE='${BASH_REMATCH[1]}',\
MASTER_LOG_POS=${BASH_REMATCH[2]}" > change_master_to.sql.in
fi
# Check if we need to complete a clone by starting replication.
if [[ -f change_master_to.sql.in ]]; then
echo "Waiting for mysqld to be ready (accepting connections)"
until mysql -h 127.0.0.1 -e "SELECT 1"; do sleep 1; done
echo "Initializing replication from clone position"
mysql -h 127.0.0.1 \
-e "$(<change_master_to.sql.in), \
MASTER_HOST='mysql-0.mysql', \
MASTER_USER='root', \
MASTER_PASSWORD='', \
MASTER_CONNECT_RETRY=10; \
START SLAVE;" || exit 1
# In case of container restart, attempt this at-most-once.
mv change_master_to.sql.in change_master_to.sql.orig
fi
# Start a server to send backups when requested by peers.
exec ncat --listen --keep-open --send-only --max-conns=1 3307 -c \
"xtrabackup --backup --slave-info --stream=xbstream --host=127.0.0.1 --user=root"
volumeMounts:
- name: data
mountPath: /var/lib/mysql
subPath: mysql
- name: conf
mountPath: /etc/mysql/conf.d
resources:
requests:
cpu: 100m
memory: 100Mi
volumes:
- name: conf
emptyDir: {}
- name: config-map
configMap:
name: mysql
volumeClaimTemplates:
- metadata:
name: data
spec:
accessModes: ["ReadWriteOnce"]
resources:
requests:
storage: 10Gi
kubectl apply -f https://k8s.io/examples/application/mysql/mysql-statefulset.yaml
You can watch the startup progress by running:
kubectl get pods -l app=mysql --watch
After a while, you should see all 3 Pods become Running:
NAME READY STATUS RESTARTS AGE
mysql-0 2/2 Running 0 2m
mysql-1 2/2 Running 0 1m
mysql-2 2/2 Running 0 1m
Press Ctrl+C to cancel the watch. If you don't see any progress, make sure you have a dynamic PersistentVolume provisioner enabled as mentioned in the prerequisites.
This manifest uses a variety of techniques for managing stateful Pods as part of a StatefulSet. The next section highlights some of these techniques to explain what happens as the StatefulSet creates Pods.
The StatefulSet controller starts Pods one at a time, in order by their ordinal index. It waits until each Pod reports being Ready before starting the next one.
In addition, the controller assigns each Pod a unique, stable name of the form
<statefulset-name>-<ordinal-index>, which results in Pods named mysql-0,
mysql-1, and mysql-2.
The Pod template in the above StatefulSet manifest takes advantage of these properties to perform orderly startup of MySQL replication.
Before starting any of the containers in the Pod spec, the Pod first runs any Init Containers in the order defined.
The first Init Container, named init-mysql, generates special MySQL config
files based on the ordinal index.
The script determines its own ordinal index by extracting it from the end of
the Pod name, which is returned by the hostname command.
Then it saves the ordinal (with a numeric offset to avoid reserved values)
into a file called server-id.cnf in the MySQL conf.d directory.
This translates the unique, stable identity provided by the StatefulSet
controller into the domain of MySQL server IDs, which require the same
properties.
The script in the init-mysql container also applies either primary.cnf or
replica.cnf from the ConfigMap by copying the contents into conf.d.
Because the example topology consists of a single primary MySQL server and any number of
replicas, the script assigns ordinal 0 to be the primary server, and everyone
else to be replicas.
Combined with the StatefulSet controller's
deployment order guarantee,
this ensures the primary MySQL server is Ready before creating replicas, so they can begin
replicating.
In general, when a new Pod joins the set as a replica, it must assume the primary MySQL server might already have data on it. It also must assume that the replication logs might not go all the way back to the beginning of time. These conservative assumptions are the key to allow a running StatefulSet to scale up and down over time, rather than being fixed at its initial size.
The second Init Container, named clone-mysql, performs a clone operation on
a replica Pod the first time it starts up on an empty PersistentVolume.
That means it copies all existing data from another running Pod,
so its local state is consistent enough to begin replicating from the primary server.
MySQL itself does not provide a mechanism to do this, so the example uses a
popular open-source tool called Percona XtraBackup.
During the clone, the source MySQL server might suffer reduced performance.
To minimize impact on the primary MySQL server, the script instructs each Pod to clone
from the Pod whose ordinal index is one lower.
This works because the StatefulSet controller always ensures Pod N is
Ready before starting Pod N+1.
After the Init Containers complete successfully, the regular containers run.
The MySQL Pods consist of a mysql container that runs the actual mysqld
server, and an xtrabackup container that acts as a
sidecar.
The xtrabackup sidecar looks at the cloned data files and determines if
it's necessary to initialize MySQL replication on the replica.
If so, it waits for mysqld to be ready and then executes the
CHANGE MASTER TO and START SLAVE commands with replication parameters
extracted from the XtraBackup clone files.
Once a replica begins replication, it remembers its primary MySQL server and
reconnects automatically if the server restarts or the connection dies.
Also, because replicas look for the primary server at its stable DNS name
(mysql-0.mysql), they automatically find the primary server even if it gets a new
Pod IP due to being rescheduled.
Lastly, after starting replication, the xtrabackup container listens for
connections from other Pods requesting a data clone.
This server remains up indefinitely in case the StatefulSet scales up, or in
case the next Pod loses its PersistentVolumeClaim and needs to redo the clone.
You can send test queries to the primary MySQL server (hostname mysql-0.mysql)
by running a temporary container with the mysql:5.7 image and running the
mysql client binary.
kubectl run mysql-client --image=mysql:5.7 -i --rm --restart=Never --\
mysql -h mysql-0.mysql <<EOF
CREATE DATABASE test;
CREATE TABLE test.messages (message VARCHAR(250));
INSERT INTO test.messages VALUES ('hello');
EOF
Use the hostname mysql-read to send test queries to any server that reports
being Ready:
kubectl run mysql-client --image=mysql:5.7 -i -t --rm --restart=Never --\
mysql -h mysql-read -e "SELECT * FROM test.messages"
You should get output like this:
Waiting for pod default/mysql-client to be running, status is Pending, pod ready: false
+---------+
| message |
+---------+
| hello |
+---------+
pod "mysql-client" deleted
To demonstrate that the mysql-read Service distributes connections across
servers, you can run SELECT @@server_id in a loop:
kubectl run mysql-client-loop --image=mysql:5.7 -i -t --rm --restart=Never --\
bash -ic "while sleep 1; do mysql -h mysql-read -e 'SELECT @@server_id,NOW()'; done"
You should see the reported @@server_id change randomly, because a different
endpoint might be selected upon each connection attempt:
+-------------+---------------------+
| @@server_id | NOW() |
+-------------+---------------------+
| 100 | 2006-01-02 15:04:05 |
+-------------+---------------------+
+-------------+---------------------+
| @@server_id | NOW() |
+-------------+---------------------+
| 102 | 2006-01-02 15:04:06 |
+-------------+---------------------+
+-------------+---------------------+
| @@server_id | NOW() |
+-------------+---------------------+
| 101 | 2006-01-02 15:04:07 |
+-------------+---------------------+
You can press Ctrl+C when you want to stop the loop, but it's useful to keep it running in another window so you can see the effects of the following steps.
To demonstrate the increased availability of reading from the pool of replicas
instead of a single server, keep the SELECT @@server_id loop from above
running while you force a Pod out of the Ready state.
The readiness probe
for the mysql container runs the command mysql -h 127.0.0.1 -e 'SELECT 1'
to make sure the server is up and able to execute queries.
One way to force this readiness probe to fail is to break that command:
kubectl exec mysql-2 -c mysql -- mv /usr/bin/mysql /usr/bin/mysql.off
This reaches into the actual container's filesystem for Pod mysql-2 and
renames the mysql command so the readiness probe can't find it.
After a few seconds, the Pod should report one of its containers as not Ready,
which you can check by running:
kubectl get pod mysql-2
Look for 1/2 in the READY column:
NAME READY STATUS RESTARTS AGE
mysql-2 1/2 Running 0 3m
At this point, you should see your SELECT @@server_id loop continue to run,
although it never reports 102 anymore.
Recall that the init-mysql script defined server-id as 100 + $ordinal,
so server ID 102 corresponds to Pod mysql-2.
Now repair the Pod and it should reappear in the loop output after a few seconds:
kubectl exec mysql-2 -c mysql -- mv /usr/bin/mysql.off /usr/bin/mysql
The StatefulSet also recreates Pods if they're deleted, similar to what a ReplicaSet does for stateless Pods.
kubectl delete pod mysql-2
The StatefulSet controller notices that no mysql-2 Pod exists anymore,
and creates a new one with the same name and linked to the same
PersistentVolumeClaim.
You should see server ID 102 disappear from the loop output for a while
and then return on its own.
If your Kubernetes cluster has multiple Nodes, you can simulate Node downtime (such as when Nodes are upgraded) by issuing a drain.
First determine which Node one of the MySQL Pods is on:
kubectl get pod mysql-2 -o wide
The Node name should show up in the last column:
NAME READY STATUS RESTARTS AGE IP NODE
mysql-2 2/2 Running 0 15m 10.244.5.27 kubernetes-node-9l2t
Then drain the Node by running the following command, which cordons it so
no new Pods may schedule there, and then evicts any existing Pods.
Replace <node-name> with the name of the Node you found in the last step.
This might impact other applications on the Node, so it's best to only do this in a test cluster.
kubectl drain <node-name> --force --delete-emptydir-data --ignore-daemonsets
Now you can watch as the Pod reschedules on a different Node:
kubectl get pod mysql-2 -o wide --watch
It should look something like this:
NAME READY STATUS RESTARTS AGE IP NODE
mysql-2 2/2 Terminating 0 15m 10.244.1.56 kubernetes-node-9l2t
[...]
mysql-2 0/2 Pending 0 0s <none> kubernetes-node-fjlm
mysql-2 0/2 Init:0/2 0 0s <none> kubernetes-node-fjlm
mysql-2 0/2 Init:1/2 0 20s 10.244.5.32 kubernetes-node-fjlm
mysql-2 0/2 PodInitializing 0 21s 10.244.5.32 kubernetes-node-fjlm
mysql-2 1/2 Running 0 22s 10.244.5.32 kubernetes-node-fjlm
mysql-2 2/2 Running 0 30s 10.244.5.32 kubernetes-node-fjlm
And again, you should see server ID 102 disappear from the
SELECT @@server_id loop output for a while and then return.
Now uncordon the Node to return it to a normal state:
kubectl uncordon <node-name>
With MySQL replication, you can scale your read query capacity by adding replicas. With StatefulSet, you can do this with a single command:
kubectl scale statefulset mysql --replicas=5
Watch the new Pods come up by running:
kubectl get pods -l app=mysql --watch
Once they're up, you should see server IDs 103 and 104 start appearing in
the SELECT @@server_id loop output.
You can also verify that these new servers have the data you added before they existed:
kubectl run mysql-client --image=mysql:5.7 -i -t --rm --restart=Never --\
mysql -h mysql-3.mysql -e "SELECT * FROM test.messages"
Waiting for pod default/mysql-client to be running, status is Pending, pod ready: false
+---------+
| message |
+---------+
| hello |
+---------+
pod "mysql-client" deleted
Scaling back down is also seamless:
kubectl scale statefulset mysql --replicas=3
Note, however, that while scaling up creates new PersistentVolumeClaims automatically, scaling down does not automatically delete these PVCs. This gives you the choice to keep those initialized PVCs around to make scaling back up quicker, or to extract data before deleting them.
You can see this by running:
kubectl get pvc -l app=mysql
Which shows that all 5 PVCs still exist, despite having scaled the StatefulSet down to 3:
NAME STATUS VOLUME CAPACITY ACCESSMODES AGE
data-mysql-0 Bound pvc-8acbf5dc-b103-11e6-93fa-42010a800002 10Gi RWO 20m
data-mysql-1 Bound pvc-8ad39820-b103-11e6-93fa-42010a800002 10Gi RWO 20m
data-mysql-2 Bound pvc-8ad69a6d-b103-11e6-93fa-42010a800002 10Gi RWO 20m
data-mysql-3 Bound pvc-50043c45-b1c5-11e6-93fa-42010a800002 10Gi RWO 2m
data-mysql-4 Bound pvc-500a9957-b1c5-11e6-93fa-42010a800002 10Gi RWO 2m
If you don't intend to reuse the extra PVCs, you can delete them:
kubectl delete pvc data-mysql-3
kubectl delete pvc data-mysql-4
Cancel the SELECT @@server_id loop by pressing Ctrl+C in its terminal,
or running the following from another terminal:
kubectl delete pod mysql-client-loop --now
Delete the StatefulSet. This also begins terminating the Pods.
kubectl delete statefulset mysql
Verify that the Pods disappear. They might take some time to finish terminating.
kubectl get pods -l app=mysql
You'll know the Pods have terminated when the above returns:
No resources found.
Delete the ConfigMap, Services, and PersistentVolumeClaims.
kubectl delete configmap,service,pvc -l app=mysql
If you manually provisioned PersistentVolumes, you also need to manually delete them, as well as release the underlying resources. If you used a dynamic provisioner, it automatically deletes the PersistentVolumes when it sees that you deleted the PersistentVolumeClaims. Some dynamic provisioners (such as those for EBS and PD) also release the underlying resources upon deleting the PersistentVolumes.
This task shows how to scale a StatefulSet. Scaling a StatefulSet refers to increasing or decreasing the number of replicas.
StatefulSets are only available in Kubernetes version 1.5 or later.
To check your version of Kubernetes, run kubectl version.
Not all stateful applications scale nicely. If you are unsure about whether to scale your StatefulSets, see StatefulSet concepts or StatefulSet tutorial for further information.
You should perform scaling only when you are confident that your stateful application cluster is completely healthy.
First, find the StatefulSet you want to scale.
kubectl get statefulsets <stateful-set-name>
Change the number of replicas of your StatefulSet:
kubectl scale statefulsets <stateful-set-name> --replicas=<new-replicas>
Alternatively, you can do in-place updates on your StatefulSets.
If your StatefulSet was initially created with kubectl apply,
update .spec.replicas of the StatefulSet manifests, and then do a kubectl apply:
kubectl apply -f <stateful-set-file-updated>
Otherwise, edit that field with kubectl edit:
kubectl edit statefulsets <stateful-set-name>
Or use kubectl patch:
kubectl patch statefulsets <stateful-set-name> -p '{"spec":{"replicas":<new-replicas>}}'
You cannot scale down a StatefulSet when any of the stateful Pods it manages is unhealthy. Scaling down only takes place after those stateful Pods become running and ready.
If spec.replicas > 1, Kubernetes cannot determine the reason for an unhealthy Pod. It might be the result of a permanent fault or of a transient fault. A transient fault can be caused by a restart required by upgrading or maintenance.
If the Pod is unhealthy due to a permanent fault, scaling without correcting the fault may lead to a state where the StatefulSet membership drops below a certain minimum number of replicas that are needed to function correctly. This may cause your StatefulSet to become unavailable.
If the Pod is unhealthy due to a transient fault and the Pod might become available again, the transient error may interfere with your scale-up or scale-down operation. Some distributed databases have issues when nodes join and leave at the same time. It is better to reason about scaling operations at the application level in these cases, and perform scaling only when you are sure that your stateful application cluster is completely healthy.
This task shows you how to delete a StatefulSet.
You can delete a StatefulSet in the same way you delete other resources in Kubernetes: use the kubectl delete command, and specify the StatefulSet either by file or by name.
kubectl delete -f <file.yaml>
kubectl delete statefulsets <statefulset-name>
You may need to delete the associated headless service separately after the StatefulSet itself is deleted.
kubectl delete service <service-name>
When deleting a StatefulSet through kubectl, the StatefulSet scales down to 0. All Pods that are part of this workload are also deleted. If you want to delete only the StatefulSet and not the Pods, use --cascade=orphan.
For example:
kubectl delete -f <file.yaml> --cascade=orphan
By passing --cascade=orphan to kubectl delete, the Pods managed by the StatefulSet are left behind even after the StatefulSet object itself is deleted. If the pods have a label app=myapp, you can then delete them as follows:
kubectl delete pods -l app=myapp
Deleting the Pods in a StatefulSet will not delete the associated volumes. This is to ensure that you have the chance to copy data off the volume before deleting it. Deleting the PVC after the pods have terminated might trigger deletion of the backing Persistent Volumes depending on the storage class and reclaim policy. You should never assume ability to access a volume after claim deletion.
To delete everything in a StatefulSet, including the associated pods, you can run a series of commands similar to the following:
grace=$(kubectl get pods <stateful-set-pod> --template '{{.spec.terminationGracePeriodSeconds}}')
kubectl delete statefulset -l app=myapp
sleep $grace
kubectl delete pvc -l app=myapp
In the example above, the Pods have the label app=myapp; substitute your own label as appropriate.
If you find that some pods in your StatefulSet are stuck in the 'Terminating' or 'Unknown' states for an extended period of time, you may need to manually intervene to forcefully delete the pods from the apiserver. This is a potentially dangerous task. Refer to Force Delete StatefulSet Pods for details.
Learn more about force deleting StatefulSet Pods.
This page shows how to delete Pods which are part of a stateful set, and explains the considerations to keep in mind when doing so.
In normal operation of a StatefulSet, there is never a need to force delete a StatefulSet Pod. The StatefulSet controller is responsible for creating, scaling and deleting members of the StatefulSet. It tries to ensure that the specified number of Pods from ordinal 0 through N-1 are alive and ready. StatefulSet ensures that, at any time, there is at most one Pod with a given identity running in a cluster. This is referred to as at most one semantics provided by a StatefulSet.
Manual force deletion should be undertaken with caution, as it has the potential to violate the at most one semantics inherent to StatefulSet. StatefulSets may be used to run distributed and clustered applications which have a need for a stable network identity and stable storage. These applications often have configuration which relies on an ensemble of a fixed number of members with fixed identities. Having multiple members with the same identity can be disastrous and may lead to data loss (e.g. split brain scenario in quorum-based systems).
You can perform a graceful pod deletion with the following command:
kubectl delete pods <pod>
For the above to lead to graceful termination, the Pod must not specify a
pod.Spec.TerminationGracePeriodSeconds of 0. The practice of setting a
pod.Spec.TerminationGracePeriodSeconds of 0 seconds is unsafe and strongly discouraged
for StatefulSet Pods. Graceful deletion is safe and will ensure that the Pod
shuts down gracefully
before the kubelet deletes the name from the apiserver.
A Pod is not deleted automatically when a node is unreachable. The Pods running on an unreachable Node enter the 'Terminating' or 'Unknown' state after a timeout. Pods may also enter these states when the user attempts graceful deletion of a Pod on an unreachable Node. The only ways in which a Pod in such a state can be removed from the apiserver are as follows:
The recommended best practice is to use the first or second approach. If a Node is confirmed to be dead (e.g. permanently disconnected from the network, powered down, etc), then delete the Node object. If the Node is suffering from a network partition, then try to resolve this or wait for it to resolve. When the partition heals, the kubelet will complete the deletion of the Pod and free up its name in the apiserver.
Normally, the system completes the deletion once the Pod is no longer running on a Node, or the Node is deleted by an administrator. You may override this by force deleting the Pod.
Force deletions do not wait for confirmation from the kubelet that the Pod has been terminated. Irrespective of whether a force deletion is successful in killing a Pod, it will immediately free up the name from the apiserver. This would let the StatefulSet controller create a replacement Pod with that same identity; this can lead to the duplication of a still-running Pod, and if said Pod can still communicate with the other members of the StatefulSet, will violate the at most one semantics that StatefulSet is designed to guarantee.
When you force delete a StatefulSet pod, you are asserting that the Pod in question will never again make contact with other Pods in the StatefulSet and its name can be safely freed up for a replacement to be created.
If you want to delete a Pod forcibly using kubectl version >= 1.5, do the following:
kubectl delete pods <pod> --grace-period=0 --force
If you're using any version of kubectl <= 1.4, you should omit the --force option and use:
kubectl delete pods <pod> --grace-period=0
If even after these commands the pod is stuck on Unknown state, use the following command to remove the pod from the cluster:
kubectl patch pod <pod> -p '{"metadata":{"finalizers":null}}'
Always perform force deletion of StatefulSet Pods carefully and with complete knowledge of the risks involved.
Learn more about debugging a StatefulSet.
The Horizontal Pod Autoscaler automatically scales the number of Pods in a replication controller, deployment, replica set or stateful set based on observed CPU utilization (or, with custom metrics support, on some other application-provided metrics). Note that Horizontal Pod Autoscaling does not apply to objects that can't be scaled, for example, DaemonSets.
The Horizontal Pod Autoscaler is implemented as a Kubernetes API resource and a controller. The resource determines the behavior of the controller. The controller periodically adjusts the number of replicas in a replication controller or deployment to match the observed metrics such as average CPU utilisation, average memory utilisation or any other custom metric to the target specified by the user.
The Horizontal Pod Autoscaler is implemented as a control loop, with a period controlled
by the controller manager's --horizontal-pod-autoscaler-sync-period flag (with a default
value of 15 seconds).
During each period, the controller manager queries the resource utilization against the metrics specified in each HorizontalPodAutoscaler definition. The controller manager obtains the metrics from either the resource metrics API (for per-pod resource metrics), or the custom metrics API (for all other metrics).
For per-pod resource metrics (like CPU), the controller fetches the metrics from the resource metrics API for each Pod targeted by the HorizontalPodAutoscaler. Then, if a target utilization value is set, the controller calculates the utilization value as a percentage of the equivalent resource request on the containers in each Pod. If a target raw value is set, the raw metric values are used directly. The controller then takes the mean of the utilization or the raw value (depending on the type of target specified) across all targeted Pods, and produces a ratio used to scale the number of desired replicas.
Please note that if some of the Pod's containers do not have the relevant resource request set, CPU utilization for the Pod will not be defined and the autoscaler will not take any action for that metric. See the algorithm details section below for more information about how the autoscaling algorithm works.
For per-pod custom metrics, the controller functions similarly to per-pod resource metrics, except that it works with raw values, not utilization values.
For object metrics and external metrics, a single metric is fetched, which describes
the object in question. This metric is compared to the target
value, to produce a ratio as above. In the autoscaling/v2beta2 API
version, this value can optionally be divided by the number of Pods before the
comparison is made.
The HorizontalPodAutoscaler normally fetches metrics from a series of aggregated APIs (metrics.k8s.io,
custom.metrics.k8s.io, and external.metrics.k8s.io). The metrics.k8s.io API is usually provided by
metrics-server, which needs to be launched separately. For more information about resource metrics, see Metrics Server.
See Support for metrics APIs for more details.
The autoscaler accesses corresponding scalable controllers (such as replication controllers, deployments, and replica sets) by using the scale sub-resource. Scale is an interface that allows you to dynamically set the number of replicas and examine each of their current states. More details on scale sub-resource can be found here.
From the most basic perspective, the Horizontal Pod Autoscaler controller operates on the ratio between desired metric value and current metric value:
desiredReplicas = ceil[currentReplicas * ( currentMetricValue / desiredMetricValue )]
For example, if the current metric value is 200m, and the desired value
is 100m, the number of replicas will be doubled, since 200.0 / 100.0 == 2.0 If the current value is instead 50m, we'll halve the number of
replicas, since 50.0 / 100.0 == 0.5. We'll skip scaling if the ratio is
sufficiently close to 1.0 (within a globally-configurable tolerance, from
the --horizontal-pod-autoscaler-tolerance flag, which defaults to 0.1).
When a targetAverageValue or targetAverageUtilization is specified,
the currentMetricValue is computed by taking the average of the given
metric across all Pods in the HorizontalPodAutoscaler's scale target.
Before checking the tolerance and deciding on the final values, we take
pod readiness and missing metrics into consideration, however.
All Pods with a deletion timestamp set (i.e. Pods in the process of being shut down) and all failed Pods are discarded.
If a particular Pod is missing metrics, it is set aside for later; Pods with missing metrics will be used to adjust the final scaling amount.
When scaling on CPU, if any pod has yet to become ready (i.e. it's still initializing) or the most recent metric point for the pod was before it became ready, that pod is set aside as well.
Due to technical constraints, the HorizontalPodAutoscaler controller
cannot exactly determine the first time a pod becomes ready when
determining whether to set aside certain CPU metrics. Instead, it
considers a Pod "not yet ready" if it's unready and transitioned to
unready within a short, configurable window of time since it started.
This value is configured with the --horizontal-pod-autoscaler-initial-readiness-delay flag, and its default is 30
seconds. Once a pod has become ready, it considers any transition to
ready to be the first if it occurred within a longer, configurable time
since it started. This value is configured with the --horizontal-pod-autoscaler-cpu-initialization-period flag, and its
default is 5 minutes.
The currentMetricValue / desiredMetricValue base scale ratio is then
calculated using the remaining pods not set aside or discarded from above.
If there were any missing metrics, we recompute the average more conservatively, assuming those pods were consuming 100% of the desired value in case of a scale down, and 0% in case of a scale up. This dampens the magnitude of any potential scale.
Furthermore, if any not-yet-ready pods were present, and we would have scaled up without factoring in missing metrics or not-yet-ready pods, we conservatively assume the not-yet-ready pods are consuming 0% of the desired metric, further dampening the magnitude of a scale up.
After factoring in the not-yet-ready pods and missing metrics, we recalculate the usage ratio. If the new ratio reverses the scale direction, or is within the tolerance, we skip scaling. Otherwise, we use the new ratio to scale.
Note that the original value for the average utilization is reported back via the HorizontalPodAutoscaler status, without factoring in the not-yet-ready pods or missing metrics, even when the new usage ratio is used.
If multiple metrics are specified in a HorizontalPodAutoscaler, this
calculation is done for each metric, and then the largest of the desired
replica counts is chosen. If any of these metrics cannot be converted
into a desired replica count (e.g. due to an error fetching the metrics
from the metrics APIs) and a scale down is suggested by the metrics which
can be fetched, scaling is skipped. This means that the HPA is still capable
of scaling up if one or more metrics give a desiredReplicas greater than
the current value.
Finally, right before HPA scales the target, the scale recommendation is recorded. The
controller considers all recommendations within a configurable window choosing the
highest recommendation from within that window. This value can be configured using the --horizontal-pod-autoscaler-downscale-stabilization flag, which defaults to 5 minutes.
This means that scaledowns will occur gradually, smoothing out the impact of rapidly
fluctuating metric values.
The Horizontal Pod Autoscaler is an API resource in the Kubernetes autoscaling API group.
The current stable version, which only includes support for CPU autoscaling,
can be found in the autoscaling/v1 API version.
The beta version, which includes support for scaling on memory and custom metrics,
can be found in autoscaling/v2beta2. The new fields introduced in autoscaling/v2beta2
are preserved as annotations when working with autoscaling/v1.
When you create a HorizontalPodAutoscaler API object, make sure the name specified is a valid DNS subdomain name. More details about the API object can be found at HorizontalPodAutoscaler Object.
Horizontal Pod Autoscaler, like every API resource, is supported in a standard way by kubectl.
We can create a new autoscaler using kubectl create command.
We can list autoscalers by kubectl get hpa and get detailed description by kubectl describe hpa.
Finally, we can delete an autoscaler using kubectl delete hpa.
In addition, there is a special kubectl autoscale command for creating a HorizontalPodAutoscaler object.
For instance, executing kubectl autoscale rs foo --min=2 --max=5 --cpu-percent=80
will create an autoscaler for replication set foo, with target CPU utilization set to 80%
and the number of replicas between 2 and 5.
The detailed documentation of kubectl autoscale can be found here.
Kubernetes lets you perform a rolling update on a Deployment. In that
case, the Deployment manages the underlying ReplicaSets for you.
When you configure autoscaling for a Deployment, you bind a
HorizontalPodAutoscaler to a single Deployment. The HorizontalPodAutoscaler
manages the replicas field of the Deployment. The deployment controller is responsible
for setting the replicas of the underlying ReplicaSets so that they add up to a suitable
number during the rollout and also afterwards.
If you perform a rolling update of a StatefulSet that has an autoscaled number of replicas, the StatefulSet directly manages its set of Pods (there is no intermediate resource similar to ReplicaSet).
When managing the scale of a group of replicas using the Horizontal Pod Autoscaler, it is possible that the number of replicas keeps fluctuating frequently due to the dynamic nature of the metrics evaluated. This is sometimes referred to as thrashing.
Starting from v1.6, a cluster operator can mitigate this problem by tuning
the global HPA settings exposed as flags for the kube-controller-manager component:
Starting from v1.12, a new algorithmic update removes the need for the upscale delay.
--horizontal-pod-autoscaler-downscale-stabilization: Specifies the duration of the
downscale stabilization time window. Horizontal Pod Autoscaler remembers
the historical recommended sizes and only acts on the largest size within this time window.
The default value is 5 minutes (5m0s).Any HPA target can be scaled based on the resource usage of the pods in the scaling target.
When defining the pod specification the resource requests like cpu and memory should
be specified. This is used to determine the resource utilization and used by the HPA controller
to scale the target up or down. To use resource utilization based scaling specify a metric source
like this:
type: Resource
resource:
name: cpu
target:
type: Utilization
averageUtilization: 60
With this metric the HPA controller will keep the average utilization of the pods in the scaling target at 60%. Utilization is the ratio between the current usage of resource to the requested resources of the pod. See Algorithm for more details about how the utilization is calculated and averaged.
Kubernetes v1.20 [alpha]
HorizontalPodAutoscaler also supports a container metric source where the HPA can track the
resource usage of individual containers across a set of Pods, in order to scale the target resource.
This lets you configure scaling thresholds for the containers that matter most in a particular Pod.
For example, if you have a web application and a logging sidecar, you can scale based on the resource
use of the web application, ignoring the sidecar container and its resource use.
If you revise the target resource to have a new Pod specification with a different set of containers, you should revise the HPA spec if that newly added container should also be used for scaling. If the specified container in the metric source is not present or only present in a subset of the pods then those pods are ignored and the recommendation is recalculated. See Algorithm for more details about the calculation. To use container resources for autoscaling define a metric source as follows:
type: ContainerResource
containerResource:
name: cpu
container: application
target:
type: Utilization
averageUtilization: 60
In the above example the HPA controller scales the target such that the average utilization of the cpu
in the application container of all the pods is 60%.
If you change the name of a container that a HorizontalPodAutoscaler is tracking, you can make that change in a specific order to ensure scaling remains available and effective whilst the change is being applied. Before you update the resource that defines the container (such as a Deployment), you should update the associated HPA to track both the new and old container names. This way, the HPA is able to calculate a scaling recommendation throughout the update process.
Once you have rolled out the container name change to the workload resource, tidy up by removing the old container name from the HPA specification.
Kubernetes 1.6 adds support for scaling based on multiple metrics. You can use the autoscaling/v2beta2 API
version to specify multiple metrics for the Horizontal Pod Autoscaler to scale on. Then, the Horizontal Pod
Autoscaler controller will evaluate each metric, and propose a new scale based on that metric. The largest of the
proposed scales will be used as the new scale.
Kubernetes 1.6 adds support for making use of custom metrics in the Horizontal Pod Autoscaler.
You can add custom metrics for the Horizontal Pod Autoscaler to use in the autoscaling/v2beta2 API.
Kubernetes then queries the new custom metrics API to fetch the values of the appropriate custom metrics.
See Support for metrics APIs for the requirements.
By default, the HorizontalPodAutoscaler controller retrieves metrics from a series of APIs. In order for it to access these APIs, cluster administrators must ensure that:
The API aggregation layer is enabled.
The corresponding APIs are registered:
For resource metrics, this is the metrics.k8s.io API, generally provided by metrics-server.
It can be launched as a cluster addon.
For custom metrics, this is the custom.metrics.k8s.io API. It's provided by "adapter" API servers provided by metrics solution vendors.
Check with your metrics pipeline, or the list of known solutions.
If you would like to write your own, check out the boilerplate to get started.
For external metrics, this is the external.metrics.k8s.io API. It may be provided by the custom metrics adapters provided above.
For more information on these different metrics paths and how they differ please see the relevant design proposals for the HPA V2, custom.metrics.k8s.io and external.metrics.k8s.io.
For examples of how to use them see the walkthrough for using custom metrics and the walkthrough for using external metrics.
Starting from
v1.18
the v2beta2 API allows scaling behavior to be configured through the HPA
behavior field. Behaviors are specified separately for scaling up and down in
scaleUp or scaleDown section under the behavior field. A stabilization
window can be specified for both directions which prevents the flapping of the
number of the replicas in the scaling target. Similarly specifying scaling
policies controls the rate of change of replicas while scaling.
One or more scaling policies can be specified in the behavior section of the spec.
When multiple policies are specified the policy which allows the highest amount of
change is the policy which is selected by default. The following example shows this behavior
while scaling down:
behavior:
scaleDown:
policies:
- type: Pods
value: 4
periodSeconds: 60
- type: Percent
value: 10
periodSeconds: 60
periodSeconds indicates the length of time in the past for which the policy must hold true.
The first policy (Pods) allows at most 4 replicas to be scaled down in one minute. The second policy
(Percent) allows at most 10% of the current replicas to be scaled down in one minute.
Since by default the policy which allows the highest amount of change is selected, the second policy will only be used when the number of pod replicas is more than 40. With 40 or less replicas, the first policy will be applied. For instance if there are 80 replicas and the target has to be scaled down to 10 replicas then during the first step 8 replicas will be reduced. In the next iteration when the number of replicas is 72, 10% of the pods is 7.2 but the number is rounded up to 8. On each loop of the autoscaler controller the number of pods to be change is re-calculated based on the number of current replicas. When the number of replicas falls below 40 the first policy (Pods) is applied and 4 replicas will be reduced at a time.
The policy selection can be changed by specifying the selectPolicy field for a scaling
direction. By setting the value to Min which would select the policy which allows the
smallest change in the replica count. Setting the value to Disabled completely disables
scaling in that direction.
The stabilization window is used to restrict the flapping of replicas when the metrics
used for scaling keep fluctuating. The stabilization window is used by the autoscaling
algorithm to consider the computed desired state from the past to prevent scaling. In
the following example the stabilization window is specified for scaleDown.
scaleDown:
stabilizationWindowSeconds: 300
When the metrics indicate that the target should be scaled down the algorithm looks into previously computed desired states and uses the highest value from the specified interval. In above example all desired states from the past 5 minutes will be considered.
To use the custom scaling not all fields have to be specified. Only values which need to be customized can be specified. These custom values are merged with default values. The default values match the existing behavior in the HPA algorithm.
behavior:
scaleDown:
stabilizationWindowSeconds: 300
policies:
- type: Percent
value: 100
periodSeconds: 15
scaleUp:
stabilizationWindowSeconds: 0
policies:
- type: Percent
value: 100
periodSeconds: 15
- type: Pods
value: 4
periodSeconds: 15
selectPolicy: Max
For scaling down the stabilization window is 300 seconds (or the value of the
--horizontal-pod-autoscaler-downscale-stabilization flag if provided). There is only a single policy
for scaling down which allows a 100% of the currently running replicas to be removed which
means the scaling target can be scaled down to the minimum allowed replicas.
For scaling up there is no stabilization window. When the metrics indicate that the target should be
scaled up the target is scaled up immediately. There are 2 policies where 4 pods or a 100% of the currently
running replicas will be added every 15 seconds till the HPA reaches its steady state.
To provide a custom downscale stabilization window of 1 minute, the following behavior would be added to the HPA:
behavior:
scaleDown:
stabilizationWindowSeconds: 60
To limit the rate at which pods are removed by the HPA to 10% per minute, the following behavior would be added to the HPA:
behavior:
scaleDown:
policies:
- type: Percent
value: 10
periodSeconds: 60
To ensure that no more than 5 Pods are removed per minute, you can add a second scale-down
policy with a fixed size of 5, and set selectPolicy to minimum. Setting selectPolicy to Min means
that the autoscaler chooses the policy that affects the smallest number of Pods:
behavior:
scaleDown:
policies:
- type: Percent
value: 10
periodSeconds: 60
- type: Pods
value: 5
periodSeconds: 60
selectPolicy: Min
The selectPolicy value of Disabled turns off scaling the given direction.
So to prevent downscaling the following policy would be used:
behavior:
scaleDown:
selectPolicy: Disabled
You can implicitly deactivate the HPA for a target without the
need to change the HPA configuration itself. If the target's desired replica count
is set to 0, and the HPA's minimum replica count is greater than 0, the HPA
stops adjusting the target (and sets the ScalingActive Condition on itself
to false) until you reactivate it by manually adjusting the target's desired
replica count or HPA's minimum replica count.
Horizontal Pod Autoscaler automatically scales the number of Pods in a replication controller, deployment, replica set or stateful set based on observed CPU utilization (or, with beta support, on some other, application-provided metrics).
This document walks you through an example of enabling Horizontal Pod Autoscaler for the php-apache server. For more information on how Horizontal Pod Autoscaler behaves, see the Horizontal Pod Autoscaler user guide.
This example requires a running Kubernetes cluster and kubectl, version 1.2 or later. Metrics server monitoring needs to be deployed in the cluster to provide metrics through the Metrics API. Horizontal Pod Autoscaler uses this API to collect metrics. To learn how to deploy the metrics-server, see the metrics-server documentation.
To specify multiple resource metrics for a Horizontal Pod Autoscaler, you must have a Kubernetes cluster and kubectl at version 1.6 or later. To make use of custom metrics, your cluster must be able to communicate with the API server providing the custom Metrics API. Finally, to use metrics not related to any Kubernetes object you must have a Kubernetes cluster at version 1.10 or later, and you must be able to communicate with the API server that provides the external Metrics API. See the Horizontal Pod Autoscaler user guide for more details.
To demonstrate Horizontal Pod Autoscaler we will use a custom docker image based on the php-apache image. The Dockerfile has the following content:
FROM php:5-apache
COPY index.php /var/www/html/index.php
RUN chmod a+rx index.php
It defines an index.php page which performs some CPU intensive computations:
<?php
$x = 0.0001;
for ($i = 0; $i <= 1000000; $i++) {
$x += sqrt($x);
}
echo "OK!";
?>
First, we will start a deployment running the image and expose it as a service using the following configuration:
apiVersion: apps/v1
kind: Deployment
metadata:
name: php-apache
spec:
selector:
matchLabels:
run: php-apache
replicas: 1
template:
metadata:
labels:
run: php-apache
spec:
containers:
- name: php-apache
image: k8s.gcr.io/hpa-example
ports:
- containerPort: 80
resources:
limits:
cpu: 500m
requests:
cpu: 200m
---
apiVersion: v1
kind: Service
metadata:
name: php-apache
labels:
run: php-apache
spec:
ports:
- port: 80
selector:
run: php-apache
Run the following command:
kubectl apply -f https://k8s.io/examples/application/php-apache.yaml
deployment.apps/php-apache created
service/php-apache created
Now that the server is running, we will create the autoscaler using
kubectl autoscale.
The following command will create a Horizontal Pod Autoscaler that maintains between 1 and 10 replicas of the Pods
controlled by the php-apache deployment we created in the first step of these instructions.
Roughly speaking, HPA will increase and decrease the number of replicas
(via the deployment) to maintain an average CPU utilization across all Pods of 50%.
Since each pod requests 200 milli-cores by kubectl run, this means an average CPU usage of 100 milli-cores.
See here for more details on the algorithm.
kubectl autoscale deployment php-apache --cpu-percent=50 --min=1 --max=10
horizontalpodautoscaler.autoscaling/php-apache autoscaled
We may check the current status of autoscaler by running:
kubectl get hpa
NAME REFERENCE TARGET MINPODS MAXPODS REPLICAS AGE
php-apache Deployment/php-apache/scale 0% / 50% 1 10 1 18s
Please note that the current CPU consumption is 0% as we are not sending any requests to the server
(the TARGET column shows the average across all the pods controlled by the corresponding deployment).
Now, we will see how the autoscaler reacts to increased load. We will start a container, and send an infinite loop of queries to the php-apache service (please run it in a different terminal):
kubectl run -i --tty load-generator --rm --image=busybox --restart=Never -- /bin/sh -c "while sleep 0.01; do wget -q -O- http://php-apache; done"
Within a minute or so, we should see the higher CPU load by executing:
kubectl get hpa
NAME REFERENCE TARGET MINPODS MAXPODS REPLICAS AGE
php-apache Deployment/php-apache/scale 305% / 50% 1 10 1 3m
Here, CPU consumption has increased to 305% of the request. As a result, the deployment was resized to 7 replicas:
kubectl get deployment php-apache
NAME READY UP-TO-DATE AVAILABLE AGE
php-apache 7/7 7 7 19m
We will finish our example by stopping the user load.
In the terminal where we created the container with busybox image, terminate
the load generation by typing <Ctrl> + C.
Then we will verify the result state (after a minute or so):
kubectl get hpa
NAME REFERENCE TARGET MINPODS MAXPODS REPLICAS AGE
php-apache Deployment/php-apache/scale 0% / 50% 1 10 1 11m
kubectl get deployment php-apache
NAME READY UP-TO-DATE AVAILABLE AGE
php-apache 1/1 1 1 27m
Here CPU utilization dropped to 0, and so HPA autoscaled the number of replicas back down to 1.
You can introduce additional metrics to use when autoscaling the php-apache Deployment
by making use of the autoscaling/v2beta2 API version.
First, get the YAML of your HorizontalPodAutoscaler in the autoscaling/v2beta2 form:
kubectl get hpa php-apache -o yaml > /tmp/hpa-v2.yaml
Open the /tmp/hpa-v2.yaml file in an editor, and you should see YAML which looks like this:
apiVersion: autoscaling/v2beta2
kind: HorizontalPodAutoscaler
metadata:
name: php-apache
spec:
scaleTargetRef:
apiVersion: apps/v1
kind: Deployment
name: php-apache
minReplicas: 1
maxReplicas: 10
metrics:
- type: Resource
resource:
name: cpu
target:
type: Utilization
averageUtilization: 50
status:
observedGeneration: 1
lastScaleTime: <some-time>
currentReplicas: 1
desiredReplicas: 1
currentMetrics:
- type: Resource
resource:
name: cpu
current:
averageUtilization: 0
averageValue: 0
Notice that the targetCPUUtilizationPercentage field has been replaced with an array called metrics.
The CPU utilization metric is a resource metric, since it is represented as a percentage of a resource
specified on pod containers. Notice that you can specify other resource metrics besides CPU. By default,
the only other supported resource metric is memory. These resources do not change names from cluster
to cluster, and should always be available, as long as the metrics.k8s.io API is available.
You can also specify resource metrics in terms of direct values, instead of as percentages of the
requested value, by using a target.type of AverageValue instead of Utilization, and
setting the corresponding target.averageValue field instead of the target.averageUtilization.
There are two other types of metrics, both of which are considered custom metrics: pod metrics and object metrics. These metrics may have names which are cluster specific, and require a more advanced cluster monitoring setup.
The first of these alternative metric types is pod metrics. These metrics describe Pods, and
are averaged together across Pods and compared with a target value to determine the replica count.
They work much like resource metrics, except that they only support a target type of AverageValue.
Pod metrics are specified using a metric block like this:
type: Pods
pods:
metric:
name: packets-per-second
target:
type: AverageValue
averageValue: 1k
The second alternative metric type is object metrics. These metrics describe a different
object in the same namespace, instead of describing Pods. The metrics are not necessarily
fetched from the object; they only describe it. Object metrics support target types of
both Value and AverageValue. With Value, the target is compared directly to the returned
metric from the API. With AverageValue, the value returned from the custom metrics API is divided
by the number of Pods before being compared to the target. The following example is the YAML
representation of the requests-per-second metric.
type: Object
object:
metric:
name: requests-per-second
describedObject:
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
name: main-route
target:
type: Value
value: 2k
If you provide multiple such metric blocks, the HorizontalPodAutoscaler will consider each metric in turn. The HorizontalPodAutoscaler will calculate proposed replica counts for each metric, and then choose the one with the highest replica count.
For example, if you had your monitoring system collecting metrics about network traffic,
you could update the definition above using kubectl edit to look like this:
apiVersion: autoscaling/v2beta2
kind: HorizontalPodAutoscaler
metadata:
name: php-apache
spec:
scaleTargetRef:
apiVersion: apps/v1
kind: Deployment
name: php-apache
minReplicas: 1
maxReplicas: 10
metrics:
- type: Resource
resource:
name: cpu
target:
type: Utilization
averageUtilization: 50
- type: Pods
pods:
metric:
name: packets-per-second
target:
type: AverageValue
averageValue: 1k
- type: Object
object:
metric:
name: requests-per-second
describedObject:
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
name: main-route
target:
type: Value
value: 10k
status:
observedGeneration: 1
lastScaleTime: <some-time>
currentReplicas: 1
desiredReplicas: 1
currentMetrics:
- type: Resource
resource:
name: cpu
current:
averageUtilization: 0
averageValue: 0
- type: Object
object:
metric:
name: requests-per-second
describedObject:
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
name: main-route
current:
value: 10k
Then, your HorizontalPodAutoscaler would attempt to ensure that each pod was consuming roughly 50% of its requested CPU, serving 1000 packets per second, and that all pods behind the main-route Ingress were serving a total of 10000 requests per second.
Many metrics pipelines allow you to describe metrics either by name or by a set of additional
descriptors called labels. For all non-resource metric types (pod, object, and external,
described below), you can specify an additional label selector which is passed to your metric
pipeline. For instance, if you collect a metric http_requests with the verb
label, you can specify the following metric block to scale only on GET requests:
type: Object
object:
metric:
name: http_requests
selector: {matchLabels: {verb: GET}}
This selector uses the same syntax as the full Kubernetes label selectors. The monitoring pipeline
determines how to collapse multiple series into a single value, if the name and selector
match multiple series. The selector is additive, and cannot select metrics
that describe objects that are not the target object (the target pods in the case of the Pods
type, and the described object in the case of the Object type).
Applications running on Kubernetes may need to autoscale based on metrics that don't have an obvious relationship to any object in the Kubernetes cluster, such as metrics describing a hosted service with no direct correlation to Kubernetes namespaces. In Kubernetes 1.10 and later, you can address this use case with external metrics.
Using external metrics requires knowledge of your monitoring system; the setup is
similar to that required when using custom metrics. External metrics allow you to autoscale your cluster
based on any metric available in your monitoring system. Provide a metric block with a
name and selector, as above, and use the External metric type instead of Object.
If multiple time series are matched by the metricSelector,
the sum of their values is used by the HorizontalPodAutoscaler.
External metrics support both the Value and AverageValue target types, which function exactly the same
as when you use the Object type.
For example if your application processes tasks from a hosted queue service, you could add the following section to your HorizontalPodAutoscaler manifest to specify that you need one worker per 30 outstanding tasks.
- type: External
external:
metric:
name: queue_messages_ready
selector:
matchLabels:
queue: "worker_tasks"
target:
type: AverageValue
averageValue: 30
When possible, it's preferable to use the custom metric target types instead of external metrics, since it's easier for cluster administrators to secure the custom metrics API. The external metrics API potentially allows access to any metric, so cluster administrators should take care when exposing it.
When using the autoscaling/v2beta2 form of the HorizontalPodAutoscaler, you will be able to see
status conditions set by Kubernetes on the HorizontalPodAutoscaler. These status conditions indicate
whether or not the HorizontalPodAutoscaler is able to scale, and whether or not it is currently restricted
in any way.
The conditions appear in the status.conditions field. To see the conditions affecting a HorizontalPodAutoscaler,
we can use kubectl describe hpa:
kubectl describe hpa cm-test
Name: cm-test
Namespace: prom
Labels: <none>
Annotations: <none>
CreationTimestamp: Fri, 16 Jun 2017 18:09:22 +0000
Reference: ReplicationController/cm-test
Metrics: ( current / target )
"http_requests" on pods: 66m / 500m
Min replicas: 1
Max replicas: 4
ReplicationController pods: 1 current / 1 desired
Conditions:
Type Status Reason Message
---- ------ ------ -------
AbleToScale True ReadyForNewScale the last scale time was sufficiently old as to warrant a new scale
ScalingActive True ValidMetricFound the HPA was able to successfully calculate a replica count from pods metric http_requests
ScalingLimited False DesiredWithinRange the desired replica count is within the acceptable range
Events:
For this HorizontalPodAutoscaler, we can see several conditions in a healthy state. The first,
AbleToScale, indicates whether or not the HPA is able to fetch and update scales, as well as
whether or not any backoff-related conditions would prevent scaling. The second, ScalingActive,
indicates whether or not the HPA is enabled (i.e. the replica count of the target is not zero) and
is able to calculate desired scales. When it is False, it generally indicates problems with
fetching metrics. Finally, the last condition, ScalingLimited, indicates that the desired scale
was capped by the maximum or minimum of the HorizontalPodAutoscaler. This is an indication that
you may wish to raise or lower the minimum or maximum replica count constraints on your
HorizontalPodAutoscaler.
All metrics in the HorizontalPodAutoscaler and metrics APIs are specified using
a special whole-number notation known in Kubernetes as a
quantity. For example,
the quantity 10500m would be written as 10.5 in decimal notation. The metrics APIs
will return whole numbers without a suffix when possible, and will generally return
quantities in milli-units otherwise. This means you might see your metric value fluctuate
between 1 and 1500m, or 1 and 1.5 when written in decimal notation.
Instead of using kubectl autoscale command to create a HorizontalPodAutoscaler imperatively we
can use the following file to create it declaratively:
apiVersion: autoscaling/v1
kind: HorizontalPodAutoscaler
metadata:
name: php-apache
spec:
scaleTargetRef:
apiVersion: apps/v1
kind: Deployment
name: php-apache
minReplicas: 1
maxReplicas: 10
targetCPUUtilizationPercentage: 50
We will create the autoscaler by executing the following command:
kubectl create -f https://k8s.io/examples/application/hpa/php-apache.yaml
horizontalpodautoscaler.autoscaling/php-apache created
Kubernetes v1.21 [stable]
This page shows how to limit the number of concurrent disruptions that your application experiences, allowing for higher availability while permitting the cluster administrator to manage the clusters nodes.
kubectl version.
The most common use case when you want to protect an application specified by one of the built-in Kubernetes controllers:
In this case, make a note of the controller's .spec.selector; the same
selector goes into the PDBs .spec.selector.
From version 1.15 PDBs support custom controllers where the scale subresource is enabled.
You can also use PDBs with pods which are not controlled by one of the above controllers, or arbitrary groups of pods, but there are some restrictions, described in Arbitrary Controllers and Selectors.
Decide how many instances can be down at the same time for a short period due to a voluntary disruption.
Values for minAvailable or maxUnavailable can be expressed as integers or as a percentage.
minAvailable to 10, then 10
Pods must always be available, even during a disruption."50%"), it represents a percentage of
total Pods. For instance, if you set maxUnavailable to "50%", then only 50% of the Pods can be unavailable during a
disruption.When you specify the value as a percentage, it may not map to an exact number of Pods. For example, if you have 7 Pods and
you set minAvailable to "50%", it's not immediately obvious whether that means 3 Pods or 4 Pods must be available.
Kubernetes rounds up to the nearest integer, so in this case, 4 Pods must be available. You can examine the
code
that controls this behavior.
A PodDisruptionBudget has three fields:
.spec.selector to specify the set of
pods to which it applies. This field is required..spec.minAvailable which is a description of the number of pods from that
set that must still be available after the eviction, even in the absence
of the evicted pod. minAvailable can be either an absolute number or a percentage..spec.maxUnavailable (available in Kubernetes 1.7 and higher) which is a description
of the number of pods from that set that can be unavailable after the eviction.
It can be either an absolute number or a percentage.You can specify only one of maxUnavailable and minAvailable in a single PodDisruptionBudget.
maxUnavailable can only be used to control the eviction of pods
that have an associated controller managing them. In the examples below, "desired replicas"
is the scale of the controller managing the pods being selected by the
PodDisruptionBudget.
Example 1: With a minAvailable of 5, evictions are allowed as long as they leave behind
5 or more healthy pods among those selected by the PodDisruptionBudget's selector.
Example 2: With a minAvailable of 30%, evictions are allowed as long as at least 30%
of the number of desired replicas are healthy.
Example 3: With a maxUnavailable of 5, evictions are allowed as long as there are at most 5
unhealthy replicas among the total number of desired replicas.
Example 4: With a maxUnavailable of 30%, evictions are allowed as long as no more than 30%
of the desired replicas are unhealthy.
In typical usage, a single budget would be used for a collection of pods managed by a controller—for example, the pods in a single ReplicaSet or StatefulSet.
If you set maxUnavailable to 0% or 0, or you set minAvailable to 100% or the number of replicas,
you are requiring zero voluntary evictions. When you set zero voluntary evictions for a workload
object such as ReplicaSet, then you cannot successfully drain a Node running one of those Pods.
If you try to drain a Node where an unevictable Pod is running, the drain never completes. This is permitted as per the
semantics of PodDisruptionBudget.
You can find examples of pod disruption budgets defined below. They match pods with the label
app: zookeeper.
Example PDB Using minAvailable:
Example PDB Using maxUnavailable:
For example, if the above zk-pdb object selects the pods of a StatefulSet of size 3, both
specifications have the exact same meaning. The use of maxUnavailable is recommended as it
automatically responds to changes in the number of replicas of the corresponding controller.
You can create or update the PDB object with a command like kubectl apply -f mypdb.yaml.
Use kubectl to check that your PDB is created.
Assuming you don't actually have pods matching app: zookeeper in your namespace,
then you'll see something like this:
kubectl get poddisruptionbudgets
NAME MIN AVAILABLE MAX UNAVAILABLE ALLOWED DISRUPTIONS AGE
zk-pdb 2 N/A 0 7s
If there are matching pods (say, 3), then you would see something like this:
kubectl get poddisruptionbudgets
NAME MIN AVAILABLE MAX UNAVAILABLE ALLOWED DISRUPTIONS AGE
zk-pdb 2 N/A 1 7s
The non-zero value for ALLOWED DISRUPTIONS means that the disruption controller has seen the pods,
counted the matching pods, and updated the status of the PDB.
You can get more information about the status of a PDB with this command:
kubectl get poddisruptionbudgets zk-pdb -o yaml
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
annotations:
…
creationTimestamp: "2020-03-04T04:22:56Z"
generation: 1
name: zk-pdb
…
status:
currentHealthy: 3
desiredHealthy: 2
disruptionsAllowed: 1
expectedPods: 3
observedGeneration: 1
You can skip this section if you only use PDBs with the built-in application controllers (Deployment, ReplicationController, ReplicaSet, and StatefulSet), with the PDB selector matching the controller's selector.
You can use a PDB with pods controlled by another type of controller, by an "operator", or bare pods, but with these restrictions:
.spec.minAvailable can be used, not .spec.maxUnavailable..spec.minAvailable, not a percentage.You can use a selector which selects a subset or superset of the pods belonging to a built-in controller. The eviction API will disallow eviction of any pod covered by multiple PDBs, so most users will want to avoid overlapping selectors. One reasonable use of overlapping PDBs is when pods are being transitioned from one PDB to another.
This guide demonstrates how to access the Kubernetes API from within a pod.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
When accessing the API from within a Pod, locating and authenticating to the API server are slightly different to the external client case.
The easiest way to use the Kubernetes API from a Pod is to use one of the official client libraries. These libraries can automatically discover the API server and authenticate.
From within a Pod, the recommended ways to connect to the Kubernetes API are:
For a Go client, use the official Go client library.
The rest.InClusterConfig() function handles API host discovery and authentication automatically.
See an example here.
For a Python client, use the official Python client library.
The config.load_incluster_config() function handles API host discovery and authentication automatically.
See an example here.
There are a number of other libraries available, please refer to the Client Libraries page.
In each case, the service account credentials of the Pod are used to communicate securely with the API server.
While running in a Pod, the Kubernetes apiserver is accessible via a Service named
kubernetes in the default namespace. Therefore, Pods can use the
kubernetes.default.svc hostname to query the API server. Official client libraries
do this automatically.
The recommended way to authenticate to the API server is with a
service account credential. By default, a Pod
is associated with a service account, and a credential (token) for that
service account is placed into the filesystem tree of each container in that Pod,
at /var/run/secrets/kubernetes.io/serviceaccount/token.
If available, a certificate bundle is placed into the filesystem tree of each
container at /var/run/secrets/kubernetes.io/serviceaccount/ca.crt, and should be
used to verify the serving certificate of the API server.
Finally, the default namespace to be used for namespaced API operations is placed in a file
at /var/run/secrets/kubernetes.io/serviceaccount/namespace in each container.
If you would like to query the API without an official client library, you can run kubectl proxy
as the command
of a new sidecar container in the Pod. This way, kubectl proxy will authenticate
to the API and expose it on the localhost interface of the Pod, so that other containers
in the Pod can use it directly.
It is possible to avoid using the kubectl proxy by passing the authentication token directly to the API server. The internal certificate secures the connection.
# Point to the internal API server hostname
APISERVER=https://kubernetes.default.svc
# Path to ServiceAccount token
SERVICEACCOUNT=/var/run/secrets/kubernetes.io/serviceaccount
# Read this Pod's namespace
NAMESPACE=$(cat ${SERVICEACCOUNT}/namespace)
# Read the ServiceAccount bearer token
TOKEN=$(cat ${SERVICEACCOUNT}/token)
# Reference the internal certificate authority (CA)
CACERT=${SERVICEACCOUNT}/ca.crt
# Explore the API with TOKEN
curl --cacert ${CACERT} --header "Authorization: Bearer ${TOKEN}" -X GET ${APISERVER}/api
The output will be similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
CronJobs was promoted to general availability in Kubernetes v1.21. If you are using an older version of
Kubernetes, please refer to the documentation for the version of Kubernetes that you are using,
so that you see accurate information. Older Kubernetes versions do not support the batch/v1 CronJob API.
You can use a CronJob to run Jobs on a time-based schedule. These automated jobs run like Cron tasks on a Linux or UNIX system.
Cron jobs are useful for creating periodic and recurring tasks, like running backups or sending emails. Cron jobs can also schedule individual tasks for a specific time, such as if you want to schedule a job for a low activity period.
Cron jobs have limitations and idiosyncrasies. For example, in certain circumstances, a single cron job can create multiple jobs. Therefore, jobs should be idempotent.
For more limitations, see CronJobs.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Cron jobs require a config file.
This example cron job config .spec file prints the current time and a hello message every minute:
apiVersion: batch/v1
kind: CronJob
metadata:
name: hello
spec:
schedule: "*/1 * * * *"
jobTemplate:
spec:
template:
spec:
containers:
- name: hello
image: busybox
imagePullPolicy: IfNotPresent
command:
- /bin/sh
- -c
- date; echo Hello from the Kubernetes cluster
restartPolicy: OnFailure
Run the example CronJob by using this command:
kubectl create -f https://k8s.io/examples/application/job/cronjob.yaml
The output is similar to this:
cronjob.batch/hello created
After creating the cron job, get its status using this command:
kubectl get cronjob hello
The output is similar to this:
NAME SCHEDULE SUSPEND ACTIVE LAST SCHEDULE AGE
hello */1 * * * * False 0 <none> 10s
As you can see from the results of the command, the cron job has not scheduled or run any jobs yet. Watch for the job to be created in around one minute:
kubectl get jobs --watch
The output is similar to this:
NAME COMPLETIONS DURATION AGE
hello-4111706356 0/1 0s
hello-4111706356 0/1 0s 0s
hello-4111706356 1/1 5s 5s
Now you've seen one running job scheduled by the "hello" cron job. You can stop watching the job and view the cron job again to see that it scheduled the job:
kubectl get cronjob hello
The output is similar to this:
NAME SCHEDULE SUSPEND ACTIVE LAST SCHEDULE AGE
hello */1 * * * * False 0 50s 75s
You should see that the cron job hello successfully scheduled a job at the time specified in LAST SCHEDULE. There are currently 0 active jobs, meaning that the job has completed or failed.
Now, find the pods that the last scheduled job created and view the standard output of one of the pods.
# Replace "hello-4111706356" with the job name in your system
pods=$(kubectl get pods --selector=job-name=hello-4111706356 --output=jsonpath={.items[*].metadata.name})
Show pod log:
kubectl logs $pods
The output is similar to this:
Fri Feb 22 11:02:09 UTC 2019
Hello from the Kubernetes cluster
When you don't need a cron job any more, delete it with kubectl delete cronjob <cronjob name>:
kubectl delete cronjob hello
Deleting the cron job removes all the jobs and pods it created and stops it from creating additional jobs. You can read more about removing jobs in garbage collection.
As with all other Kubernetes configs, a cron job needs apiVersion, kind, and metadata fields. For general
information about working with config files, see deploying applications,
and using kubectl to manage resources documents.
A cron job config also needs a .spec section.
.spec, are applied only to the following runs.
The .spec.schedule is a required field of the .spec.
It takes a Cron format string, such as 0 * * * * or @hourly, as schedule time of its jobs to be created and executed.
The format also includes extended "Vixie cron" step values. As explained in the FreeBSD manual:
Step values can be used in conjunction with ranges. Following a range with
/<number>specifies skips of the number's value through the range. For example,0-23/2can be used in the hours field to specify command execution every other hour (the alternative in the V7 standard is0,2,4,6,8,10,12,14,16,18,20,22). Steps are also permitted after an asterisk, so if you want to say "every two hours", just use*/2.
?) in the schedule has the same meaning as an asterisk *, that is, it stands for any of available value for a given field.
The .spec.jobTemplate is the template for the job, and it is required.
It has exactly the same schema as a Job, except that it is nested and does not have an apiVersion or kind.
For information about writing a job .spec, see Writing a Job Spec.
The .spec.startingDeadlineSeconds field is optional.
It stands for the deadline in seconds for starting the job if it misses its scheduled time for any reason.
After the deadline, the cron job does not start the job.
Jobs that do not meet their deadline in this way count as failed jobs.
If this field is not specified, the jobs have no deadline.
If the .spec.startingDeadlineSeconds field is set (not null), the CronJob
controller measures the time between when a job is expected to be created and
now. If the difference is higher than that limit, it will skip this execution.
For example, if it is set to 200, it allows a job to be created for up to 200
seconds after the actual schedule.
The .spec.concurrencyPolicy field is also optional.
It specifies how to treat concurrent executions of a job that is created by this cron job.
The spec may specify only one of the following concurrency policies:
Allow (default): The cron job allows concurrently running jobsForbid: The cron job does not allow concurrent runs; if it is time for a new job run and the previous job run hasn't finished yet, the cron job skips the new job runReplace: If it is time for a new job run and the previous job run hasn't finished yet, the cron job replaces the currently running job run with a new job runNote that concurrency policy only applies to the jobs created by the same cron job. If there are multiple cron jobs, their respective jobs are always allowed to run concurrently.
The .spec.suspend field is also optional.
If it is set to true, all subsequent executions are suspended.
This setting does not apply to already started executions.
Defaults to false.
.spec.suspend changes from true to false on an existing cron job without a starting deadline, the missed jobs are scheduled immediately.
The .spec.successfulJobsHistoryLimit and .spec.failedJobsHistoryLimit fields are optional.
These fields specify how many completed and failed jobs should be kept.
By default, they are set to 3 and 1 respectively. Setting a limit to 0 corresponds to keeping none of the corresponding kind of jobs after they finish.
In this example, we will run a Kubernetes Job with multiple parallel worker processes.
In this example, as each pod is created, it picks up one unit of work from a task queue, completes it, deletes it from the queue, and exits.
Here is an overview of the steps in this example:
Be familiar with the basic, non-parallel, use of Job.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
This example uses RabbitMQ, however, you can adapt the example to use another AMQP-type message service.
In practice you could set up a message queue service once in a cluster and reuse it for many jobs, as well as for long-running services.
Start RabbitMQ as follows:
kubectl create -f https://raw.githubusercontent.com/kubernetes/kubernetes/release-1.3/examples/celery-rabbitmq/rabbitmq-service.yaml
service "rabbitmq-service" created
kubectl create -f https://raw.githubusercontent.com/kubernetes/kubernetes/release-1.3/examples/celery-rabbitmq/rabbitmq-controller.yaml
replicationcontroller "rabbitmq-controller" created
We will only use the rabbitmq part from the celery-rabbitmq example.
Now, we can experiment with accessing the message queue. We will create a temporary interactive pod, install some tools on it, and experiment with queues.
First create a temporary interactive Pod.
# Create a temporary interactive container
kubectl run -i --tty temp --image ubuntu:18.04
Waiting for pod default/temp-loe07 to be running, status is Pending, pod ready: false
... [ previous line repeats several times .. hit return when it stops ] ...
Note that your pod name and command prompt will be different.
Next install the amqp-tools so we can work with message queues.
# Install some tools
root@temp-loe07:/# apt-get update
.... [ lots of output ] ....
root@temp-loe07:/# apt-get install -y curl ca-certificates amqp-tools python dnsutils
.... [ lots of output ] ....
Later, we will make a docker image that includes these packages.
Next, we will check that we can discover the rabbitmq service:
# Note the rabbitmq-service has a DNS name, provided by Kubernetes:
root@temp-loe07:/# nslookup rabbitmq-service
Server: 10.0.0.10
Address: 10.0.0.10#53
Name: rabbitmq-service.default.svc.cluster.local
Address: 10.0.147.152
# Your address will vary.
If Kube-DNS is not setup correctly, the previous step may not work for you. You can also find the service IP in an env var:
# env | grep RABBIT | grep HOST
RABBITMQ_SERVICE_SERVICE_HOST=10.0.147.152
# Your address will vary.
Next we will verify we can create a queue, and publish and consume messages.
# In the next line, rabbitmq-service is the hostname where the rabbitmq-service
# can be reached. 5672 is the standard port for rabbitmq.
root@temp-loe07:/# export BROKER_URL=amqp://guest:guest@rabbitmq-service:5672
# If you could not resolve "rabbitmq-service" in the previous step,
# then use this command instead:
# root@temp-loe07:/# BROKER_URL=amqp://guest:guest@$RABBITMQ_SERVICE_SERVICE_HOST:5672
# Now create a queue:
root@temp-loe07:/# /usr/bin/amqp-declare-queue --url=$BROKER_URL -q foo -d
foo
# Publish one message to it:
root@temp-loe07:/# /usr/bin/amqp-publish --url=$BROKER_URL -r foo -p -b Hello
# And get it back.
root@temp-loe07:/# /usr/bin/amqp-consume --url=$BROKER_URL -q foo -c 1 cat && echo
Hello
root@temp-loe07:/#
In the last command, the amqp-consume tool takes one message (-c 1)
from the queue, and passes that message to the standard input of an arbitrary command. In this case, the program cat prints out the characters read from standard input, and the echo adds a carriage
return so the example is readable.
Now let's fill the queue with some "tasks". In our example, our tasks are strings to be printed.
In a practice, the content of the messages might be:
In practice, if there is large data that is needed in a read-only mode by all pods of the Job, you will typically put that in a shared file system like NFS and mount that readonly on all the pods, or the program in the pod will natively read data from a cluster file system like HDFS.
For our example, we will create the queue and fill it using the amqp command line tools. In practice, you might write a program to fill the queue using an amqp client library.
/usr/bin/amqp-declare-queue --url=$BROKER_URL -q job1 -d
job1
for f in apple banana cherry date fig grape lemon melon
do
/usr/bin/amqp-publish --url=$BROKER_URL -r job1 -p -b $f
done
So, we filled the queue with 8 messages.
Now we are ready to create an image that we will run as a job.
We will use the amqp-consume utility to read the message
from the queue and run our actual program. Here is a very simple
example program:
#!/usr/bin/env python
# Just prints standard out and sleeps for 10 seconds.
import sys
import time
print("Processing " + sys.stdin.readlines()[0])
time.sleep(10)
Give the script execution permission:
chmod +x worker.py
Now, build an image. If you are working in the source
tree, then change directory to examples/job/work-queue-1.
Otherwise, make a temporary directory, change to it,
download the Dockerfile,
and worker.py. In either case,
build the image with this command:
docker build -t job-wq-1 .
For the Docker Hub, tag your app image with
your username and push to the Hub with the below commands. Replace
<username> with your Hub username.
docker tag job-wq-1 <username>/job-wq-1
docker push <username>/job-wq-1
If you are using Google Container
Registry, tag
your app image with your project ID, and push to GCR. Replace
<project> with your project ID.
docker tag job-wq-1 gcr.io/<project>/job-wq-1
gcloud docker -- push gcr.io/<project>/job-wq-1
Here is a job definition. You'll need to make a copy of the Job and edit the
image to match the name you used, and call it ./job.yaml.
apiVersion: batch/v1
kind: Job
metadata:
name: job-wq-1
spec:
completions: 8
parallelism: 2
template:
metadata:
name: job-wq-1
spec:
containers:
- name: c
image: gcr.io/<project>/job-wq-1
env:
- name: BROKER_URL
value: amqp://guest:guest@rabbitmq-service:5672
- name: QUEUE
value: job1
restartPolicy: OnFailure
In this example, each pod works on one item from the queue and then exits.
So, the completion count of the Job corresponds to the number of work items
done. So we set, .spec.completions: 8 for the example, since we put 8 items in the queue.
So, now run the Job:
kubectl apply -f ./job.yaml
Now wait a bit, then check on the job.
kubectl describe jobs/job-wq-1
Name: job-wq-1
Namespace: default
Selector: controller-uid=41d75705-92df-11e7-b85e-fa163ee3c11f
Labels: controller-uid=41d75705-92df-11e7-b85e-fa163ee3c11f
job-name=job-wq-1
Annotations: <none>
Parallelism: 2
Completions: 8
Start Time: Wed, 06 Sep 2017 16:42:02 +0800
Pods Statuses: 0 Running / 8 Succeeded / 0 Failed
Pod Template:
Labels: controller-uid=41d75705-92df-11e7-b85e-fa163ee3c11f
job-name=job-wq-1
Containers:
c:
Image: gcr.io/causal-jigsaw-637/job-wq-1
Port:
Environment:
BROKER_URL: amqp://guest:guest@rabbitmq-service:5672
QUEUE: job1
Mounts: <none>
Volumes: <none>
Events:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
───────── ──────── ───── ──── ───────────── ────── ────── ───────
27s 27s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-hcobb
27s 27s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-weytj
27s 27s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-qaam5
27s 27s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-b67sr
26s 26s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-xe5hj
15s 15s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-w2zqe
14s 14s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-d6ppa
14s 14s 1 {job } Normal SuccessfulCreate Created pod: job-wq-1-p17e0
All our pods succeeded. Yay.
This approach has the advantage that you do not need to modify your "worker" program to be aware that there is a work queue.
It does require that you run a message queue service. If running a queue service is inconvenient, you may want to consider one of the other job patterns.
This approach creates a pod for every work item. If your work items only take a few seconds, though, creating a Pod for every work item may add a lot of overhead. Consider another example, that executes multiple work items per Pod.
In this example, we use the amqp-consume utility to read the message
from the queue and run our actual program. This has the advantage that you
do not need to modify your program to be aware of the queue.
A different example, shows how to
communicate with the work queue using a client library.
If the number of completions is set to less than the number of items in the queue, then not all items will be processed.
If the number of completions is set to more than the number of items in the queue, then the Job will not appear to be completed, even though all items in the queue have been processed. It will start additional pods which will block waiting for a message.
There is an unlikely race with this pattern. If the container is killed in between the time that the message is acknowledged by the amqp-consume command and the time that the container exits with success, or if the node crashes before the kubelet is able to post the success of the pod back to the api-server, then the Job will not appear to be complete, even though all items in the queue have been processed.
In this example, we will run a Kubernetes Job with multiple parallel worker processes in a given pod.
In this example, as each pod is created, it picks up one unit of work from a task queue, processes it, and repeats until the end of the queue is reached.
Here is an overview of the steps in this example:
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Be familiar with the basic, non-parallel, use of Job.
For this example, for simplicity, we will start a single instance of Redis. See the Redis Example for an example of deploying Redis scalably and redundantly.
You could also download the following files directly:
Now let's fill the queue with some "tasks". In our example, our tasks are strings to be printed.
Start a temporary interactive pod for running the Redis CLI.
kubectl run -i --tty temp --image redis --command "/bin/sh"
Waiting for pod default/redis2-c7h78 to be running, status is Pending, pod ready: false
Hit enter for command prompt
Now hit enter, start the redis CLI, and create a list with some work items in it.
# redis-cli -h redis
redis:6379> rpush job2 "apple"
(integer) 1
redis:6379> rpush job2 "banana"
(integer) 2
redis:6379> rpush job2 "cherry"
(integer) 3
redis:6379> rpush job2 "date"
(integer) 4
redis:6379> rpush job2 "fig"
(integer) 5
redis:6379> rpush job2 "grape"
(integer) 6
redis:6379> rpush job2 "lemon"
(integer) 7
redis:6379> rpush job2 "melon"
(integer) 8
redis:6379> rpush job2 "orange"
(integer) 9
redis:6379> lrange job2 0 -1
1) "apple"
2) "banana"
3) "cherry"
4) "date"
5) "fig"
6) "grape"
7) "lemon"
8) "melon"
9) "orange"
So, the list with key job2 will be our work queue.
Note: if you do not have Kube DNS setup correctly, you may need to change
the first step of the above block to redis-cli -h $REDIS_SERVICE_HOST.
Now we are ready to create an image that we will run.
We will use a python worker program with a redis client to read the messages from the message queue.
A simple Redis work queue client library is provided, called rediswq.py (Download).
The "worker" program in each Pod of the Job uses the work queue client library to get work. Here it is:
#!/usr/bin/env python
import time
import rediswq
host="redis"
# Uncomment next two lines if you do not have Kube-DNS working.
# import os
# host = os.getenv("REDIS_SERVICE_HOST")
q = rediswq.RedisWQ(name="job2", host=host)
print("Worker with sessionID: " + q.sessionID())
print("Initial queue state: empty=" + str(q.empty()))
while not q.empty():
item = q.lease(lease_secs=10, block=True, timeout=2)
if item is not None:
itemstr = item.decode("utf-8")
print("Working on " + itemstr)
time.sleep(10) # Put your actual work here instead of sleep.
q.complete(item)
else:
print("Waiting for work")
print("Queue empty, exiting")
You could also download worker.py,
rediswq.py, and
Dockerfile files, then build
the image:
docker build -t job-wq-2 .
For the Docker Hub, tag your app image with
your username and push to the Hub with the below commands. Replace
<username> with your Hub username.
docker tag job-wq-2 <username>/job-wq-2
docker push <username>/job-wq-2
You need to push to a public repository or configure your cluster to be able to access your private repository.
If you are using Google Container
Registry, tag
your app image with your project ID, and push to GCR. Replace
<project> with your project ID.
docker tag job-wq-2 gcr.io/<project>/job-wq-2
gcloud docker -- push gcr.io/<project>/job-wq-2
Here is the job definition:
apiVersion: batch/v1
kind: Job
metadata:
name: job-wq-2
spec:
parallelism: 2
template:
metadata:
name: job-wq-2
spec:
containers:
- name: c
image: gcr.io/myproject/job-wq-2
restartPolicy: OnFailure
Be sure to edit the job template to
change gcr.io/myproject to your own path.
In this example, each pod works on several items from the queue and then exits when there are no more items. Since the workers themselves detect when the workqueue is empty, and the Job controller does not know about the workqueue, it relies on the workers to signal when they are done working. The workers signal that the queue is empty by exiting with success. So, as soon as any worker exits with success, the controller knows the work is done, and the Pods will exit soon. So, we set the completion count of the Job to 1. The job controller will wait for the other pods to complete too.
So, now run the Job:
kubectl apply -f ./job.yaml
Now wait a bit, then check on the job.
kubectl describe jobs/job-wq-2
Name: job-wq-2
Namespace: default
Selector: controller-uid=b1c7e4e3-92e1-11e7-b85e-fa163ee3c11f
Labels: controller-uid=b1c7e4e3-92e1-11e7-b85e-fa163ee3c11f
job-name=job-wq-2
Annotations: <none>
Parallelism: 2
Completions: <unset>
Start Time: Mon, 11 Jan 2016 17:07:59 -0800
Pods Statuses: 1 Running / 0 Succeeded / 0 Failed
Pod Template:
Labels: controller-uid=b1c7e4e3-92e1-11e7-b85e-fa163ee3c11f
job-name=job-wq-2
Containers:
c:
Image: gcr.io/exampleproject/job-wq-2
Port:
Environment: <none>
Mounts: <none>
Volumes: <none>
Events:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
33s 33s 1 {job-controller } Normal SuccessfulCreate Created pod: job-wq-2-lglf8
kubectl logs pods/job-wq-2-7r7b2
Worker with sessionID: bbd72d0a-9e5c-4dd6-abf6-416cc267991f
Initial queue state: empty=False
Working on banana
Working on date
Working on lemon
As you can see, one of our pods worked on several work units.
If running a queue service or modifying your containers to use a work queue is inconvenient, you may want to consider one of the other job patterns.
If you have a continuous stream of background processing work to run, then
consider running your background workers with a ReplicaSet instead,
and consider running a background processing library such as
https://github.com/resque/resque.
Kubernetes v1.22 [beta]
In this example, you will run a Kubernetes Job that uses multiple parallel worker processes. Each worker is a different container running in its own Pod. The Pods have an index number that the control plane sets automatically, which allows each Pod to identify which part of the overall task to work on.
The pod index is available in the annotation
batch.kubernetes.io/job-completion-index as a string representing its
decimal value. In order for the containerized task process to obtain this index,
you can publish the value of the annotation using the downward API
mechanism.
For convenience, the control plane automatically sets the downward API to
expose the index in the JOB_COMPLETION_INDEX environment variable.
Here is an overview of the steps in this example:
Indexed Job based on that manifest.You should already be familiar with the basic, non-parallel, use of Job.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.21. To check the version, enterkubectl version.
To access the work item from the worker program, you have a few options:
JOB_COMPLETION_INDEX environment variable. The Job
controller
automatically links this variable to the annotation containing the completion
index.For this example, imagine that you chose option 3 and you want to run the rev utility. This program accepts a file as an argument and prints its content reversed.
rev data.txt
You'll use the rev tool from the
busybox container image.
As this is only an example, each Pod only does a tiny piece of work (reversing a short string). In a real workload you might, for example, create a Job that represents the task of producing 60 seconds of video based on scene data. Each work item in the video rendering Job would be to render a particular frame of that video clip. Indexed completion would mean that each Pod in the Job knows which frame to render and publish, by counting frames from the start of the clip.
Here is a sample Job manifest that uses Indexed completion mode:
apiVersion: batch/v1
kind: Job
metadata:
name: 'indexed-job'
spec:
completions: 5
parallelism: 3
completionMode: Indexed
template:
spec:
restartPolicy: Never
initContainers:
- name: 'input'
image: 'docker.io/library/bash'
command:
- "bash"
- "-c"
- |
items=(foo bar baz qux xyz)
echo ${items[$JOB_COMPLETION_INDEX]} > /input/data.txt
volumeMounts:
- mountPath: /input
name: input
containers:
- name: 'worker'
image: 'docker.io/library/busybox'
command:
- "rev"
- "/input/data.txt"
volumeMounts:
- mountPath: /input
name: input
volumes:
- name: input
emptyDir: {}
In the example above, you use the builtin JOB_COMPLETION_INDEX environment
variable set by the Job controller for all containers. An init container
maps the index to a static value and writes it to a file that is shared with the
container running the worker through an emptyDir volume.
Optionally, you can define your own environment variable through the downward
API
to publish the index to containers. You can also choose to load a list of values
from a ConfigMap as an environment variable or file.
Alternatively, you can directly use the downward API to pass the annotation value as a volume file, like shown in the following example:
apiVersion: batch/v1
kind: Job
metadata:
name: 'indexed-job'
spec:
completions: 5
parallelism: 3
completionMode: Indexed
template:
spec:
restartPolicy: Never
containers:
- name: 'worker'
image: 'docker.io/library/busybox'
command:
- "rev"
- "/input/data.txt"
volumeMounts:
- mountPath: /input
name: input
volumes:
- name: input
downwardAPI:
items:
- path: "data.txt"
fieldRef:
fieldPath: metadata.annotations['batch.kubernetes.io/job-completion-index']Now run the Job:
# This uses the first approach (relying on $JOB_COMPLETION_INDEX)
kubectl apply -f https://kubernetes.io/examples/application/job/indexed-job.yaml
When you create this Job, the control plane creates a series of Pods, one for each index you specified. The value of .spec.parallelism determines how many can run at once whereas .spec.completions determines how many Pods the Job creates in total.
Because .spec.parallelism is less than .spec.completions, the control plane waits for some of the first Pods to complete before starting more of them.
Once you have created the Job, wait a moment then check on progress:
kubectl describe jobs/indexed-job
The output is similar to:
Name: indexed-job
Namespace: default
Selector: controller-uid=bf865e04-0b67-483b-9a90-74cfc4c3e756
Labels: controller-uid=bf865e04-0b67-483b-9a90-74cfc4c3e756
job-name=indexed-job
Annotations: <none>
Parallelism: 3
Completions: 5
Start Time: Thu, 11 Mar 2021 15:47:34 +0000
Pods Statuses: 2 Running / 3 Succeeded / 0 Failed
Completed Indexes: 0-2
Pod Template:
Labels: controller-uid=bf865e04-0b67-483b-9a90-74cfc4c3e756
job-name=indexed-job
Init Containers:
input:
Image: docker.io/library/bash
Port: <none>
Host Port: <none>
Command:
bash
-c
items=(foo bar baz qux xyz)
echo ${items[$JOB_COMPLETION_INDEX]} > /input/data.txt
Environment: <none>
Mounts:
/input from input (rw)
Containers:
worker:
Image: docker.io/library/busybox
Port: <none>
Host Port: <none>
Command:
rev
/input/data.txt
Environment: <none>
Mounts:
/input from input (rw)
Volumes:
input:
Type: EmptyDir (a temporary directory that shares a pod's lifetime)
Medium:
SizeLimit: <unset>
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal SuccessfulCreate 4s job-controller Created pod: indexed-job-njkjj
Normal SuccessfulCreate 4s job-controller Created pod: indexed-job-9kd4h
Normal SuccessfulCreate 4s job-controller Created pod: indexed-job-qjwsz
Normal SuccessfulCreate 1s job-controller Created pod: indexed-job-fdhq5
Normal SuccessfulCreate 1s job-controller Created pod: indexed-job-ncslj
In this example, you run the Job with custom values for each index. You can inspect the output of one of the pods:
kubectl logs indexed-job-fdhq5 # Change this to match the name of a Pod from that Job
The output is similar to:
xuq
This task demonstrates running multiple Jobs based on a common template. You can use this approach to process batches of work in parallel.
For this example there are only three items: apple, banana, and cherry. The sample Jobs process each item by printing a string then pausing.
See using Jobs in real workloads to learn about how this pattern fits more realistic use cases.
You should be familiar with the basic, non-parallel, use of Job.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
For basic templating you need the command-line utility sed.
To follow the advanced templating example, you need a working installation of Python, and the Jinja2 template library for Python.
Once you have Python set up, you can install Jinja2 by running:
pip install --user jinja2
First, download the following template of a Job to a file called job-tmpl.yaml.
Here's what you'll download:
apiVersion: batch/v1
kind: Job
metadata:
name: process-item-$ITEM
labels:
jobgroup: jobexample
spec:
template:
metadata:
name: jobexample
labels:
jobgroup: jobexample
spec:
containers:
- name: c
image: busybox
command: ["sh", "-c", "echo Processing item $ITEM && sleep 5"]
restartPolicy: Never
# Use curl to download job-tmpl.yaml
curl -L -s -O https://k8s.io/examples/application/job/job-tmpl.yaml
The file you downloaded is not yet a valid Kubernetes
manifest.
Instead that template is a YAML representation of a Job object with some placeholders
that need to be filled in before it can be used. The $ITEM syntax is not meaningful to Kubernetes.
The following shell snippet uses sed to replace the string $ITEM with the loop
variable, writing into a temporary directory named jobs. Run this now:
# Expand the template into multiple files, one for each item to be processed.
mkdir ./jobs
for i in apple banana cherry
do
cat job-tmpl.yaml | sed "s/\$ITEM/$i/" > ./jobs/job-$i.yaml
done
Check if it worked:
ls jobs/
The output is similar to this:
job-apple.yaml
job-banana.yaml
job-cherry.yaml
You could use any type of template language (for example: Jinja2; ERB), or write a program to generate the Job manifests.
Next, create all the Jobs with one kubectl command:
kubectl create -f ./jobs
The output is similar to this:
job.batch/process-item-apple created
job.batch/process-item-banana created
job.batch/process-item-cherry created
Now, check on the jobs:
kubectl get jobs -l jobgroup=jobexample
The output is similar to this:
NAME COMPLETIONS DURATION AGE
process-item-apple 1/1 14s 22s
process-item-banana 1/1 12s 21s
process-item-cherry 1/1 12s 20s
Using the -l option to kubectl selects only the Jobs that are part
of this group of jobs (there might be other unrelated jobs in the system).
You can check on the Pods as well using the same label selector:
kubectl get pods -l jobgroup=jobexample
The output is similar to:
NAME READY STATUS RESTARTS AGE
process-item-apple-kixwv 0/1 Completed 0 4m
process-item-banana-wrsf7 0/1 Completed 0 4m
process-item-cherry-dnfu9 0/1 Completed 0 4m
We can use this single command to check on the output of all jobs at once:
kubectl logs -f -l jobgroup=jobexample
The output should be:
Processing item apple
Processing item banana
Processing item cherry
# Remove the Jobs you created
# Your cluster automatically cleans up their Pods
kubectl delete job -l jobgroup=jobexample
In the first example, each instance of the template had one parameter, and that parameter was also used in the Job's name. However, names are restricted to contain only certain characters.
This slightly more complex example uses the Jinja template language to generate manifests and then objects from those manifests, with a multiple parameters for each Job.
For this part of the task, you are going to use a one-line Python script to convert the template to a set of manifests.
First, copy and paste the following template of a Job object, into a file called job.yaml.jinja2:
{%- set params = [{ "name": "apple", "url": "http://dbpedia.org/resource/Apple", },
{ "name": "banana", "url": "http://dbpedia.org/resource/Banana", },
{ "name": "cherry", "url": "http://dbpedia.org/resource/Cherry" }]
%}
{%- for p in params %}
{%- set name = p["name"] %}
{%- set url = p["url"] %}
---
apiVersion: batch/v1
kind: Job
metadata:
name: jobexample-{{ name }}
labels:
jobgroup: jobexample
spec:
template:
metadata:
name: jobexample
labels:
jobgroup: jobexample
spec:
containers:
- name: c
image: busybox
command: ["sh", "-c", "echo Processing URL {{ url }} && sleep 5"]
restartPolicy: Never
{%- endfor %}
The above template defines two parameters for each Job object using a list of
python dicts (lines 1-4). A for loop emits one Job manifest for each
set of parameters (remaining lines).
This example relies on a feature of YAML. One YAML file can contain multiple
documents (Kubernetes manifests, in this case), separated by --- on a line
by itself.
You can pipe the output directly to kubectl to create the Jobs.
Next, use this one-line Python program to expand the template:
alias render_template='python -c "from jinja2 import Template; import sys; print(Template(sys.stdin.read()).render());"'
Use render_template to convert the parameters and template into a single
YAML file containing Kubernetes manifests:
# This requires the alias you defined earlier
cat job.yaml.jinja2 | render_template > jobs.yaml
You can view jobs.yaml to verify that the render_template script worked
correctly.
Once you are happy that render_template is working how you intend,
you can pipe its output into kubectl:
cat job.yaml.jinja2 | render_template | kubectl apply -f -
Kubernetes accepts and runs the Jobs you created.
# Remove the Jobs you created
# Your cluster automatically cleans up their Pods
kubectl delete job -l jobgroup=jobexample
In a real use case, each Job performs some substantial computation, such as rendering a frame
of a movie, or processing a range of rows in a database. If you were rendering a movie
you would set $ITEM to the frame number. If you were processing rows from a database
table, you would set $ITEM to represent the range of database rows to process.
In the task, you ran a command to collect the output from Pods by fetching
their logs. In a real use case, each Pod for a Job writes its output to
durable storage before completing. You can use a PersistentVolume for each Job,
or an external storage service. For example, if you are rendering frames for a movie,
use HTTP to PUT the rendered frame data to a URL, using a different URL for each
frame.
After you create a Job, Kubernetes automatically adds additional labels that distinguish one Job's pods from another Job's pods.
In this example, each Job and its Pod template have a label:
jobgroup=jobexample.
Kubernetes itself pays no attention to labels named jobgroup. Setting a label
for all the Jobs you create from a template makes it convenient to operate on all
those Jobs at once.
In the first example you used a template to
create several Jobs. The template ensures that each Pod also gets the same label, so
you can check on all Pods for these templated Jobs with a single command.
jobgroup is not special or reserved.
You can pick your own labelling scheme.
There are recommended labels
that you can use if you wish.
If you plan to create a large number of Job objects, you may find that:
There are other job patterns that you can use to process large amounts of work without creating very many Job objects.
You could also consider writing your own controller to manage Job objects automatically.
Dashboard is a web-based Kubernetes user interface. You can use Dashboard to deploy containerized applications to a Kubernetes cluster, troubleshoot your containerized application, and manage the cluster resources. You can use Dashboard to get an overview of applications running on your cluster, as well as for creating or modifying individual Kubernetes resources (such as Deployments, Jobs, DaemonSets, etc). For example, you can scale a Deployment, initiate a rolling update, restart a pod or deploy new applications using a deploy wizard.
Dashboard also provides information on the state of Kubernetes resources in your cluster and on any errors that may have occurred.
The Dashboard UI is not deployed by default. To deploy it, run the following command:
kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v2.3.1/aio/deploy/recommended.yaml
To protect your cluster data, Dashboard deploys with a minimal RBAC configuration by default. Currently, Dashboard only supports logging in with a Bearer Token. To create a token for this demo, you can follow our guide on creating a sample user.
You can enable access to the Dashboard using the kubectl command-line tool,
by running the following command:
kubectl proxy
Kubectl will make Dashboard available at http://localhost:8001/api/v1/namespaces/kubernetes-dashboard/services/https:kubernetes-dashboard:/proxy/.
The UI can only be accessed from the machine where the command is executed. See kubectl proxy --help for more options.
When you access Dashboard on an empty cluster, you'll see the welcome page.
This page contains a link to this document as well as a button to deploy your first application.
In addition, you can view which system applications are running by default in the kube-system
namespace of your cluster, for example the Dashboard itself.
Dashboard lets you create and deploy a containerized application as a Deployment and optional Service with a simple wizard. You can either manually specify application details, or upload a YAML or JSON manifest file containing application configuration.
Click the CREATE button in the upper right corner of any page to begin.
The deploy wizard expects that you provide the following information:
App name (mandatory): Name for your application. A label with the name will be added to the Deployment and Service, if any, that will be deployed.
The application name must be unique within the selected Kubernetes namespace. It must start with a lowercase character, and end with a lowercase character or a number, and contain only lowercase letters, numbers and dashes (-). It is limited to 24 characters. Leading and trailing spaces are ignored.
Container image (mandatory): The URL of a public Docker container image on any registry, or a private image (commonly hosted on the Google Container Registry or Docker Hub). The container image specification must end with a colon.
Number of pods (mandatory): The target number of Pods you want your application to be deployed in. The value must be a positive integer.
A Deployment will be created to maintain the desired number of Pods across your cluster.
Service (optional): For some parts of your application (e.g. frontends) you may want to expose a Service onto an external, maybe public IP address outside of your cluster (external Service).
Other Services that are only visible from inside the cluster are called internal Services.
Irrespective of the Service type, if you choose to create a Service and your container listens on a port (incoming), you need to specify two ports. The Service will be created mapping the port (incoming) to the target port seen by the container. This Service will route to your deployed Pods. Supported protocols are TCP and UDP. The internal DNS name for this Service will be the value you specified as application name above.
If needed, you can expand the Advanced options section where you can specify more settings:
Description: The text you enter here will be added as an annotation to the Deployment and displayed in the application's details.
Labels: Default labels to be used for your application are application name and version. You can specify additional labels to be applied to the Deployment, Service (if any), and Pods, such as release, environment, tier, partition, and release track.
Example:
release=1.0
tier=frontend
environment=pod
track=stable
Namespace: Kubernetes supports multiple virtual clusters backed by the same physical cluster. These virtual clusters are called namespaces. They let you partition resources into logically named groups.
Dashboard offers all available namespaces in a dropdown list, and allows you to create a new namespace. The namespace name may contain a maximum of 63 alphanumeric characters and dashes (-) but can not contain capital letters. Namespace names should not consist of only numbers. If the name is set as a number, such as 10, the pod will be put in the default namespace.
In case the creation of the namespace is successful, it is selected by default. If the creation fails, the first namespace is selected.
Image Pull Secret: In case the specified Docker container image is private, it may require pull secret credentials.
Dashboard offers all available secrets in a dropdown list, and allows you to create a new secret.
The secret name must follow the DNS domain name syntax, for example new.image-pull.secret.
The content of a secret must be base64-encoded and specified in a
.dockercfg file.
The secret name may consist of a maximum of 253 characters.
In case the creation of the image pull secret is successful, it is selected by default. If the creation fails, no secret is applied.
CPU requirement (cores) and Memory requirement (MiB): You can specify the minimum resource limits for the container. By default, Pods run with unbounded CPU and memory limits.
Run command and Run command arguments: By default, your containers run the specified Docker image's default entrypoint command. You can use the command options and arguments to override the default.
Run as privileged: This setting determines whether processes in privileged containers are equivalent to processes running as root on the host. Privileged containers can make use of capabilities like manipulating the network stack and accessing devices.
Environment variables: Kubernetes exposes Services through
environment variables.
You can compose environment variable or pass arguments to your commands using the values of environment variables.
They can be used in applications to find a Service.
Values can reference other variables using the $(VAR_NAME) syntax.
Kubernetes supports declarative configuration. In this style, all configuration is stored in manifests (YAML or JSON configuration files). The manifests use Kubernetes API resource schemas.
As an alternative to specifying application details in the deploy wizard, you can define your application in one or more manifests, and upload the files using Dashboard.
Following sections describe views of the Kubernetes Dashboard UI; what they provide and how can they be used.
When there are Kubernetes objects defined in the cluster, Dashboard shows them in the initial view. By default only objects from the default namespace are shown and this can be changed using the namespace selector located in the navigation menu.
Dashboard shows most Kubernetes object kinds and groups them in a few menu categories.
For cluster and namespace administrators, Dashboard lists Nodes, Namespaces and PersistentVolumes and has detail views for them. Node list view contains CPU and memory usage metrics aggregated across all Nodes. The details view shows the metrics for a Node, its specification, status, allocated resources, events and pods running on the node.
Shows all applications running in the selected namespace. The view lists applications by workload kind (for example: Deployments, ReplicaSets, StatefulSets). and each workload kind can be viewed separately. The lists summarize actionable information about the workloads, such as the number of ready pods for a ReplicaSet or current memory usage for a Pod.
Detail views for workloads show status and specification information and surface relationships between objects. For example, Pods that ReplicaSet is controlling or new ReplicaSets and HorizontalPodAutoscalers for Deployments.
Shows Kubernetes resources that allow for exposing services to external world and discovering them within a cluster. For that reason, Service and Ingress views show Pods targeted by them, internal endpoints for cluster connections and external endpoints for external users.
Storage view shows PersistentVolumeClaim resources which are used by applications for storing data.
Shows all Kubernetes resources that are used for live configuration of applications running in clusters. The view allows for editing and managing config objects and displays secrets hidden by default.
Pod lists and detail pages link to a logs viewer that is built into Dashboard. The viewer allows for drilling down logs from containers belonging to a single Pod.
For more information, see the Kubernetes Dashboard project page.
This topic discusses multiple ways to interact with clusters.
When accessing the Kubernetes API for the first time, we suggest using the
Kubernetes CLI, kubectl.
To access a cluster, you need to know the location of the cluster and have credentials to access it. Typically, this is automatically set-up when you work through a Getting started guide, or someone else setup the cluster and provided you with credentials and a location.
Check the location and credentials that kubectl knows about with this command:
kubectl config view
Many of the examples provide an introduction to using kubectl and complete documentation is found in the kubectl manual.
Kubectl handles locating and authenticating to the apiserver. If you want to directly access the REST API with an http client like curl or wget, or a browser, there are several ways to locate and authenticate:
The following command runs kubectl in a mode where it acts as a reverse proxy. It handles locating the apiserver and authenticating. Run it like this:
kubectl proxy --port=8080
See kubectl proxy for more details.
Then you can explore the API with curl, wget, or a browser, replacing localhost with [::1] for IPv6, like so:
curl http://localhost:8080/api/
The output is similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
Use kubectl describe secret... to get the token for the default service account with grep/cut:
APISERVER=$(kubectl config view --minify | grep server | cut -f 2- -d ":" | tr -d " ")
SECRET_NAME=$(kubectl get secrets | grep ^default | cut -f1 -d ' ')
TOKEN=$(kubectl describe secret $SECRET_NAME | grep -E '^token' | cut -f2 -d':' | tr -d " ")
curl $APISERVER/api --header "Authorization: Bearer $TOKEN" --insecure
The output is similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
Using jsonpath:
APISERVER=$(kubectl config view --minify -o jsonpath='{.clusters[0].cluster.server}')
SECRET_NAME=$(kubectl get serviceaccount default -o jsonpath='{.secrets[0].name}')
TOKEN=$(kubectl get secret $SECRET_NAME -o jsonpath='{.data.token}' | base64 --decode)
curl $APISERVER/api --header "Authorization: Bearer $TOKEN" --insecure
The output is similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.1.149:443"
}
]
}
The above examples use the --insecure flag. This leaves it subject to MITM
attacks. When kubectl accesses the cluster it uses a stored root certificate
and client certificates to access the server. (These are installed in the
~/.kube directory). Since cluster certificates are typically self-signed, it
may take special configuration to get your http client to use root
certificate.
On some clusters, the apiserver does not require authentication; it may serve on localhost, or be protected by a firewall. There is not a standard for this. Controlling Access to the API describes how a cluster admin can configure this.
Kubernetes officially supports Go and Python client libraries.
go get k8s.io/client-go@kubernetes-<kubernetes-version-number>, see INSTALL.md for detailed installation instructions. See https://github.com/kubernetes/client-go to see which versions are supported.import "k8s.io/client-go/kubernetes" is correct.The Go client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the apiserver. See this example.
If the application is deployed as a Pod in the cluster, please refer to the next section.
To use Python client, run the following command: pip install kubernetes. See Python Client Library page for more installation options.
The Python client can use the same kubeconfig file as the kubectl CLI does to locate and authenticate to the apiserver. See this example.
There are client libraries for accessing the API from other languages. See documentation for other libraries for how they authenticate.
When accessing the API from a pod, locating and authenticating to the apiserver are somewhat different.
The recommended way to locate the apiserver within the pod is with
the kubernetes.default.svc DNS name, which resolves to a Service IP which in turn
will be routed to an apiserver.
The recommended way to authenticate to the apiserver is with a
service account credential. By kube-system, a pod
is associated with a service account, and a credential (token) for that
service account is placed into the filesystem tree of each container in that pod,
at /var/run/secrets/kubernetes.io/serviceaccount/token.
If available, a certificate bundle is placed into the filesystem tree of each
container at /var/run/secrets/kubernetes.io/serviceaccount/ca.crt, and should be
used to verify the serving certificate of the apiserver.
Finally, the default namespace to be used for namespaced API operations is placed in a file
at /var/run/secrets/kubernetes.io/serviceaccount/namespace in each container.
From within a pod the recommended ways to connect to API are:
kubectl proxy in a sidecar container in the pod, or as a background
process within the container. This proxies the
Kubernetes API to the localhost interface of the pod, so that other processes
in any container of the pod can access it.rest.InClusterConfig() and kubernetes.NewForConfig() functions.
They handle locating and authenticating to the apiserver. exampleIn each case, the credentials of the pod are used to communicate securely with the apiserver.
The previous section was about connecting the Kubernetes API server. This section is about connecting to other services running on Kubernetes cluster. In Kubernetes, the nodes, pods and services all have their own IPs. In many cases, the node IPs, pod IPs, and some service IPs on a cluster will not be routable, so they will not be reachable from a machine outside the cluster, such as your desktop machine.
You have several options for connecting to nodes, pods and services from outside the cluster:
NodePort or LoadBalancer to make the service reachable outside
the cluster. See the services and
kubectl expose documentation.Typically, there are several services which are started on a cluster by kube-system. Get a list of these
with the kubectl cluster-info command:
kubectl cluster-info
The output is similar to this:
Kubernetes master is running at https://104.197.5.247
elasticsearch-logging is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy
kibana-logging is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/kibana-logging/proxy
kube-dns is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/kube-dns/proxy
grafana is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/monitoring-grafana/proxy
heapster is running at https://104.197.5.247/api/v1/namespaces/kube-system/services/monitoring-heapster/proxy
This shows the proxy-verb URL for accessing each service.
For example, this cluster has cluster-level logging enabled (using Elasticsearch), which can be reached
at https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/ if suitable credentials are passed. Logging can also be reached through a kubectl proxy, for example at:
http://localhost:8080/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/.
(See Access Clusters Using the Kubernetes API for how to pass credentials or use kubectl proxy.)
As mentioned above, you use the kubectl cluster-info command to retrieve the service's proxy URL. To create proxy URLs that include service endpoints, suffixes, and parameters, you append to the service's proxy URL:
http://kubernetes_master_address/api/v1/namespaces/namespace_name/services/service_name[:port_name]/proxy
If you haven't specified a name for your port, you don't have to specify port_name in the URL. You can also use the port number in place of the port_name for both named and unnamed ports.
By default, the API server proxies to your service using http. To use https, prefix the service name with https::
http://kubernetes_master_address/api/v1/namespaces/namespace_name/services/https:service_name:[port_name]/proxy
The supported formats for the name segment of the URL are:
<service_name> - proxies to the default or unnamed port using http<service_name>:<port_name> - proxies to the specified port name or port number using httphttps:<service_name>: - proxies to the default or unnamed port using https (note the trailing colon)https:<service_name>:<port_name> - proxies to the specified port name or port number using https_search?q=user:kimchy, you would use: http://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/_search?q=user:kimchy_cluster/health?pretty=true, you would use: https://104.197.5.247/api/v1/namespaces/kube-system/services/elasticsearch-logging/proxy/_cluster/health?pretty=true{
"cluster_name" : "kubernetes_logging",
"status" : "yellow",
"timed_out" : false,
"number_of_nodes" : 1,
"number_of_data_nodes" : 1,
"active_primary_shards" : 5,
"active_shards" : 5,
"relocating_shards" : 0,
"initializing_shards" : 0,
"unassigned_shards" : 5
}
You may be able to put an apiserver proxy url into the address bar of a browser. However:
The redirect capabilities have been deprecated and removed. Please use a proxy (see below) instead.
There are several different proxies you may encounter when using Kubernetes:
The kubectl proxy:
The apiserver proxy:
The kube proxy:
A Proxy/Load-balancer in front of apiserver(s):
Cloud Load Balancers on external services:
LoadBalancerKubernetes users will typically not need to worry about anything other than the first two types. The cluster admin will typically ensure that the latter types are setup correctly.
This page shows how to configure access to multiple clusters by using
configuration files. After your clusters, users, and contexts are defined in
one or more configuration files, you can quickly switch between clusters by using the
kubectl config use-context command.
kubeconfig.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check that kubectl is installed,
run kubectl version --client. The kubectl version should be
within one minor version of your
cluster's API server.
Suppose you have two clusters, one for development work and one for scratch work.
In the development cluster, your frontend developers work in a namespace called frontend,
and your storage developers work in a namespace called storage. In your scratch cluster,
developers work in the default namespace, or they create auxiliary namespaces as they
see fit. Access to the development cluster requires authentication by certificate. Access
to the scratch cluster requires authentication by username and password.
Create a directory named config-exercise. In your
config-exercise directory, create a file named config-demo with this content:
apiVersion: v1
kind: Config
preferences: {}
clusters:
- cluster:
name: development
- cluster:
name: scratch
users:
- name: developer
- name: experimenter
contexts:
- context:
name: dev-frontend
- context:
name: dev-storage
- context:
name: exp-scratch
A configuration file describes clusters, users, and contexts. Your config-demo file
has the framework to describe two clusters, two users, and three contexts.
Go to your config-exercise directory. Enter these commands to add cluster details to
your configuration file:
kubectl config --kubeconfig=config-demo set-cluster development --server=https://1.2.3.4 --certificate-authority=fake-ca-file
kubectl config --kubeconfig=config-demo set-cluster scratch --server=https://5.6.7.8 --insecure-skip-tls-verify
Add user details to your configuration file:
kubectl config --kubeconfig=config-demo set-credentials developer --client-certificate=fake-cert-file --client-key=fake-key-seefile
kubectl config --kubeconfig=config-demo set-credentials experimenter --username=exp --password=some-password
kubectl --kubeconfig=config-demo config unset users.<name>kubectl --kubeconfig=config-demo config unset clusters.<name>kubectl --kubeconfig=config-demo config unset contexts.<name>Add context details to your configuration file:
kubectl config --kubeconfig=config-demo set-context dev-frontend --cluster=development --namespace=frontend --user=developer
kubectl config --kubeconfig=config-demo set-context dev-storage --cluster=development --namespace=storage --user=developer
kubectl config --kubeconfig=config-demo set-context exp-scratch --cluster=scratch --namespace=default --user=experimenter
Open your config-demo file to see the added details. As an alternative to opening the
config-demo file, you can use the config view command.
kubectl config --kubeconfig=config-demo view
The output shows the two clusters, two users, and three contexts:
apiVersion: v1
clusters:
- cluster:
certificate-authority: fake-ca-file
server: https://1.2.3.4
name: development
- cluster:
insecure-skip-tls-verify: true
server: https://5.6.7.8
name: scratch
contexts:
- context:
cluster: development
namespace: frontend
user: developer
name: dev-frontend
- context:
cluster: development
namespace: storage
user: developer
name: dev-storage
- context:
cluster: scratch
namespace: default
user: experimenter
name: exp-scratch
current-context: ""
kind: Config
preferences: {}
users:
- name: developer
user:
client-certificate: fake-cert-file
client-key: fake-key-file
- name: experimenter
user:
password: some-password
username: exp
The fake-ca-file, fake-cert-file and fake-key-file above are the placeholders
for the pathnames of the certificate files. You need to change these to the actual pathnames
of certificate files in your environment.
Sometimes you may want to use Base64-encoded data embedded here instead of separate
certificate files; in that case you need to add the suffix -data to the keys, for example,
certificate-authority-data, client-certificate-data, client-key-data.
Each context is a triple (cluster, user, namespace). For example, the
dev-frontend context says, "Use the credentials of the developer
user to access the frontend namespace of the development cluster".
Set the current context:
kubectl config --kubeconfig=config-demo use-context dev-frontend
Now whenever you enter a kubectl command, the action will apply to the cluster,
and namespace listed in the dev-frontend context. And the command will use
the credentials of the user listed in the dev-frontend context.
To see only the configuration information associated with
the current context, use the --minify flag.
kubectl config --kubeconfig=config-demo view --minify
The output shows configuration information associated with the dev-frontend context:
apiVersion: v1
clusters:
- cluster:
certificate-authority: fake-ca-file
server: https://1.2.3.4
name: development
contexts:
- context:
cluster: development
namespace: frontend
user: developer
name: dev-frontend
current-context: dev-frontend
kind: Config
preferences: {}
users:
- name: developer
user:
client-certificate: fake-cert-file
client-key: fake-key-file
Now suppose you want to work for a while in the scratch cluster.
Change the current context to exp-scratch:
kubectl config --kubeconfig=config-demo use-context exp-scratch
Now any kubectl command you give will apply to the default namespace of
the scratch cluster. And the command will use the credentials of the user
listed in the exp-scratch context.
View configuration associated with the new current context, exp-scratch.
kubectl config --kubeconfig=config-demo view --minify
Finally, suppose you want to work for a while in the storage namespace of the
development cluster.
Change the current context to dev-storage:
kubectl config --kubeconfig=config-demo use-context dev-storage
View configuration associated with the new current context, dev-storage.
kubectl config --kubeconfig=config-demo view --minify
In your config-exercise directory, create a file named config-demo-2 with this content:
apiVersion: v1
kind: Config
preferences: {}
contexts:
- context:
cluster: development
namespace: ramp
user: developer
name: dev-ramp-up
The preceding configuration file defines a new context named dev-ramp-up.
See whether you have an environment variable named KUBECONFIG. If so, save the
current value of your KUBECONFIG environment variable, so you can restore it later.
For example:
export KUBECONFIG_SAVED=$KUBECONFIG
$Env:KUBECONFIG_SAVED=$ENV:KUBECONFIG
The KUBECONFIG environment variable is a list of paths to configuration files. The list is
colon-delimited for Linux and Mac, and semicolon-delimited for Windows. If you have
a KUBECONFIG environment variable, familiarize yourself with the configuration files
in the list.
Temporarily append two paths to your KUBECONFIG environment variable. For example:
export KUBECONFIG=$KUBECONFIG:config-demo:config-demo-2
$Env:KUBECONFIG=("config-demo;config-demo-2")
In your config-exercise directory, enter this command:
kubectl config view
The output shows merged information from all the files listed in your KUBECONFIG
environment variable. In particular, notice that the merged information has the
dev-ramp-up context from the config-demo-2 file and the three contexts from
the config-demo file:
contexts:
- context:
cluster: development
namespace: frontend
user: developer
name: dev-frontend
- context:
cluster: development
namespace: ramp
user: developer
name: dev-ramp-up
- context:
cluster: development
namespace: storage
user: developer
name: dev-storage
- context:
cluster: scratch
namespace: default
user: experimenter
name: exp-scratch
For more information about how kubeconfig files are merged, see Organizing Cluster Access Using kubeconfig Files
If you already have a cluster, and you can use kubectl to interact with
the cluster, then you probably have a file named config in the $HOME/.kube
directory.
Go to $HOME/.kube, and see what files are there. Typically, there is a file named
config. There might also be other configuration files in this directory. Briefly
familiarize yourself with the contents of these files.
If you have a $HOME/.kube/config file, and it's not already listed in your
KUBECONFIG environment variable, append it to your KUBECONFIG environment variable now.
For example:
export KUBECONFIG=$KUBECONFIG:$HOME/.kube/config
$Env:KUBECONFIG="$Env:KUBECONFIG;$HOME\.kube\config"
View configuration information merged from all the files that are now listed
in your KUBECONFIG environment variable. In your config-exercise directory, enter:
kubectl config view
Return your KUBECONFIG environment variable to its original value. For example:
export KUBECONFIG=$KUBECONFIG_SAVED
$Env:KUBECONFIG=$ENV:KUBECONFIG_SAVED
This page shows how to use kubectl port-forward to connect to a MongoDB
server running in a Kubernetes cluster. This type of connection can be useful
for database debugging.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.10. To check the version, enterkubectl version.
Install MongoDB Shell.
Create a Deployment that runs MongoDB:
kubectl apply -f https://k8s.io/examples/application/mongodb/mongo-deployment.yaml
The output of a successful command verifies that the deployment was created:
deployment.apps/mongo created
View the pod status to check that it is ready:
kubectl get pods
The output displays the pod created:
NAME READY STATUS RESTARTS AGE
mongo-75f59d57f4-4nd6q 1/1 Running 0 2m4s
View the Deployment's status:
kubectl get deployment
The output displays that the Deployment was created:
NAME READY UP-TO-DATE AVAILABLE AGE
mongo 1/1 1 1 2m21s
The Deployment automatically manages a ReplicaSet. View the ReplicaSet status using:
kubectl get replicaset
The output displays that the ReplicaSet was created:
NAME DESIRED CURRENT READY AGE
mongo-75f59d57f4 1 1 1 3m12s
Create a Service to expose MongoDB on the network:
kubectl apply -f https://k8s.io/examples/application/mongodb/mongo-service.yaml
The output of a successful command verifies that the Service was created:
service/mongo created
Check the Service created:
kubectl get service mongo
The output displays the service created:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
mongo ClusterIP 10.96.41.183 <none> 27017/TCP 11s
Verify that the MongoDB server is running in the Pod, and listening on port 27017:
# Change mongo-75f59d57f4-4nd6q to the name of the Pod
kubectl get pod mongo-75f59d57f4-4nd6q --template='{{(index (index .spec.containers 0).ports 0).containerPort}}{{"\n"}}'
The output displays the port for MongoDB in that Pod:
27017
(this is the TCP port allocated to MongoDB on the internet).
kubectl port-forward allows using resource name, such as a pod name, to select a matching pod to port forward to.
# Change mongo-75f59d57f4-4nd6q to the name of the Pod
kubectl port-forward mongo-75f59d57f4-4nd6q 28015:27017
which is the same as
kubectl port-forward pods/mongo-75f59d57f4-4nd6q 28015:27017
or
kubectl port-forward deployment/mongo 28015:27017
or
kubectl port-forward replicaset/mongo-75f59d57f4 28015:27017
or
kubectl port-forward service/mongo 28015:27017
Any of the above commands works. The output is similar to this:
Forwarding from 127.0.0.1:28015 -> 27017
Forwarding from [::1]:28015 -> 27017
kubectl port-forward does not return. To continue with the exercises, you will need to open another terminal.
Start the MongoDB command line interface:
mongosh --port 28015
At the MongoDB command line prompt, enter the ping command:
db.runCommand( { ping: 1 } )
A successful ping request returns:
{ ok: 1 }
If you don't need a specific local port, you can let kubectl choose and allocate
the local port and thus relieve you from having to manage local port conflicts, with
the slightly simpler syntax:
kubectl port-forward deployment/mongo :27017
The kubectl tool finds a local port number that is not in use (avoiding low ports numbers,
because these might be used by other applications). The output is similar to:
Forwarding from 127.0.0.1:63753 -> 27017
Forwarding from [::1]:63753 -> 27017
Connections made to local port 28015 are forwarded to port 27017 of the Pod that is running the MongoDB server. With this connection in place, you can use your local workstation to debug the database that is running in the Pod.
kubectl port-forward is implemented for TCP ports only.
The support for UDP protocol is tracked in
issue 47862.
Learn more about kubectl port-forward.
This page shows how to create a Kubernetes Service object that external clients can use to access an application running in a cluster. The Service provides load balancing for an application that has two running instances.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Here is the configuration file for the application Deployment:
apiVersion: apps/v1
kind: Deployment
metadata:
name: hello-world
spec:
selector:
matchLabels:
run: load-balancer-example
replicas: 2
template:
metadata:
labels:
run: load-balancer-example
spec:
containers:
- name: hello-world
image: gcr.io/google-samples/node-hello:1.0
ports:
- containerPort: 8080
protocol: TCP
Run a Hello World application in your cluster: Create the application Deployment using the file above:
kubectl apply -f https://k8s.io/examples/service/access/hello-application.yaml
The preceding command creates a Deployment and an associated ReplicaSet. The ReplicaSet has two Pods each of which runs the Hello World application.
Display information about the Deployment:
kubectl get deployments hello-world
kubectl describe deployments hello-world
Display information about your ReplicaSet objects:
kubectl get replicasets
kubectl describe replicasets
Create a Service object that exposes the deployment:
kubectl expose deployment hello-world --type=NodePort --name=example-service
Display information about the Service:
kubectl describe services example-service
The output is similar to this:
Name: example-service
Namespace: default
Labels: run=load-balancer-example
Annotations: <none>
Selector: run=load-balancer-example
Type: NodePort
IP: 10.32.0.16
Port: <unset> 8080/TCP
TargetPort: 8080/TCP
NodePort: <unset> 31496/TCP
Endpoints: 10.200.1.4:8080,10.200.2.5:8080
Session Affinity: None
Events: <none>
Make a note of the NodePort value for the service. For example, in the preceding output, the NodePort value is 31496.
List the pods that are running the Hello World application:
kubectl get pods --selector="run=load-balancer-example" --output=wide
The output is similar to this:
NAME READY STATUS ... IP NODE
hello-world-2895499144-bsbk5 1/1 Running ... 10.200.1.4 worker1
hello-world-2895499144-m1pwt 1/1 Running ... 10.200.2.5 worker2
Get the public IP address of one of your nodes that is running
a Hello World pod. How you get this address depends on how you set
up your cluster. For example, if you are using Minikube, you can
see the node address by running kubectl cluster-info. If you are
using Google Compute Engine instances, you can use the
gcloud compute instances list command to see the public addresses of your
nodes.
On your chosen node, create a firewall rule that allows TCP traffic on your node port. For example, if your Service has a NodePort value of 31568, create a firewall rule that allows TCP traffic on port 31568. Different cloud providers offer different ways of configuring firewall rules.
Use the node address and node port to access the Hello World application:
curl http://<public-node-ip>:<node-port>
where <public-node-ip> is the public IP address of your node,
and <node-port> is the NodePort value for your service. The
response to a successful request is a hello message:
Hello Kubernetes!
As an alternative to using kubectl expose, you can use a
service configuration file
to create a Service.
To delete the Service, enter this command:
kubectl delete services example-service
To delete the Deployment, the ReplicaSet, and the Pods that are running the Hello World application, enter this command:
kubectl delete deployment hello-world
Learn more about connecting applications with services.
This task shows how to create a frontend and a backend microservice. The backend microservice is a hello greeter. The frontend exposes the backend using nginx and a Kubernetes Service object.
hello backend microservice using a
Deployment object.nginx frontend microservice, also using a Deployment object.type=LoadBalancer to expose the frontend microservice
outside the cluster.You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
This task uses Services with external load balancers, which require a supported environment. If your environment does not support this, you can use a Service of type NodePort instead.
The backend is a simple hello greeter microservice. Here is the configuration file for the backend Deployment:
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: backend
spec:
selector:
matchLabels:
app: hello
tier: backend
track: stable
replicas: 3
template:
metadata:
labels:
app: hello
tier: backend
track: stable
spec:
containers:
- name: hello
image: "gcr.io/google-samples/hello-go-gke:1.0"
ports:
- name: http
containerPort: 80
...Create the backend Deployment:
kubectl apply -f https://k8s.io/examples/service/access/backend-deployment.yaml
View information about the backend Deployment:
kubectl describe deployment backend
The output is similar to this:
Name: backend
Namespace: default
CreationTimestamp: Mon, 24 Oct 2016 14:21:02 -0700
Labels: app=hello
tier=backend
track=stable
Annotations: deployment.kubernetes.io/revision=1
Selector: app=hello,tier=backend,track=stable
Replicas: 3 desired | 3 updated | 3 total | 3 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 1 max unavailable, 1 max surge
Pod Template:
Labels: app=hello
tier=backend
track=stable
Containers:
hello:
Image: "gcr.io/google-samples/hello-go-gke:1.0"
Port: 80/TCP
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Available True MinimumReplicasAvailable
Progressing True NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet: hello-3621623197 (3/3 replicas created)
Events:
...
hello Service objectThe key to sending requests from a frontend to a backend is the backend Service. A Service creates a persistent IP address and DNS name entry so that the backend microservice can always be reached. A Service uses selectors to find the Pods that it routes traffic to.
First, explore the Service configuration file:
---
apiVersion: v1
kind: Service
metadata:
name: hello
spec:
selector:
app: hello
tier: backend
ports:
- protocol: TCP
port: 80
targetPort: http
...In the configuration file, you can see that the Service, named hello routes
traffic to Pods that have the labels app: hello and tier: backend.
Create the backend Service:
kubectl apply -f https://k8s.io/examples/service/access/backend-service.yaml
At this point, you have a backend Deployment running three replicas of your hello
application, and you have a Service that can route traffic to them. However, this
service is neither available nor resolvable outside the cluster.
Now that you have your backend running, you can create a frontend that is accessible outside the cluster, and connects to the backend by proxying requests to it.
The frontend sends requests to the backend worker Pods by using the DNS name
given to the backend Service. The DNS name is hello, which is the value
of the name field in the examples/service/access/backend-service.yaml
configuration file.
The Pods in the frontend Deployment run a nginx image that is configured
to proxy requests to the hello backend Service. Here is the nginx configuration file:
# The identifier Backend is internal to nginx, and used to name this specific upstream
upstream Backend {
# hello is the internal DNS name used by the backend Service inside Kubernetes
server hello;
}
server {
listen 80;
location / {
# The following statement will proxy traffic to the upstream named Backend
proxy_pass http://Backend;
}
}
Similar to the backend, the frontend has a Deployment and a Service. An important
difference to notice between the backend and frontend services, is that the
configuration for the frontend Service has type: LoadBalancer, which means that
the Service uses a load balancer provisioned by your cloud provider and will be
accessible from outside the cluster.
---
apiVersion: v1
kind: Service
metadata:
name: frontend
spec:
selector:
app: hello
tier: frontend
ports:
- protocol: "TCP"
port: 80
targetPort: 80
type: LoadBalancer
...---
apiVersion: apps/v1
kind: Deployment
metadata:
name: frontend
spec:
selector:
matchLabels:
app: hello
tier: frontend
track: stable
replicas: 1
template:
metadata:
labels:
app: hello
tier: frontend
track: stable
spec:
containers:
- name: nginx
image: "gcr.io/google-samples/hello-frontend:1.0"
lifecycle:
preStop:
exec:
command: ["/usr/sbin/nginx","-s","quit"]
...Create the frontend Deployment and Service:
kubectl apply -f https://k8s.io/examples/service/access/frontend-deployment.yaml
kubectl apply -f https://k8s.io/examples/service/access/frontend-service.yaml
The output verifies that both resources were created:
deployment.apps/frontend created
service/frontend created
Once you've created a Service of type LoadBalancer, you can use this command to find the external IP:
kubectl get service frontend --watch
This displays the configuration for the frontend Service and watches for
changes. Initially, the external IP is listed as <pending>:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
frontend LoadBalancer 10.51.252.116 <pending> 80/TCP 10s
As soon as an external IP is provisioned, however, the configuration updates
to include the new IP under the EXTERNAL-IP heading:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
frontend LoadBalancer 10.51.252.116 XXX.XXX.XXX.XXX 80/TCP 1m
That IP can now be used to interact with the frontend service from outside the
cluster.
The frontend and backend are now connected. You can hit the endpoint by using the curl command on the external IP of your frontend Service.
curl http://${EXTERNAL_IP} # replace this with the EXTERNAL-IP you saw earlier
The output shows the message generated by the backend:
{"message":"Hello"}
To delete the Services, enter this command:
kubectl delete services frontend backend
To delete the Deployments, the ReplicaSets and the Pods that are running the backend and frontend applications, enter this command:
kubectl delete deployment frontend backend
This page shows how to create an external load balancer.
When creating a Service, you have the option of automatically creating a cloud load balancer. This provides an externally-accessible IP address that sends traffic to the correct port on your cluster nodes, provided your cluster runs in a supported environment and is configured with the correct cloud load balancer provider package.
You can also use an Ingress in place of Service. For more information, check the Ingress documentation.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your cluster must be running in a cloud or other environment that already has support for configuring external load balancers.
To create an external load balancer, add the following line to your Service manifest:
type: LoadBalancer
Your manifest might then look like:
apiVersion: v1
kind: Service
metadata:
name: example-service
spec:
selector:
app: example
ports:
- port: 8765
targetPort: 9376
type: LoadBalancer
You can alternatively create the service with the kubectl expose command and
its --type=LoadBalancer flag:
kubectl expose deployment example --port=8765 --target-port=9376 \
--name=example-service --type=LoadBalancer
This command creates a new Service using the same selectors as the referenced
resource (in the case of the example above, a
Deployment named example).
For more information, including optional flags, refer to the
kubectl expose reference.
You can find the IP address created for your service by getting the service
information through kubectl:
kubectl describe services example-service
which should produce output similar to:
Name: example-service
Namespace: default
Labels: app=example
Annotations: <none>
Selector: app=example
Type: LoadBalancer
IP Families: <none>
IP: 10.3.22.96
IPs: 10.3.22.96
LoadBalancer Ingress: 192.0.2.89
Port: <unset> 8765/TCP
TargetPort: 9376/TCP
NodePort: <unset> 30593/TCP
Endpoints: 172.17.0.3:9376
Session Affinity: None
External Traffic Policy: Cluster
Events: <none>
The load balancer's IP address is listed next to LoadBalancer Ingress.
If you are running your service on Minikube, you can find the assigned IP address and port with:
minikube service example-service --url
By default, the source IP seen in the target container is not the original
source IP of the client. To enable preservation of the client IP, the following
fields can be configured in the .spec of the Service:
.spec.externalTrafficPolicy - denotes if this Service desires to route
external traffic to node-local or cluster-wide endpoints. There are two available
options: Cluster (default) and Local. Cluster obscures the client source
IP and may cause a second hop to another node, but should have good overall
load-spreading. Local preserves the client source IP and avoids a second hop
for LoadBalancer and NodePort type Services, but risks potentially imbalanced
traffic spreading..spec.healthCheckNodePort - specifies the health check node port
(numeric port number) for the service. If you don't specify
healthCheckNodePort, the service controller allocates a port from your
cluster's NodePort range.--service-node-port-range. The Service will use the user-specified
healthCheckNodePort value if you specify it, provided that the
Service type is set to LoadBalancer and externalTrafficPolicy is set
to Local.Setting externalTrafficPolicy to Local in the Service manifest
activates this feature. For example:
apiVersion: v1
kind: Service
metadata:
name: example-service
spec:
selector:
app: example
ports:
- port: 8765
targetPort: 9376
externalTrafficPolicy: Local
type: LoadBalancer
Load balancing services from some cloud providers do not let you configure different weights for each target.
With each target weighted equally in terms of sending traffic to Nodes, external traffic is not equally load balanced across different Pods. The external load balancer is unaware of the number of Pods on each node that are used as a target.
Where NumServicePods << _NumNodes or NumServicePods >> NumNodes, a fairly close-to-equal
distribution will be seen, even without weights.
Internal pod to pod traffic should behave similar to ClusterIP services, with equal probability across all pods.
Kubernetes v1.17 [stable]
In usual case, the correlating load balancer resources in cloud provider should be cleaned up soon after a LoadBalancer type Service is deleted. But it is known that there are various corner cases where cloud resources are orphaned after the associated Service is deleted. Finalizer Protection for Service LoadBalancers was introduced to prevent this from happening. By using finalizers, a Service resource will never be deleted until the correlating load balancer resources are also deleted.
Specifically, if a Service has type LoadBalancer, the service controller will attach
a finalizer named service.kubernetes.io/load-balancer-cleanup.
The finalizer will only be removed after the load balancer resource is cleaned up.
This prevents dangling load balancer resources even in corner cases such as the
service controller crashing.
It is important to note that the datapath for this functionality is provided by a load balancer external to the Kubernetes cluster.
When the Service type is set to LoadBalancer, Kubernetes provides functionality equivalent to type equals ClusterIP to pods
within the cluster and extends it by programming the (external to Kubernetes) load balancer with entries for the nodes
hosting the relevant Kubernetes pods. The Kubernetes control plane automates the creation of the external load balancer,
health checks (if needed), and packet filtering rules (if needed). Once the cloud provider allocates an IP address for the load
balancer, the control plane looks up that external IP address and populates it into the Service object.
This page shows how to use kubectl to list all of the Container images for Pods running in a cluster.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise you will use kubectl to fetch all of the Pods running in a cluster, and format the output to pull out the list of Containers for each.
kubectl get pods --all-namespaces-o jsonpath={.items[*].spec.containers[*].image}. This will recursively parse out the
image field from the returned json.
tr, sort, uniq
tr to replace spaces with newlinessort to sort the resultsuniq to aggregate image countskubectl get pods --all-namespaces -o jsonpath="{.items[*].spec.containers[*].image}" |\
tr -s '[[:space:]]' '\n' |\
sort |\
uniq -c
The above command will recursively return all fields named image
for all items returned.
As an alternative, it is possible to use the absolute path to the image
field within the Pod. This ensures the correct field is retrieved
even when the field name is repeated,
e.g. many fields are called name within a given item:
kubectl get pods --all-namespaces -o jsonpath="{.items[*].spec.containers[*].image}"
The jsonpath is interpreted as follows:
.items[*]: for each returned value.spec: get the spec.containers[*]: for each container.image: get the imagekubectl get pod nginx,
the .items[*] portion of the path should be omitted because a single
Pod is returned instead of a list of items.
The formatting can be controlled further by using the range operation to
iterate over elements individually.
kubectl get pods --all-namespaces -o=jsonpath='{range .items[*]}{"\n"}{.metadata.name}{":\t"}{range .spec.containers[*]}{.image}{", "}{end}{end}' |\
sort
To target only Pods matching a specific label, use the -l flag. The
following matches only Pods with labels matching app=nginx.
kubectl get pods --all-namespaces -o=jsonpath="{.items[*].spec.containers[*].image}" -l app=nginx
To target only pods in a specific namespace, use the namespace flag. The
following matches only Pods in the kube-system namespace.
kubectl get pods --namespace kube-system -o jsonpath="{.items[*].spec.containers[*].image}"
As an alternative to jsonpath, Kubectl supports using go-templates for formatting the output:
kubectl get pods --all-namespaces -o go-template --template="{{range .items}}{{range .spec.containers}}{{.image}} {{end}}{{end}}"
An Ingress is an API object that defines rules which allow external access to services in a cluster. An Ingress controller fulfills the rules set in the Ingress.
This page shows you how to set up a simple Ingress which routes requests to Service web or web2 depending on the HTTP URI.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.19. To check the version, enterkubectl version.
If you are using an older Kubernetes version, switch to the documentation
for that version.
minikube start to create a cluster.To enable the NGINX Ingress controller, run the following command:
minikube addons enable ingress
Verify that the NGINX Ingress controller is running
kubectl get pods -n ingress-nginx
The output is similar to:
NAME READY STATUS RESTARTS AGE
ingress-nginx-admission-create-g9g49 0/1 Completed 0 11m
ingress-nginx-admission-patch-rqp78 0/1 Completed 1 11m
ingress-nginx-controller-59b45fb494-26npt 1/1 Running 0 11m
kubectl get pods -n kube-system
The output is similar to:
NAME READY STATUS RESTARTS AGE
default-http-backend-59868b7dd6-xb8tq 1/1 Running 0 1m
kube-addon-manager-minikube 1/1 Running 0 3m
kube-dns-6dcb57bcc8-n4xd4 3/3 Running 0 2m
kubernetes-dashboard-5498ccf677-b8p5h 1/1 Running 0 2m
nginx-ingress-controller-5984b97644-rnkrg 1/1 Running 0 1m
storage-provisioner 1/1 Running 0 2m
Make sure that you see a Pod with a name that starts with nginx-ingress-controller-.
Create a Deployment using the following command:
kubectl create deployment web --image=gcr.io/google-samples/hello-app:1.0
The output should be:
deployment.apps/web created
Expose the Deployment:
kubectl expose deployment web --type=NodePort --port=8080
The output should be:
service/web exposed
Verify the Service is created and is available on a node port:
kubectl get service web
The output is similar to:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
web NodePort 10.104.133.249 <none> 8080:31637/TCP 12m
Visit the Service via NodePort:
minikube service web --url
The output is similar to:
http://172.17.0.15:31637
31637, and then click Display Port.
The output is similar to:
Hello, world!
Version: 1.0.0
Hostname: web-55b8c6998d-8k564
You can now access the sample app via the Minikube IP address and NodePort. The next step lets you access the app using the Ingress resource.
The following manifest defines an Ingress that sends traffic to your Service via hello-world.info.
Create example-ingress.yaml from the following file:
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
name: example-ingress
annotations:
nginx.ingress.kubernetes.io/rewrite-target: /$1
spec:
rules:
- host: hello-world.info
http:
paths:
- path: /
pathType: Prefix
backend:
service:
name: web
port:
number: 8080Create the Ingress object by running the following command:
kubectl apply -f https://k8s.io/examples/service/networking/example-ingress.yaml
The output should be:
ingress.networking.k8s.io/example-ingress created
Verify the IP address is set:
kubectl get ingress
You should see an IPv4 address in the ADDRESS column; for example:
NAME CLASS HOSTS ADDRESS PORTS AGE
example-ingress <none> hello-world.info 172.17.0.15 80 38s
Add the following line to the bottom of the /etc/hosts file on
your computer (you will need adminstrator access):
172.17.0.15 hello-world.info
minikube ip to get the external IP. The IP address displayed within the ingress list will be the internal IP.
After you make this change, your web browser sends requests for hello-world.info URLs to Minikube.
Verify that the Ingress controller is directing traffic:
curl hello-world.info
You should see:
Hello, world!
Version: 1.0.0
Hostname: web-55b8c6998d-8k564
Create another Deployment using the following command:
kubectl create deployment web2 --image=gcr.io/google-samples/hello-app:2.0
The output should be:
deployment.apps/web2 created
Expose the second Deployment:
kubectl expose deployment web2 --port=8080 --type=NodePort
The output should be:
service/web2 exposed
Edit the existing example-ingress.yaml manifest, and add the
following lines at the end:
- path: /v2
pathType: Prefix
backend:
service:
name: web2
port:
number: 8080
Apply the changes:
kubectl apply -f example-ingress.yaml
You should see:
ingress.networking/example-ingress configured
Access the 1st version of the Hello World app.
curl hello-world.info
The output is similar to:
Hello, world!
Version: 1.0.0
Hostname: web-55b8c6998d-8k564
Access the 2nd version of the Hello World app.
curl hello-world.info/v2
The output is similar to:
Hello, world!
Version: 2.0.0
Hostname: web2-75cd47646f-t8cjk
This page shows how to use a Volume to communicate between two Containers running in the same Pod. See also how to allow processes to communicate by sharing process namespace between containers.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise, you create a Pod that runs two Containers. The two containers share a Volume that they can use to communicate. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: two-containers
spec:
restartPolicy: Never
volumes:
- name: shared-data
emptyDir: {}
containers:
- name: nginx-container
image: nginx
volumeMounts:
- name: shared-data
mountPath: /usr/share/nginx/html
- name: debian-container
image: debian
volumeMounts:
- name: shared-data
mountPath: /pod-data
command: ["/bin/sh"]
args: ["-c", "echo Hello from the debian container > /pod-data/index.html"]
In the configuration file, you can see that the Pod has a Volume named
shared-data.
The first container listed in the configuration file runs an nginx server. The
mount path for the shared Volume is /usr/share/nginx/html.
The second container is based on the debian image, and has a mount path of
/pod-data. The second container runs the following command and then terminates.
echo Hello from the debian container > /pod-data/index.html
Notice that the second container writes the index.html file in the root
directory of the nginx server.
Create the Pod and the two Containers:
kubectl apply -f https://k8s.io/examples/pods/two-container-pod.yaml
View information about the Pod and the Containers:
kubectl get pod two-containers --output=yaml
Here is a portion of the output:
apiVersion: v1
kind: Pod
metadata:
...
name: two-containers
namespace: default
...
spec:
...
containerStatuses:
- containerID: docker://c1d8abd1 ...
image: debian
...
lastState:
terminated:
...
name: debian-container
...
- containerID: docker://96c1ff2c5bb ...
image: nginx
...
name: nginx-container
...
state:
running:
...
You can see that the debian Container has terminated, and the nginx Container is still running.
Get a shell to nginx Container:
kubectl exec -it two-containers -c nginx-container -- /bin/bash
In your shell, verify that nginx is running:
root@two-containers:/# apt-get update
root@two-containers:/# apt-get install curl procps
root@two-containers:/# ps aux
The output is similar to this:
USER PID ... STAT START TIME COMMAND
root 1 ... Ss 21:12 0:00 nginx: master process nginx -g daemon off;
Recall that the debian Container created the index.html file in the nginx root
directory. Use curl to send a GET request to the nginx server:
root@two-containers:/# curl localhost
The output shows that nginx serves a web page written by the debian container:
Hello from the debian container
The primary reason that Pods can have multiple containers is to support helper applications that assist a primary application. Typical examples of helper applications are data pullers, data pushers, and proxies. Helper and primary applications often need to communicate with each other. Typically this is done through a shared filesystem, as shown in this exercise, or through the loopback network interface, localhost. An example of this pattern is a web server along with a helper program that polls a Git repository for new updates.
The Volume in this exercise provides a way for Containers to communicate during the life of the Pod. If the Pod is deleted and recreated, any data stored in the shared Volume is lost.
Learn more about patterns for composite containers.
Learn about composite containers for modular architecture.
See Configure a Pod to share process namespace between containers in a Pod
See Volume.
See Pod.
Kubernetes offers a DNS cluster addon, which most of the supported environments enable by default. In Kubernetes version 1.11 and later, CoreDNS is recommended and is installed by default with kubeadm.
For more information on how to configure CoreDNS for a Kubernetes cluster, see the Customizing DNS Service. An example demonstrating how to use Kubernetes DNS with kube-dns, see the Kubernetes DNS sample plugin.
Once your application is running, you'll inevitably need to debug problems with it.
Earlier we described how you can use kubectl get pods to retrieve simple status information about
your pods. But there are a number of ways to get even more information about your application.
kubectl describe pod to fetch details about podsFor this example we'll use a Deployment to create two pods, similar to the earlier example.
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
selector:
matchLabels:
app: nginx
replicas: 2
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx
resources:
limits:
memory: "128Mi"
cpu: "500m"
ports:
- containerPort: 80
Create deployment by running following command:
kubectl apply -f https://k8s.io/examples/application/nginx-with-request.yaml
deployment.apps/nginx-deployment created
Check pod status by following command:
kubectl get pods
NAME READY STATUS RESTARTS AGE
nginx-deployment-1006230814-6winp 1/1 Running 0 11s
nginx-deployment-1006230814-fmgu3 1/1 Running 0 11s
We can retrieve a lot more information about each of these pods using kubectl describe pod. For example:
kubectl describe pod nginx-deployment-1006230814-6winp
Name: nginx-deployment-1006230814-6winp
Namespace: default
Node: kubernetes-node-wul5/10.240.0.9
Start Time: Thu, 24 Mar 2016 01:39:49 +0000
Labels: app=nginx,pod-template-hash=1006230814
Annotations: kubernetes.io/created-by={"kind":"SerializedReference","apiVersion":"v1","reference":{"kind":"ReplicaSet","namespace":"default","name":"nginx-deployment-1956810328","uid":"14e607e7-8ba1-11e7-b5cb-fa16" ...
Status: Running
IP: 10.244.0.6
Controllers: ReplicaSet/nginx-deployment-1006230814
Containers:
nginx:
Container ID: docker://90315cc9f513c724e9957a4788d3e625a078de84750f244a40f97ae355eb1149
Image: nginx
Image ID: docker://6f62f48c4e55d700cf3eb1b5e33fa051802986b77b874cc351cce539e5163707
Port: 80/TCP
QoS Tier:
cpu: Guaranteed
memory: Guaranteed
Limits:
cpu: 500m
memory: 128Mi
Requests:
memory: 128Mi
cpu: 500m
State: Running
Started: Thu, 24 Mar 2016 01:39:51 +0000
Ready: True
Restart Count: 0
Environment: <none>
Mounts:
/var/run/secrets/kubernetes.io/serviceaccount from default-token-5kdvl (ro)
Conditions:
Type Status
Initialized True
Ready True
PodScheduled True
Volumes:
default-token-4bcbi:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-4bcbi
Optional: false
QoS Class: Guaranteed
Node-Selectors: <none>
Tolerations: <none>
Events:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
54s 54s 1 {default-scheduler } Normal Scheduled Successfully assigned nginx-deployment-1006230814-6winp to kubernetes-node-wul5
54s 54s 1 {kubelet kubernetes-node-wul5} spec.containers{nginx} Normal Pulling pulling image "nginx"
53s 53s 1 {kubelet kubernetes-node-wul5} spec.containers{nginx} Normal Pulled Successfully pulled image "nginx"
53s 53s 1 {kubelet kubernetes-node-wul5} spec.containers{nginx} Normal Created Created container with docker id 90315cc9f513
53s 53s 1 {kubelet kubernetes-node-wul5} spec.containers{nginx} Normal Started Started container with docker id 90315cc9f513
Here you can see configuration information about the container(s) and Pod (labels, resource requirements, etc.), as well as status information about the container(s) and Pod (state, readiness, restart count, events, etc.).
The container state is one of Waiting, Running, or Terminated. Depending on the state, additional information will be provided -- here you can see that for a container in Running state, the system tells you when the container started.
Ready tells you whether the container passed its last readiness probe. (In this case, the container does not have a readiness probe configured; the container is assumed to be ready if no readiness probe is configured.)
Restart Count tells you how many times the container has been restarted; this information can be useful for detecting crash loops in containers that are configured with a restart policy of 'always.'
Currently the only Condition associated with a Pod is the binary Ready condition, which indicates that the pod is able to service requests and should be added to the load balancing pools of all matching services.
Lastly, you see a log of recent events related to your Pod. The system compresses multiple identical events by indicating the first and last time it was seen and the number of times it was seen. "From" indicates the component that is logging the event, "SubobjectPath" tells you which object (e.g. container within the pod) is being referred to, and "Reason" and "Message" tell you what happened.
A common scenario that you can detect using events is when you've created a Pod that won't fit on any node. For example, the Pod might request more resources than are free on any node, or it might specify a label selector that doesn't match any nodes. Let's say we created the previous Deployment with 5 replicas (instead of 2) and requesting 600 millicores instead of 500, on a four-node cluster where each (virtual) machine has 1 CPU. In that case one of the Pods will not be able to schedule. (Note that because of the cluster addon pods such as fluentd, skydns, etc., that run on each node, if we requested 1000 millicores then none of the Pods would be able to schedule.)
kubectl get pods
NAME READY STATUS RESTARTS AGE
nginx-deployment-1006230814-6winp 1/1 Running 0 7m
nginx-deployment-1006230814-fmgu3 1/1 Running 0 7m
nginx-deployment-1370807587-6ekbw 1/1 Running 0 1m
nginx-deployment-1370807587-fg172 0/1 Pending 0 1m
nginx-deployment-1370807587-fz9sd 0/1 Pending 0 1m
To find out why the nginx-deployment-1370807587-fz9sd pod is not running, we can use kubectl describe pod on the pending Pod and look at its events:
kubectl describe pod nginx-deployment-1370807587-fz9sd
Name: nginx-deployment-1370807587-fz9sd
Namespace: default
Node: /
Labels: app=nginx,pod-template-hash=1370807587
Status: Pending
IP:
Controllers: ReplicaSet/nginx-deployment-1370807587
Containers:
nginx:
Image: nginx
Port: 80/TCP
QoS Tier:
memory: Guaranteed
cpu: Guaranteed
Limits:
cpu: 1
memory: 128Mi
Requests:
cpu: 1
memory: 128Mi
Environment Variables:
Volumes:
default-token-4bcbi:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-4bcbi
Events:
FirstSeen LastSeen Count From SubobjectPath Type Reason Message
--------- -------- ----- ---- ------------- -------- ------ -------
1m 48s 7 {default-scheduler } Warning FailedScheduling pod (nginx-deployment-1370807587-fz9sd) failed to fit in any node
fit failure on node (kubernetes-node-6ta5): Node didn't have enough resource: CPU, requested: 1000, used: 1420, capacity: 2000
fit failure on node (kubernetes-node-wul5): Node didn't have enough resource: CPU, requested: 1000, used: 1100, capacity: 2000
Here you can see the event generated by the scheduler saying that the Pod failed to schedule for reason FailedScheduling (and possibly others). The message tells us that there were not enough resources for the Pod on any of the nodes.
To correct this situation, you can use kubectl scale to update your Deployment to specify four or fewer replicas. (Or you could leave the one Pod pending, which is harmless.)
Events such as the ones you saw at the end of kubectl describe pod are persisted in etcd and provide high-level information on what is happening in the cluster. To list all events you can use
kubectl get events
but you have to remember that events are namespaced. This means that if you're interested in events for some namespaced object (e.g. what happened with Pods in namespace my-namespace) you need to explicitly provide a namespace to the command:
kubectl get events --namespace=my-namespace
To see events from all namespaces, you can use the --all-namespaces argument.
In addition to kubectl describe pod, another way to get extra information about a pod (beyond what is provided by kubectl get pod) is to pass the -o yaml output format flag to kubectl get pod. This will give you, in YAML format, even more information than kubectl describe pod--essentially all of the information the system has about the Pod. Here you will see things like annotations (which are key-value metadata without the label restrictions, that is used internally by Kubernetes system components), restart policy, ports, and volumes.
kubectl get pod nginx-deployment-1006230814-6winp -o yaml
apiVersion: v1
kind: Pod
metadata:
annotations:
kubernetes.io/created-by: |
{"kind":"SerializedReference","apiVersion":"v1","reference":{"kind":"ReplicaSet","namespace":"default","name":"nginx-deployment-1006230814","uid":"4c84c175-f161-11e5-9a78-42010af00005","apiVersion":"extensions","resourceVersion":"133434"}}
creationTimestamp: 2016-03-24T01:39:50Z
generateName: nginx-deployment-1006230814-
labels:
app: nginx
pod-template-hash: "1006230814"
name: nginx-deployment-1006230814-6winp
namespace: default
resourceVersion: "133447"
uid: 4c879808-f161-11e5-9a78-42010af00005
spec:
containers:
- image: nginx
imagePullPolicy: Always
name: nginx
ports:
- containerPort: 80
protocol: TCP
resources:
limits:
cpu: 500m
memory: 128Mi
requests:
cpu: 500m
memory: 128Mi
terminationMessagePath: /dev/termination-log
volumeMounts:
- mountPath: /var/run/secrets/kubernetes.io/serviceaccount
name: default-token-4bcbi
readOnly: true
dnsPolicy: ClusterFirst
nodeName: kubernetes-node-wul5
restartPolicy: Always
securityContext: {}
serviceAccount: default
serviceAccountName: default
terminationGracePeriodSeconds: 30
volumes:
- name: default-token-4bcbi
secret:
secretName: default-token-4bcbi
status:
conditions:
- lastProbeTime: null
lastTransitionTime: 2016-03-24T01:39:51Z
status: "True"
type: Ready
containerStatuses:
- containerID: docker://90315cc9f513c724e9957a4788d3e625a078de84750f244a40f97ae355eb1149
image: nginx
imageID: docker://6f62f48c4e55d700cf3eb1b5e33fa051802986b77b874cc351cce539e5163707
lastState: {}
name: nginx
ready: true
restartCount: 0
state:
running:
startedAt: 2016-03-24T01:39:51Z
hostIP: 10.240.0.9
phase: Running
podIP: 10.244.0.6
startTime: 2016-03-24T01:39:49Z
Sometimes when debugging it can be useful to look at the status of a node -- for example, because you've noticed strange behavior of a Pod that's running on the node, or to find out why a Pod won't schedule onto the node. As with Pods, you can use kubectl describe node and kubectl get node -o yaml to retrieve detailed information about nodes. For example, here's what you'll see if a node is down (disconnected from the network, or kubelet dies and won't restart, etc.). Notice the events that show the node is NotReady, and also notice that the pods are no longer running (they are evicted after five minutes of NotReady status).
kubectl get nodes
NAME STATUS ROLES AGE VERSION
kubernetes-node-861h NotReady <none> 1h v1.13.0
kubernetes-node-bols Ready <none> 1h v1.13.0
kubernetes-node-st6x Ready <none> 1h v1.13.0
kubernetes-node-unaj Ready <none> 1h v1.13.0
kubectl describe node kubernetes-node-861h
Name: kubernetes-node-861h
Role
Labels: kubernetes.io/arch=amd64
kubernetes.io/os=linux
kubernetes.io/hostname=kubernetes-node-861h
Annotations: node.alpha.kubernetes.io/ttl=0
volumes.kubernetes.io/controller-managed-attach-detach=true
Taints: <none>
CreationTimestamp: Mon, 04 Sep 2017 17:13:23 +0800
Phase:
Conditions:
Type Status LastHeartbeatTime LastTransitionTime Reason Message
---- ------ ----------------- ------------------ ------ -------
OutOfDisk Unknown Fri, 08 Sep 2017 16:04:28 +0800 Fri, 08 Sep 2017 16:20:58 +0800 NodeStatusUnknown Kubelet stopped posting node status.
MemoryPressure Unknown Fri, 08 Sep 2017 16:04:28 +0800 Fri, 08 Sep 2017 16:20:58 +0800 NodeStatusUnknown Kubelet stopped posting node status.
DiskPressure Unknown Fri, 08 Sep 2017 16:04:28 +0800 Fri, 08 Sep 2017 16:20:58 +0800 NodeStatusUnknown Kubelet stopped posting node status.
Ready Unknown Fri, 08 Sep 2017 16:04:28 +0800 Fri, 08 Sep 2017 16:20:58 +0800 NodeStatusUnknown Kubelet stopped posting node status.
Addresses: 10.240.115.55,104.197.0.26
Capacity:
cpu: 2
hugePages: 0
memory: 4046788Ki
pods: 110
Allocatable:
cpu: 1500m
hugePages: 0
memory: 1479263Ki
pods: 110
System Info:
Machine ID: 8e025a21a4254e11b028584d9d8b12c4
System UUID: 349075D1-D169-4F25-9F2A-E886850C47E3
Boot ID: 5cd18b37-c5bd-4658-94e0-e436d3f110e0
Kernel Version: 4.4.0-31-generic
OS Image: Debian GNU/Linux 8 (jessie)
Operating System: linux
Architecture: amd64
Container Runtime Version: docker://1.12.5
Kubelet Version: v1.6.9+a3d1dfa6f4335
Kube-Proxy Version: v1.6.9+a3d1dfa6f4335
ExternalID: 15233045891481496305
Non-terminated Pods: (9 in total)
Namespace Name CPU Requests CPU Limits Memory Requests Memory Limits
--------- ---- ------------ ---------- --------------- -------------
......
Allocated resources:
(Total limits may be over 100 percent, i.e., overcommitted.)
CPU Requests CPU Limits Memory Requests Memory Limits
------------ ---------- --------------- -------------
900m (60%) 2200m (146%) 1009286400 (66%) 5681286400 (375%)
Events: <none>
kubectl get node kubernetes-node-861h -o yaml
apiVersion: v1
kind: Node
metadata:
creationTimestamp: 2015-07-10T21:32:29Z
labels:
kubernetes.io/hostname: kubernetes-node-861h
name: kubernetes-node-861h
resourceVersion: "757"
uid: 2a69374e-274b-11e5-a234-42010af0d969
spec:
externalID: "15233045891481496305"
podCIDR: 10.244.0.0/24
providerID: gce://striped-torus-760/us-central1-b/kubernetes-node-861h
status:
addresses:
- address: 10.240.115.55
type: InternalIP
- address: 104.197.0.26
type: ExternalIP
capacity:
cpu: "1"
memory: 3800808Ki
pods: "100"
conditions:
- lastHeartbeatTime: 2015-07-10T21:34:32Z
lastTransitionTime: 2015-07-10T21:35:15Z
reason: Kubelet stopped posting node status.
status: Unknown
type: Ready
nodeInfo:
bootID: 4e316776-b40d-4f78-a4ea-ab0d73390897
containerRuntimeVersion: docker://Unknown
kernelVersion: 3.16.0-0.bpo.4-amd64
kubeProxyVersion: v0.21.1-185-gffc5a86098dc01
kubeletVersion: v0.21.1-185-gffc5a86098dc01
machineID: ""
osImage: Debian GNU/Linux 7 (wheezy)
systemUUID: ABE5F6B4-D44B-108B-C46A-24CCE16C8B6E
Learn about additional debugging tools, including:
Kubernetes auditing provides a security-relevant, chronological set of records documenting the sequence of actions in a cluster. The cluster audits the activities generated by users, by applications that use the Kubernetes API, and by the control plane itself.
Auditing allows cluster administrators to answer the following questions:
Audit records begin their lifecycle inside the kube-apiserver component. Each request on each stage of its execution generates an audit event, which is then pre-processed according to a certain policy and written to a backend. The policy determines what's recorded and the backends persist the records. The current backend implementations include logs files and webhooks.
Each request can be recorded with an associated stage. The defined stages are:
RequestReceived - The stage for events generated as soon as the audit
handler receives the request, and before it is delegated down the handler
chain.ResponseStarted - Once the response headers are sent, but before the
response body is sent. This stage is only generated for long-running requests
(e.g. watch).ResponseComplete - The response body has been completed and no more bytes
will be sent.Panic - Events generated when a panic occurred.The audit logging feature increases the memory consumption of the API server because some context required for auditing is stored for each request. Memory consumption depends on the audit logging configuration.
Audit policy defines rules about what events should be recorded and what data
they should include. The audit policy object structure is defined in the
audit.k8s.io API group.
When an event is processed, it's
compared against the list of rules in order. The first matching rule sets the
audit level of the event. The defined audit levels are:
None - don't log events that match this rule.Metadata - log request metadata (requesting user, timestamp, resource,
verb, etc.) but not request or response body.Request - log event metadata and request body but not response body.
This does not apply for non-resource requests.RequestResponse - log event metadata, request and response bodies.
This does not apply for non-resource requests.You can pass a file with the policy to kube-apiserver
using the --audit-policy-file flag. If the flag is omitted, no events are logged.
Note that the rules field must be provided in the audit policy file.
A policy with no (0) rules is treated as illegal.
Below is an example audit policy file:
apiVersion: audit.k8s.io/v1 # This is required.
kind: Policy
# Don't generate audit events for all requests in RequestReceived stage.
omitStages:
- "RequestReceived"
rules:
# Log pod changes at RequestResponse level
- level: RequestResponse
resources:
- group: ""
# Resource "pods" doesn't match requests to any subresource of pods,
# which is consistent with the RBAC policy.
resources: ["pods"]
# Log "pods/log", "pods/status" at Metadata level
- level: Metadata
resources:
- group: ""
resources: ["pods/log", "pods/status"]
# Don't log requests to a configmap called "controller-leader"
- level: None
resources:
- group: ""
resources: ["configmaps"]
resourceNames: ["controller-leader"]
# Don't log watch requests by the "system:kube-proxy" on endpoints or services
- level: None
users: ["system:kube-proxy"]
verbs: ["watch"]
resources:
- group: "" # core API group
resources: ["endpoints", "services"]
# Don't log authenticated requests to certain non-resource URL paths.
- level: None
userGroups: ["system:authenticated"]
nonResourceURLs:
- "/api*" # Wildcard matching.
- "/version"
# Log the request body of configmap changes in kube-system.
- level: Request
resources:
- group: "" # core API group
resources: ["configmaps"]
# This rule only applies to resources in the "kube-system" namespace.
# The empty string "" can be used to select non-namespaced resources.
namespaces: ["kube-system"]
# Log configmap and secret changes in all other namespaces at the Metadata level.
- level: Metadata
resources:
- group: "" # core API group
resources: ["secrets", "configmaps"]
# Log all other resources in core and extensions at the Request level.
- level: Request
resources:
- group: "" # core API group
- group: "extensions" # Version of group should NOT be included.
# A catch-all rule to log all other requests at the Metadata level.
- level: Metadata
# Long-running requests like watches that fall under this rule will not
# generate an audit event in RequestReceived.
omitStages:
- "RequestReceived"
You can use a minimal audit policy file to log all requests at the Metadata level:
# Log all requests at the Metadata level.
apiVersion: audit.k8s.io/v1
kind: Policy
rules:
- level: Metadata
If you're crafting your own audit profile, you can use the audit profile for Google Container-Optimized OS as a starting point. You can check the configure-helper.sh script, which generates an audit policy file. You can see most of the audit policy file by looking directly at the script.
You can also refer to the Policy configuration reference
for details about the fields defined.
Audit backends persist audit events to an external storage. Out of the box, the kube-apiserver provides two backends:
In all cases, audit events follow a structure defined by the Kubernetes API in the
audit.k8s.io API group.
In case of patches, request body is a JSON array with patch operations, not a JSON object
with an appropriate Kubernetes API object. For example, the following request body is a valid patch
request to /apis/batch/v1/namespaces/some-namespace/jobs/some-job-name:
[
{
"op": "replace",
"path": "/spec/parallelism",
"value": 0
},
{
"op": "remove",
"path": "/spec/template/spec/containers/0/terminationMessagePolicy"
}
]
The log backend writes audit events to a file in JSONlines format.
You can configure the log audit backend using the following kube-apiserver flags:
--audit-log-path specifies the log file path that log backend uses to write
audit events. Not specifying this flag disables log backend. - means standard out--audit-log-maxage defined the maximum number of days to retain old audit log files--audit-log-maxbackup defines the maximum number of audit log files to retain--audit-log-maxsize defines the maximum size in megabytes of the audit log file before it gets rotatedIf your cluster's control plane runs the kube-apiserver as a Pod, remember to mount the hostPath
to the location of the policy file and log file, so that audit records are persisted. For example:
--audit-policy-file=/etc/kubernetes/audit-policy.yaml \
--audit-log-path=/var/log/audit.log
then mount the volumes:
...
volumeMounts:
- mountPath: /etc/kubernetes/audit-policy.yaml
name: audit
readOnly: true
- mountPath: /var/log/audit.log
name: audit-log
readOnly: false
and finally configure the hostPath:
...
- name: audit
hostPath:
path: /etc/kubernetes/audit-policy.yaml
type: File
- name: audit-log
hostPath:
path: /var/log/audit.log
type: FileOrCreate
The webhook audit backend sends audit events to a remote web API, which is assumed to be a form of the Kubernetes API, including means of authentication. You can configure a webhook audit backend using the following kube-apiserver flags:
--audit-webhook-config-file specifies the path to a file with a webhook
configuration. The webhook configuration is effectively a specialized
kubeconfig.--audit-webhook-initial-backoff specifies the amount of time to wait after the first failed
request before retrying. Subsequent requests are retried with exponential backoff.The webhook config file uses the kubeconfig format to specify the remote address of the service and credentials used to connect to it.
Both log and webhook backends support batching. Using webhook as an example, here's the list of
available flags. To get the same flag for log backend, replace webhook with log in the flag
name. By default, batching is enabled in webhook and disabled in log. Similarly, by default
throttling is enabled in webhook and disabled in log.
--audit-webhook-mode defines the buffering strategy. One of the following:
batch - buffer events and asynchronously process them in batches. This is the default.blocking - block API server responses on processing each individual event.blocking-strict - Same as blocking, but when there is a failure during audit logging at the
RequestReceived stage, the whole request to the kube-apiserver fails.The following flags are used only in the batch mode:
--audit-webhook-batch-buffer-size defines the number of events to buffer before batching.
If the rate of incoming events overflows the buffer, events are dropped.--audit-webhook-batch-max-size defines the maximum number of events in one batch.--audit-webhook-batch-max-wait defines the maximum amount of time to wait before unconditionally
batching events in the queue.--audit-webhook-batch-throttle-qps defines the maximum average number of batches generated
per second.--audit-webhook-batch-throttle-burst defines the maximum number of batches generated at the same
moment if the allowed QPS was underutilized previously.Parameters should be set to accommodate the load on the API server.
For example, if kube-apiserver receives 100 requests each second, and each request is audited only
on ResponseStarted and ResponseComplete stages, you should account for ≅200 audit
events being generated each second. Assuming that there are up to 100 events in a batch,
you should set throttling level at least 2 queries per second. Assuming that the backend can take up to
5 seconds to write events, you should set the buffer size to hold up to 5 seconds of events;
that is: 10 batches, or 1000 events.
In most cases however, the default parameters should be sufficient and you don't have to worry about setting them manually. You can look at the following Prometheus metrics exposed by kube-apiserver and in the logs to monitor the state of the auditing subsystem.
apiserver_audit_event_total metric contains the total number of audit events exported.apiserver_audit_error_total metric contains the total number of events dropped due to an error
during exporting.Both log and webhook backends support limiting the size of events that are logged. As an example, the following is the list of flags available for the log backend:
audit-log-truncate-enabled whether event and batch truncating is enabled.audit-log-truncate-max-batch-size maximum size in bytes of the batch sent to the underlying backend.audit-log-truncate-max-event-size maximum size in bytes of the audit event sent to the underlying backend.By default truncate is disabled in both webhook and log, a cluster administrator should set
audit-log-truncate-enabled or audit-webhook-truncate-enabled to enable the feature.
Event
and the Policy
resource types by reading the Audit configuration reference.This task shows you how to debug a StatefulSet.
In order to list all the pods which belong to a StatefulSet, which have a label app=myapp set on them,
you can use the following:
kubectl get pods -l app=myapp
If you find that any Pods listed are in Unknown or Terminating state for an extended period of time,
refer to the Deleting StatefulSet Pods task for
instructions on how to deal with them.
You can debug individual Pods in a StatefulSet using the
Debugging Pods guide.
Learn more about debugging an init-container.
This page shows how to investigate problems related to the execution of
Init Containers. The example command lines below refer to the Pod as
<pod-name> and the Init Containers as <init-container-1> and
<init-container-2>.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Display the status of your pod:
kubectl get pod <pod-name>
For example, a status of Init:1/2 indicates that one of two Init Containers
has completed successfully:
NAME READY STATUS RESTARTS AGE
<pod-name> 0/1 Init:1/2 0 7s
See Understanding Pod status for more examples of status values and their meanings.
View more detailed information about Init Container execution:
kubectl describe pod <pod-name>
For example, a Pod with two Init Containers might show the following:
Init Containers:
<init-container-1>:
Container ID: ...
...
State: Terminated
Reason: Completed
Exit Code: 0
Started: ...
Finished: ...
Ready: True
Restart Count: 0
...
<init-container-2>:
Container ID: ...
...
State: Waiting
Reason: CrashLoopBackOff
Last State: Terminated
Reason: Error
Exit Code: 1
Started: ...
Finished: ...
Ready: False
Restart Count: 3
...
You can also access the Init Container statuses programmatically by reading the
status.initContainerStatuses field on the Pod Spec:
kubectl get pod nginx --template '{{.status.initContainerStatuses}}'
This command will return the same information as above in raw JSON.
Pass the Init Container name along with the Pod name to access its logs.
kubectl logs <pod-name> -c <init-container-2>
Init Containers that run a shell script print
commands as they're executed. For example, you can do this in Bash by running
set -x at the beginning of the script.
A Pod status beginning with Init: summarizes the status of Init Container
execution. The table below describes some example status values that you might
see while debugging Init Containers.
| Status | Meaning |
|---|---|
Init:N/M |
The Pod has M Init Containers, and N have completed so far. |
Init:Error |
An Init Container has failed to execute. |
Init:CrashLoopBackOff |
An Init Container has failed repeatedly. |
Pending |
The Pod has not yet begun executing Init Containers. |
PodInitializing or Running |
The Pod has already finished executing Init Containers. |
This page shows how to debug Pods and ReplicationControllers.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The first step in debugging a pod is taking a look at it. Check the current state of the pod and recent events with the following command:
kubectl describe pods ${POD_NAME}
Look at the state of the containers in the pod. Are they all Running? Have
there been recent restarts?
Continue debugging depending on the state of the pods.
If a pod is stuck in Pending it means that it can not be scheduled onto a
node. Generally this is because there are insufficient resources of one type or
another that prevent scheduling. Look at the output of the kubectl describe ... command above. There should be messages from the scheduler about why it
can not schedule your pod. Reasons include:
You may have exhausted the supply of CPU or Memory in your cluster. In this case you can try several things:
Add more nodes to the cluster.
Terminate unneeded pods to make room for pending pods.
Check that the pod is not larger than your nodes. For example, if all
nodes have a capacity of cpu:1, then a pod with a request of cpu: 1.1
will never be scheduled.
You can check node capacities with the kubectl get nodes -o <format>
command. Here are some example command lines that extract the necessary
information:
kubectl get nodes -o yaml | egrep '\sname:|cpu:|memory:'
kubectl get nodes -o json | jq '.items[] | {name: .metadata.name, cap: .status.capacity}'
The resource quota feature can be configured to limit the total amount of resources that can be consumed. If used in conjunction with namespaces, it can prevent one team from hogging all the resources.
When you bind a pod to a hostPort there are a limited number of places that
the pod can be scheduled. In most cases, hostPort is unnecessary; try using a
service object to expose your pod. If you do require hostPort then you can
only schedule as many pods as there are nodes in your container cluster.
If a pod is stuck in the Waiting state, then it has been scheduled to a
worker node, but it can't run on that machine. Again, the information from
kubectl describe ... should be informative. The most common cause of
Waiting pods is a failure to pull the image. There are three things to check:
docker pull <image> on your machine to see if the image can be
pulled.Once your pod has been scheduled, the methods described in Debug Running Pods are available for debugging.
ReplicationControllers are fairly straightforward. They can either create pods or they can't. If they can't create pods, then please refer to the instructions above to debug your pods.
You can also use kubectl describe rc ${CONTROLLER_NAME} to inspect events
related to the replication controller.
This page explains how to debug Pods running (or crashing) on a Node.
kubectl.First, look at the logs of the affected container:
kubectl logs ${POD_NAME} ${CONTAINER_NAME}
If your container has previously crashed, you can access the previous container's crash log with:
kubectl logs --previous ${POD_NAME} ${CONTAINER_NAME}
If the container image includes
debugging utilities, as is the case with images built from Linux and Windows OS
base images, you can run commands inside a specific container with
kubectl exec:
kubectl exec ${POD_NAME} -c ${CONTAINER_NAME} -- ${CMD} ${ARG1} ${ARG2} ... ${ARGN}
-c ${CONTAINER_NAME} is optional. You can omit it for Pods that only contain a single container.
As an example, to look at the logs from a running Cassandra pod, you might run
kubectl exec cassandra -- cat /var/log/cassandra/system.log
You can run a shell that's connected to your terminal using the -i and -t
arguments to kubectl exec, for example:
kubectl exec -it cassandra -- sh
For more details, see Get a Shell to a Running Container.
Kubernetes v1.22 [alpha]
Ephemeral containers
are useful for interactive troubleshooting when kubectl exec is insufficient
because a container has crashed or a container image doesn't include debugging
utilities, such as with distroless images.
EphemeralContainers feature gate enabled in your
cluster and kubectl version v1.22 or later.
You can use the kubectl debug command to add ephemeral containers to a
running Pod. First, create a pod for the example:
kubectl run ephemeral-demo --image=k8s.gcr.io/pause:3.1 --restart=Never
The examples in this section use the pause container image because it does not
contain debugging utilities, but this method works with all container
images.
If you attempt to use kubectl exec to create a shell you will see an error
because there is no shell in this container image.
kubectl exec -it ephemeral-demo -- sh
OCI runtime exec failed: exec failed: container_linux.go:346: starting container process caused "exec: \"sh\": executable file not found in $PATH": unknown
You can instead add a debugging container using kubectl debug. If you
specify the -i/--interactive argument, kubectl will automatically attach
to the console of the Ephemeral Container.
kubectl debug -it ephemeral-demo --image=busybox --target=ephemeral-demo
Defaulting debug container name to debugger-8xzrl.
If you don't see a command prompt, try pressing enter.
/ #
This command adds a new busybox container and attaches to it. The --target
parameter targets the process namespace of another container. It's necessary
here because kubectl run does not enable process namespace sharing in the pod it
creates.
--target parameter must be supported by the Container Runtime. When not supported,
the Ephemeral Container may not be started, or it may be started with an
isolated process namespace so that ps does not reveal processes in other
containers.
You can view the state of the newly created ephemeral container using kubectl describe:
kubectl describe pod ephemeral-demo
...
Ephemeral Containers:
debugger-8xzrl:
Container ID: docker://b888f9adfd15bd5739fefaa39e1df4dd3c617b9902082b1cfdc29c4028ffb2eb
Image: busybox
Image ID: docker-pullable://busybox@sha256:1828edd60c5efd34b2bf5dd3282ec0cc04d47b2ff9caa0b6d4f07a21d1c08084
Port: <none>
Host Port: <none>
State: Running
Started: Wed, 12 Feb 2020 14:25:42 +0100
Ready: False
Restart Count: 0
Environment: <none>
Mounts: <none>
...
Use kubectl delete to remove the Pod when you're finished:
kubectl delete pod ephemeral-demo
Sometimes Pod configuration options make it difficult to troubleshoot in certain
situations. For example, you can't run kubectl exec to troubleshoot your
container if your container image does not include a shell or if your application
crashes on startup. In these situations you can use kubectl debug to create a
copy of the Pod with configuration values changed to aid debugging.
Adding a new container can be useful when your application is running but not behaving as you expect and you'd like to add additional troubleshooting utilities to the Pod.
For example, maybe your application's container images are built on busybox
but you need debugging utilities not included in busybox. You can simulate
this scenario using kubectl run:
kubectl run myapp --image=busybox --restart=Never -- sleep 1d
Run this command to create a copy of myapp named myapp-debug that adds a
new Ubuntu container for debugging:
kubectl debug myapp -it --image=ubuntu --share-processes --copy-to=myapp-debug
Defaulting debug container name to debugger-w7xmf.
If you don't see a command prompt, try pressing enter.
root@myapp-debug:/#
kubectl debug automatically generates a container name if you don't choose
one using the --container flag.-i flag causes kubectl debug to attach to the new container by
default. You can prevent this by specifying --attach=false. If your session
becomes disconnected you can reattach using kubectl attach.--share-processes allows the containers in this Pod to see processes
from the other containers in the Pod. For more information about how this
works, see Share Process Namespace between Containers in a Pod.Don't forget to clean up the debugging Pod when you're finished with it:
kubectl delete pod myapp myapp-debug
Sometimes it's useful to change the command for a container, for example to add a debugging flag or because the application is crashing.
To simulate a crashing application, use kubectl run to create a container
that immediately exits:
kubectl run --image=busybox myapp -- false
You can see using kubectl describe pod myapp that this container is crashing:
Containers:
myapp:
Image: busybox
...
Args:
false
State: Waiting
Reason: CrashLoopBackOff
Last State: Terminated
Reason: Error
Exit Code: 1
You can use kubectl debug to create a copy of this Pod with the command
changed to an interactive shell:
kubectl debug myapp -it --copy-to=myapp-debug --container=myapp -- sh
If you don't see a command prompt, try pressing enter.
/ #
Now you have an interactive shell that you can use to perform tasks like checking filesystem paths or running the container command manually.
--container or kubectl debug will instead
create a new container to run the command you specified.-i flag causes kubectl debug to attach to the container by default.
You can prevent this by specifying --attach=false. If your session becomes
disconnected you can reattach using kubectl attach.Don't forget to clean up the debugging Pod when you're finished with it:
kubectl delete pod myapp myapp-debug
In some situations you may want to change a misbehaving Pod from its normal production container images to an image containing a debugging build or additional utilities.
As an example, create a Pod using kubectl run:
kubectl run myapp --image=busybox --restart=Never -- sleep 1d
Now use kubectl debug to make a copy and change its container image
to ubuntu:
kubectl debug myapp --copy-to=myapp-debug --set-image=*=ubuntu
The syntax of --set-image uses the same container_name=image syntax as
kubectl set image. *=ubuntu means change the image of all containers
to ubuntu.
Don't forget to clean up the debugging Pod when you're finished with it:
kubectl delete pod myapp myapp-debug
If none of these approaches work, you can find the Node on which the Pod is
running and create a privileged Pod running in the host namespaces. To create
an interactive shell on a node using kubectl debug, run:
kubectl debug node/mynode -it --image=ubuntu
Creating debugging pod node-debugger-mynode-pdx84 with container debugger on node mynode.
If you don't see a command prompt, try pressing enter.
root@ek8s:/#
When creating a debugging session on a node, keep in mind that:
kubectl debug automatically generates the name of the new Pod based on
the name of the Node./host.Don't forget to clean up the debugging Pod when you're finished with it:
kubectl delete pod node-debugger-mynode-pdx84
An issue that comes up rather frequently for new installations of Kubernetes is that a Service is not working properly. You've run your Pods through a Deployment (or other workload controller) and created a Service, but you get no response when you try to access it. This document will hopefully help you to figure out what's going wrong.
For many steps here you will want to see what a Pod running in the cluster sees. The simplest way to do this is to run an interactive busybox Pod:
kubectl run -it --rm --restart=Never busybox --image=gcr.io/google-containers/busybox sh
If you already have a running Pod that you prefer to use, you can run a command in it using:
kubectl exec <POD-NAME> -c <CONTAINER-NAME> -- <COMMAND>
For the purposes of this walk-through, let's run some Pods. Since you're probably debugging your own Service you can substitute your own details, or you can follow along and get a second data point.
kubectl create deployment hostnames --image=k8s.gcr.io/serve_hostname
deployment.apps/hostnames created
kubectl commands will print the type and name of the resource created or mutated, which can then be used in subsequent commands.
Let's scale the deployment to 3 replicas.
kubectl scale deployment hostnames --replicas=3
deployment.apps/hostnames scaled
Note that this is the same as if you had started the Deployment with the following YAML:
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: hostnames
name: hostnames
spec:
selector:
matchLabels:
app: hostnames
replicas: 3
template:
metadata:
labels:
app: hostnames
spec:
containers:
- name: hostnames
image: k8s.gcr.io/serve_hostname
The label "app" is automatically set by kubectl create deployment to the name of the
Deployment.
You can confirm your Pods are running:
kubectl get pods -l app=hostnames
NAME READY STATUS RESTARTS AGE
hostnames-632524106-bbpiw 1/1 Running 0 2m
hostnames-632524106-ly40y 1/1 Running 0 2m
hostnames-632524106-tlaok 1/1 Running 0 2m
You can also confirm that your Pods are serving. You can get the list of Pod IP addresses and test them directly.
kubectl get pods -l app=hostnames \
-o go-template='{{range .items}}{{.status.podIP}}{{"\n"}}{{end}}'
10.244.0.5
10.244.0.6
10.244.0.7
The example container used for this walk-through serves its own hostname via HTTP on port 9376, but if you are debugging your own app, you'll want to use whatever port number your Pods are listening on.
From within a pod:
for ep in 10.244.0.5:9376 10.244.0.6:9376 10.244.0.7:9376; do
wget -qO- $ep
done
This should produce something like:
hostnames-632524106-bbpiw
hostnames-632524106-ly40y
hostnames-632524106-tlaok
If you are not getting the responses you expect at this point, your Pods
might not be healthy or might not be listening on the port you think they are.
You might find kubectl logs to be useful for seeing what is happening, or
perhaps you need to kubectl exec directly into your Pods and debug from
there.
Assuming everything has gone to plan so far, you can start to investigate why your Service doesn't work.
The astute reader will have noticed that you did not actually create a Service yet - that is intentional. This is a step that sometimes gets forgotten, and is the first thing to check.
What would happen if you tried to access a non-existent Service? If you have another Pod that consumes this Service by name you would get something like:
wget -O- hostnames
Resolving hostnames (hostnames)... failed: Name or service not known.
wget: unable to resolve host address 'hostnames'
The first thing to check is whether that Service actually exists:
kubectl get svc hostnames
No resources found.
Error from server (NotFound): services "hostnames" not found
Let's create the Service. As before, this is for the walk-through - you can use your own Service's details here.
kubectl expose deployment hostnames --port=80 --target-port=9376
service/hostnames exposed
And read it back:
kubectl get svc hostnames
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
hostnames ClusterIP 10.0.1.175 <none> 80/TCP 5s
Now you know that the Service exists.
As before, this is the same as if you had started the Service with YAML:
apiVersion: v1
kind: Service
metadata:
labels:
app: hostnames
name: hostnames
spec:
selector:
app: hostnames
ports:
- name: default
protocol: TCP
port: 80
targetPort: 9376
In order to highlight the full range of configuration, the Service you created here uses a different port number than the Pods. For many real-world Services, these values might be the same.
One of the most common ways that clients consume a Service is through a DNS name.
From a Pod in the same Namespace:
nslookup hostnames
Address 1: 10.0.0.10 kube-dns.kube-system.svc.cluster.local
Name: hostnames
Address 1: 10.0.1.175 hostnames.default.svc.cluster.local
If this fails, perhaps your Pod and Service are in different Namespaces, try a namespace-qualified name (again, from within a Pod):
nslookup hostnames.default
Address 1: 10.0.0.10 kube-dns.kube-system.svc.cluster.local
Name: hostnames.default
Address 1: 10.0.1.175 hostnames.default.svc.cluster.local
If this works, you'll need to adjust your app to use a cross-namespace name, or run your app and Service in the same Namespace. If this still fails, try a fully-qualified name:
nslookup hostnames.default.svc.cluster.local
Address 1: 10.0.0.10 kube-dns.kube-system.svc.cluster.local
Name: hostnames.default.svc.cluster.local
Address 1: 10.0.1.175 hostnames.default.svc.cluster.local
Note the suffix here: "default.svc.cluster.local". The "default" is the Namespace you're operating in. The "svc" denotes that this is a Service. The "cluster.local" is your cluster domain, which COULD be different in your own cluster.
You can also try this from a Node in the cluster:
nslookup hostnames.default.svc.cluster.local 10.0.0.10
Server: 10.0.0.10
Address: 10.0.0.10#53
Name: hostnames.default.svc.cluster.local
Address: 10.0.1.175
If you are able to do a fully-qualified name lookup but not a relative one, you
need to check that your /etc/resolv.conf file in your Pod is correct. From
within a Pod:
cat /etc/resolv.conf
You should see something like:
nameserver 10.0.0.10
search default.svc.cluster.local svc.cluster.local cluster.local example.com
options ndots:5
The nameserver line must indicate your cluster's DNS Service. This is
passed into kubelet with the --cluster-dns flag.
The search line must include an appropriate suffix for you to find the
Service name. In this case it is looking for Services in the local
Namespace ("default.svc.cluster.local"), Services in all Namespaces
("svc.cluster.local"), and lastly for names in the cluster ("cluster.local").
Depending on your own install you might have additional records after that (up
to 6 total). The cluster suffix is passed into kubelet with the
--cluster-domain flag. Throughout this document, the cluster suffix is
assumed to be "cluster.local". Your own clusters might be configured
differently, in which case you should change that in all of the previous
commands.
The options line must set ndots high enough that your DNS client library
considers search paths at all. Kubernetes sets this to 5 by default, which is
high enough to cover all of the DNS names it generates.
If the above still fails, DNS lookups are not working for your Service. You can take a step back and see what else is not working. The Kubernetes master Service should always work. From within a Pod:
nslookup kubernetes.default
Server: 10.0.0.10
Address 1: 10.0.0.10 kube-dns.kube-system.svc.cluster.local
Name: kubernetes.default
Address 1: 10.0.0.1 kubernetes.default.svc.cluster.local
If this fails, please see the kube-proxy section of this document, or even go back to the top of this document and start over, but instead of debugging your own Service, debug the DNS Service.
Assuming you have confirmed that DNS works, the next thing to test is whether your
Service works by its IP address. From a Pod in your cluster, access the
Service's IP (from kubectl get above).
for i in $(seq 1 3); do
wget -qO- 10.0.1.175:80
done
This should produce something like:
hostnames-632524106-bbpiw
hostnames-632524106-ly40y
hostnames-632524106-tlaok
If your Service is working, you should get correct responses. If not, there are a number of things that could be going wrong. Read on.
It might sound silly, but you should really double and triple check that your Service is correct and matches your Pod's port. Read back your Service and verify it:
kubectl get service hostnames -o json
{
"kind": "Service",
"apiVersion": "v1",
"metadata": {
"name": "hostnames",
"namespace": "default",
"uid": "428c8b6c-24bc-11e5-936d-42010af0a9bc",
"resourceVersion": "347189",
"creationTimestamp": "2015-07-07T15:24:29Z",
"labels": {
"app": "hostnames"
}
},
"spec": {
"ports": [
{
"name": "default",
"protocol": "TCP",
"port": 80,
"targetPort": 9376,
"nodePort": 0
}
],
"selector": {
"app": "hostnames"
},
"clusterIP": "10.0.1.175",
"type": "ClusterIP",
"sessionAffinity": "None"
},
"status": {
"loadBalancer": {}
}
}
spec.ports[]?targetPort correct for your Pods (some Pods use a different port than the Service)?protocol correct for your Pods?If you got this far, you have confirmed that your Service is correctly defined and is resolved by DNS. Now let's check that the Pods you ran are actually being selected by the Service.
Earlier you saw that the Pods were running. You can re-check that:
kubectl get pods -l app=hostnames
NAME READY STATUS RESTARTS AGE
hostnames-632524106-bbpiw 1/1 Running 0 1h
hostnames-632524106-ly40y 1/1 Running 0 1h
hostnames-632524106-tlaok 1/1 Running 0 1h
The -l app=hostnames argument is a label selector configured on the Service.
The "AGE" column says that these Pods are about an hour old, which implies that they are running fine and not crashing.
The "RESTARTS" column says that these pods are not crashing frequently or being restarted. Frequent restarts could lead to intermittent connectivity issues. If the restart count is high, read more about how to debug pods.
Inside the Kubernetes system is a control loop which evaluates the selector of every Service and saves the results into a corresponding Endpoints object.
kubectl get endpoints hostnames
NAME ENDPOINTS
hostnames 10.244.0.5:9376,10.244.0.6:9376,10.244.0.7:9376
This confirms that the endpoints controller has found the correct Pods for
your Service. If the ENDPOINTS column is <none>, you should check that
the spec.selector field of your Service actually selects for
metadata.labels values on your Pods. A common mistake is to have a typo or
other error, such as the Service selecting for app=hostnames, but the
Deployment specifying run=hostnames, as in versions previous to 1.18, where
the kubectl run command could have been also used to create a Deployment.
At this point, you know that your Service exists and has selected your Pods. At the beginning of this walk-through, you verified the Pods themselves. Let's check again that the Pods are actually working - you can bypass the Service mechanism and go straight to the Pods, as listed by the Endpoints above.
From within a Pod:
for ep in 10.244.0.5:9376 10.244.0.6:9376 10.244.0.7:9376; do
wget -qO- $ep
done
This should produce something like:
hostnames-632524106-bbpiw
hostnames-632524106-ly40y
hostnames-632524106-tlaok
You expect each Pod in the Endpoints list to return its own hostname. If this is not what happens (or whatever the correct behavior is for your own Pods), you should investigate what's happening there.
If you get here, your Service is running, has Endpoints, and your Pods are actually serving. At this point, the whole Service proxy mechanism is suspect. Let's confirm it, piece by piece.
The default implementation of Services, and the one used on most clusters, is kube-proxy. This is a program that runs on every node and configures one of a small set of mechanisms for providing the Service abstraction. If your cluster does not use kube-proxy, the following sections will not apply, and you will have to investigate whatever implementation of Services you are using.
Confirm that kube-proxy is running on your Nodes. Running directly on a
Node, you should get something like the below:
ps auxw | grep kube-proxy
root 4194 0.4 0.1 101864 17696 ? Sl Jul04 25:43 /usr/local/bin/kube-proxy --master=https://kubernetes-master --kubeconfig=/var/lib/kube-proxy/kubeconfig --v=2
Next, confirm that it is not failing something obvious, like contacting the
master. To do this, you'll have to look at the logs. Accessing the logs
depends on your Node OS. On some OSes it is a file, such as
/var/log/kube-proxy.log, while other OSes use journalctl to access logs. You
should see something like:
I1027 22:14:53.995134 5063 server.go:200] Running in resource-only container "/kube-proxy"
I1027 22:14:53.998163 5063 server.go:247] Using iptables Proxier.
I1027 22:14:53.999055 5063 server.go:255] Tearing down userspace rules. Errors here are acceptable.
I1027 22:14:54.038140 5063 proxier.go:352] Setting endpoints for "kube-system/kube-dns:dns-tcp" to [10.244.1.3:53]
I1027 22:14:54.038164 5063 proxier.go:352] Setting endpoints for "kube-system/kube-dns:dns" to [10.244.1.3:53]
I1027 22:14:54.038209 5063 proxier.go:352] Setting endpoints for "default/kubernetes:https" to [10.240.0.2:443]
I1027 22:14:54.038238 5063 proxier.go:429] Not syncing iptables until Services and Endpoints have been received from master
I1027 22:14:54.040048 5063 proxier.go:294] Adding new service "default/kubernetes:https" at 10.0.0.1:443/TCP
I1027 22:14:54.040154 5063 proxier.go:294] Adding new service "kube-system/kube-dns:dns" at 10.0.0.10:53/UDP
I1027 22:14:54.040223 5063 proxier.go:294] Adding new service "kube-system/kube-dns:dns-tcp" at 10.0.0.10:53/TCP
If you see error messages about not being able to contact the master, you should double-check your Node configuration and installation steps.
One of the possible reasons that kube-proxy cannot run correctly is that the
required conntrack binary cannot be found. This may happen on some Linux
systems, depending on how you are installing the cluster, for example, you are
installing Kubernetes from scratch. If this is the case, you need to manually
install the conntrack package (e.g. sudo apt install conntrack on Ubuntu)
and then retry.
Kube-proxy can run in one of a few modes. In the log listed above, the
line Using iptables Proxier indicates that kube-proxy is running in
"iptables" mode. The most common other mode is "ipvs". The older "userspace"
mode has largely been replaced by these.
In "iptables" mode, you should see something like the following on a Node:
iptables-save | grep hostnames
-A KUBE-SEP-57KPRZ3JQVENLNBR -s 10.244.3.6/32 -m comment --comment "default/hostnames:" -j MARK --set-xmark 0x00004000/0x00004000
-A KUBE-SEP-57KPRZ3JQVENLNBR -p tcp -m comment --comment "default/hostnames:" -m tcp -j DNAT --to-destination 10.244.3.6:9376
-A KUBE-SEP-WNBA2IHDGP2BOBGZ -s 10.244.1.7/32 -m comment --comment "default/hostnames:" -j MARK --set-xmark 0x00004000/0x00004000
-A KUBE-SEP-WNBA2IHDGP2BOBGZ -p tcp -m comment --comment "default/hostnames:" -m tcp -j DNAT --to-destination 10.244.1.7:9376
-A KUBE-SEP-X3P2623AGDH6CDF3 -s 10.244.2.3/32 -m comment --comment "default/hostnames:" -j MARK --set-xmark 0x00004000/0x00004000
-A KUBE-SEP-X3P2623AGDH6CDF3 -p tcp -m comment --comment "default/hostnames:" -m tcp -j DNAT --to-destination 10.244.2.3:9376
-A KUBE-SERVICES -d 10.0.1.175/32 -p tcp -m comment --comment "default/hostnames: cluster IP" -m tcp --dport 80 -j KUBE-SVC-NWV5X2332I4OT4T3
-A KUBE-SVC-NWV5X2332I4OT4T3 -m comment --comment "default/hostnames:" -m statistic --mode random --probability 0.33332999982 -j KUBE-SEP-WNBA2IHDGP2BOBGZ
-A KUBE-SVC-NWV5X2332I4OT4T3 -m comment --comment "default/hostnames:" -m statistic --mode random --probability 0.50000000000 -j KUBE-SEP-X3P2623AGDH6CDF3
-A KUBE-SVC-NWV5X2332I4OT4T3 -m comment --comment "default/hostnames:" -j KUBE-SEP-57KPRZ3JQVENLNBR
For each port of each Service, there should be 1 rule in KUBE-SERVICES and
one KUBE-SVC-<hash> chain. For each Pod endpoint, there should be a small
number of rules in that KUBE-SVC-<hash> and one KUBE-SEP-<hash> chain with
a small number of rules in it. The exact rules will vary based on your exact
config (including node-ports and load-balancers).
In "ipvs" mode, you should see something like the following on a Node:
ipvsadm -ln
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
...
TCP 10.0.1.175:80 rr
-> 10.244.0.5:9376 Masq 1 0 0
-> 10.244.0.6:9376 Masq 1 0 0
-> 10.244.0.7:9376 Masq 1 0 0
...
For each port of each Service, plus any NodePorts, external IPs, and
load-balancer IPs, kube-proxy will create a virtual server. For each Pod
endpoint, it will create corresponding real servers. In this example, service
hostnames(10.0.1.175:80) has 3 endpoints(10.244.0.5:9376,
10.244.0.6:9376, 10.244.0.7:9376).
In rare cases, you may be using "userspace" mode. From your Node:
iptables-save | grep hostnames
-A KUBE-PORTALS-CONTAINER -d 10.0.1.175/32 -p tcp -m comment --comment "default/hostnames:default" -m tcp --dport 80 -j REDIRECT --to-ports 48577
-A KUBE-PORTALS-HOST -d 10.0.1.175/32 -p tcp -m comment --comment "default/hostnames:default" -m tcp --dport 80 -j DNAT --to-destination 10.240.115.247:48577
There should be 2 rules for each port of your Service (only one in this example) - a "KUBE-PORTALS-CONTAINER" and a "KUBE-PORTALS-HOST".
Almost nobody should be using the "userspace" mode any more, so you won't spend more time on it here.
Assuming you do see one the above cases, try again to access your Service by IP from one of your Nodes:
curl 10.0.1.175:80
hostnames-632524106-bbpiw
If this fails and you are using the userspace proxy, you can try accessing the proxy directly. If you are using the iptables proxy, skip this section.
Look back at the iptables-save output above, and extract the
port number that kube-proxy is using for your Service. In the above
examples it is "48577". Now connect to that:
curl localhost:48577
hostnames-632524106-tlaok
If this still fails, look at the kube-proxy logs for specific lines like:
Setting endpoints for default/hostnames:default to [10.244.0.5:9376 10.244.0.6:9376 10.244.0.7:9376]
If you don't see those, try restarting kube-proxy with the -v flag set to 4, and
then look at the logs again.
This might sound unlikely, but it does happen and it is supposed to work.
This can happen when the network is not properly configured for "hairpin"
traffic, usually when kube-proxy is running in iptables mode and Pods
are connected with bridge network. The Kubelet exposes a hairpin-mode
flag that allows endpoints of a Service to loadbalance
back to themselves if they try to access their own Service VIP. The
hairpin-mode flag must either be set to hairpin-veth or
promiscuous-bridge.
The common steps to trouble shoot this are as follows:
hairpin-mode is set to hairpin-veth or promiscuous-bridge.
You should see something like the below. hairpin-mode is set to
promiscuous-bridge in the following example.ps auxw | grep kubelet
root 3392 1.1 0.8 186804 65208 ? Sl 00:51 11:11 /usr/local/bin/kubelet --enable-debugging-handlers=true --config=/etc/kubernetes/manifests --allow-privileged=True --v=4 --cluster-dns=10.0.0.10 --cluster-domain=cluster.local --configure-cbr0=true --cgroup-root=/ --system-cgroups=/system --hairpin-mode=promiscuous-bridge --runtime-cgroups=/docker-daemon --kubelet-cgroups=/kubelet --babysit-daemons=true --max-pods=110 --serialize-image-pulls=false --outofdisk-transition-frequency=0
hairpin-mode. To do this, you'll have to look at
kubelet log. Accessing the logs depends on your Node OS. On some OSes it
is a file, such as /var/log/kubelet.log, while other OSes use journalctl
to access logs. Please be noted that the effective hairpin mode may not
match --hairpin-mode flag due to compatibility. Check if there is any log
lines with key word hairpin in kubelet.log. There should be log lines
indicating the effective hairpin mode, like something below.I0629 00:51:43.648698 3252 kubelet.go:380] Hairpin mode set to "promiscuous-bridge"
hairpin-veth, ensure the Kubelet has
the permission to operate in /sys on node. If everything works properly,
you should see something like:for intf in /sys/devices/virtual/net/cbr0/brif/*; do cat $intf/hairpin_mode; done
1
1
1
1
promiscuous-bridge, ensure Kubelet
has the permission to manipulate linux bridge on node. If cbr0 bridge is
used and configured properly, you should see:ifconfig cbr0 |grep PROMISC
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1460 Metric:1
If you get this far, something very strange is happening. Your Service is
running, has Endpoints, and your Pods are actually serving. You have DNS
working, and kube-proxy does not seem to be misbehaving. And yet your
Service is not working. Please let us know what is going on, so we can help
investigate!
Contact us on Slack or Forum or GitHub.
Visit troubleshooting document for more information.
Kubernetes v1.11 [stable]
crictl is a command-line interface for CRI-compatible container runtimes.
You can use it to inspect and debug container runtimes and applications on a
Kubernetes node. crictl and its source are hosted in the
cri-tools repository.
crictl requires a Linux operating system with a CRI runtime.
You can download a compressed archive crictl from the cri-tools release
page, for several
different architectures. Download the version that corresponds to your version
of Kubernetes. Extract it and move it to a location on your system path, such as
/usr/local/bin/.
The crictl command has several subcommands and runtime flags. Use
crictl help or crictl <subcommand> help for more details.
crictl connects to unix:///var/run/dockershim.sock by default. For other
runtimes, you can set the endpoint in multiple different ways:
--runtime-endpoint and --image-endpointCONTAINER_RUNTIME_ENDPOINT and IMAGE_SERVICE_ENDPOINT--config=/etc/crictl.yamlYou can also specify timeout values when connecting to the server and enable or
disable debugging, by specifying timeout or debug values in the configuration
file or using the --timeout and --debug command-line flags.
To view or edit the current configuration, view or edit the contents of /etc/crictl.yaml.
cat /etc/crictl.yaml
runtime-endpoint: unix:///var/run/dockershim.sock
image-endpoint: unix:///var/run/dockershim.sock
timeout: 10
debug: true
The following examples show some crictl commands and example output.
crictl to create pod sandboxes or containers on a running
Kubernetes cluster, the Kubelet will eventually delete them. crictl is not a
general purpose workflow tool, but a tool that is useful for debugging.
List all pods:
crictl pods
The output is similar to this:
POD ID CREATED STATE NAME NAMESPACE ATTEMPT
926f1b5a1d33a About a minute ago Ready sh-84d7dcf559-4r2gq default 0
4dccb216c4adb About a minute ago Ready nginx-65899c769f-wv2gp default 0
a86316e96fa89 17 hours ago Ready kube-proxy-gblk4 kube-system 0
919630b8f81f1 17 hours ago Ready nvidia-device-plugin-zgbbv kube-system 0
List pods by name:
crictl pods --name nginx-65899c769f-wv2gp
The output is similar to this:
POD ID CREATED STATE NAME NAMESPACE ATTEMPT
4dccb216c4adb 2 minutes ago Ready nginx-65899c769f-wv2gp default 0
List pods by label:
crictl pods --label run=nginx
The output is similar to this:
POD ID CREATED STATE NAME NAMESPACE ATTEMPT
4dccb216c4adb 2 minutes ago Ready nginx-65899c769f-wv2gp default 0
List all images:
crictl images
The output is similar to this:
IMAGE TAG IMAGE ID SIZE
busybox latest 8c811b4aec35f 1.15MB
k8s-gcrio.azureedge.net/hyperkube-amd64 v1.10.3 e179bbfe5d238 665MB
k8s-gcrio.azureedge.net/pause-amd64 3.1 da86e6ba6ca19 742kB
nginx latest cd5239a0906a6 109MB
List images by repository:
crictl images nginx
The output is similar to this:
IMAGE TAG IMAGE ID SIZE
nginx latest cd5239a0906a6 109MB
Only list image IDs:
crictl images -q
The output is similar to this:
sha256:8c811b4aec35f259572d0f79207bc0678df4c736eeec50bc9fec37ed936a472a
sha256:e179bbfe5d238de6069f3b03fccbecc3fb4f2019af741bfff1233c4d7b2970c5
sha256:da86e6ba6ca197bf6bc5e9d900febd906b133eaa4750e6bed647b0fbe50ed43e
sha256:cd5239a0906a6ccf0562354852fae04bc5b52d72a2aff9a871ddb6bd57553569
List all containers:
crictl ps -a
The output is similar to this:
CONTAINER ID IMAGE CREATED STATE NAME ATTEMPT
1f73f2d81bf98 busybox@sha256:141c253bc4c3fd0a201d32dc1f493bcf3fff003b6df416dea4f41046e0f37d47 7 minutes ago Running sh 1
9c5951df22c78 busybox@sha256:141c253bc4c3fd0a201d32dc1f493bcf3fff003b6df416dea4f41046e0f37d47 8 minutes ago Exited sh 0
87d3992f84f74 nginx@sha256:d0a8828cccb73397acb0073bf34f4d7d8aa315263f1e7806bf8c55d8ac139d5f 8 minutes ago Running nginx 0
1941fb4da154f k8s-gcrio.azureedge.net/hyperkube-amd64@sha256:00d814b1f7763f4ab5be80c58e98140dfc69df107f253d7fdd714b30a714260a 18 hours ago Running kube-proxy 0
List running containers:
crictl ps
The output is similar to this:
CONTAINER ID IMAGE CREATED STATE NAME ATTEMPT
1f73f2d81bf98 busybox@sha256:141c253bc4c3fd0a201d32dc1f493bcf3fff003b6df416dea4f41046e0f37d47 6 minutes ago Running sh 1
87d3992f84f74 nginx@sha256:d0a8828cccb73397acb0073bf34f4d7d8aa315263f1e7806bf8c55d8ac139d5f 7 minutes ago Running nginx 0
1941fb4da154f k8s-gcrio.azureedge.net/hyperkube-amd64@sha256:00d814b1f7763f4ab5be80c58e98140dfc69df107f253d7fdd714b30a714260a 17 hours ago Running kube-proxy 0
crictl exec -i -t 1f73f2d81bf98 ls
The output is similar to this:
bin dev etc home proc root sys tmp usr var
Get all container logs:
crictl logs 87d3992f84f74
The output is similar to this:
10.240.0.96 - - [06/Jun/2018:02:45:49 +0000] "GET / HTTP/1.1" 200 612 "-" "curl/7.47.0" "-"
10.240.0.96 - - [06/Jun/2018:02:45:50 +0000] "GET / HTTP/1.1" 200 612 "-" "curl/7.47.0" "-"
10.240.0.96 - - [06/Jun/2018:02:45:51 +0000] "GET / HTTP/1.1" 200 612 "-" "curl/7.47.0" "-"
Get only the latest N lines of logs:
crictl logs --tail=1 87d3992f84f74
The output is similar to this:
10.240.0.96 - - [06/Jun/2018:02:45:51 +0000] "GET / HTTP/1.1" 200 612 "-" "curl/7.47.0" "-"
Using crictl to run a pod sandbox is useful for debugging container runtimes.
On a running Kubernetes cluster, the sandbox will eventually be stopped and
deleted by the Kubelet.
Create a JSON file like the following:
{
"metadata": {
"name": "nginx-sandbox",
"namespace": "default",
"attempt": 1,
"uid": "hdishd83djaidwnduwk28bcsb"
},
"logDirectory": "/tmp",
"linux": {
}
}
Use the crictl runp command to apply the JSON and run the sandbox.
crictl runp pod-config.json
The ID of the sandbox is returned.
Using crictl to create a container is useful for debugging container runtimes.
On a running Kubernetes cluster, the sandbox will eventually be stopped and
deleted by the Kubelet.
Pull a busybox image
crictl pull busybox
Image is up to date for busybox@sha256:141c253bc4c3fd0a201d32dc1f493bcf3fff003b6df416dea4f41046e0f37d47
Create configs for the pod and the container:
Pod config:
{
"metadata": {
"name": "nginx-sandbox",
"namespace": "default",
"attempt": 1,
"uid": "hdishd83djaidwnduwk28bcsb"
},
"log_directory": "/tmp",
"linux": {
}
}
Container config:
{
"metadata": {
"name": "busybox"
},
"image":{
"image": "busybox"
},
"command": [
"top"
],
"log_path":"busybox.log",
"linux": {
}
}
Create the container, passing the ID of the previously-created pod, the container config file, and the pod config file. The ID of the container is returned.
crictl create f84dd361f8dc51518ed291fbadd6db537b0496536c1d2d6c05ff943ce8c9a54f container-config.json pod-config.json
List all containers and verify that the newly-created container has its
state set to Created.
crictl ps -a
The output is similar to this:
CONTAINER ID IMAGE CREATED STATE NAME ATTEMPT
3e025dd50a72d busybox 32 seconds ago Created busybox 0
To start a container, pass its ID to crictl start:
crictl start 3e025dd50a72d956c4f14881fbb5b1080c9275674e95fb67f965f6478a957d60
The output is similar to this:
3e025dd50a72d956c4f14881fbb5b1080c9275674e95fb67f965f6478a957d60
Check the container has its state set to Running.
crictl ps
The output is similar to this:
CONTAINER ID IMAGE CREATED STATE NAME ATTEMPT
3e025dd50a72d busybox About a minute ago Running busybox 0
See kubernetes-sigs/cri-tools for more information.
The exact versions for below mapping table are for docker cli v1.40 and crictl v1.19.0. Please note that the list is not exhaustive. For example, it doesn't include experimental commands of docker cli.
| docker cli | crictl | Description | Unsupported Features |
|---|---|---|---|
attach |
attach |
Attach to a running container | --detach-keys, --sig-proxy |
exec |
exec |
Run a command in a running container | --privileged, --user, --detach-keys |
images |
images |
List images | |
info |
info |
Display system-wide information | |
inspect |
inspect, inspecti |
Return low-level information on a container, image or task | |
logs |
logs |
Fetch the logs of a container | --details |
ps |
ps |
List containers | |
stats |
stats |
Display a live stream of container(s) resource usage statistics | Column: NET/BLOCK I/O, PIDs |
version |
version |
Show the runtime (Docker, ContainerD, or others) version information |
| docker cli | crictl | Description | Unsupported Features |
|---|---|---|---|
create |
create |
Create a new container | |
kill |
stop (timeout = 0) |
Kill one or more running container | --signal |
pull |
pull |
Pull an image or a repository from a registry | --all-tags, --disable-content-trust |
rm |
rm |
Remove one or more containers | |
rmi |
rmi |
Remove one or more images | |
run |
run |
Run a command in a new container | |
start |
start |
Start one or more stopped containers | --detach-keys |
stop |
stop |
Stop one or more running containers | |
update |
update |
Update configuration of one or more containers | --restart, --blkio-weight and some other resource limit not supported by CRI. |
| crictl | Description |
|---|---|
imagefsinfo |
Return image filesystem info |
inspectp |
Display the status of one or more pods |
port-forward |
Forward local port to a pod |
pods |
List pods |
runp |
Run a new pod |
rmp |
Remove one or more pods |
stopp |
Stop one or more running pods |
This page shows how to write and read a Container termination message.
Termination messages provide a way for containers to write information about fatal events to a location where it can be easily retrieved and surfaced by tools like dashboards and monitoring software. In most cases, information that you put in a termination message should also be written to the general Kubernetes logs.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
In this exercise, you create a Pod that runs one container. The configuration file specifies a command that runs when the container starts.
apiVersion: v1
kind: Pod
metadata:
name: termination-demo
spec:
containers:
- name: termination-demo-container
image: debian
command: ["/bin/sh"]
args: ["-c", "sleep 10 && echo Sleep expired > /dev/termination-log"]
Create a Pod based on the YAML configuration file:
kubectl apply -f https://k8s.io/examples/debug/termination.yaml
In the YAML file, in the command and args fields, you can see that the
container sleeps for 10 seconds and then writes "Sleep expired" to
the /dev/termination-log file. After the container writes
the "Sleep expired" message, it terminates.
Display information about the Pod:
kubectl get pod termination-demo
Repeat the preceding command until the Pod is no longer running.
Display detailed information about the Pod:
kubectl get pod termination-demo --output=yaml
The output includes the "Sleep expired" message:
apiVersion: v1
kind: Pod
...
lastState:
terminated:
containerID: ...
exitCode: 0
finishedAt: ...
message: |
Sleep expired
...
Use a Go template to filter the output so that it includes only the termination message:
kubectl get pod termination-demo -o go-template="{{range .status.containerStatuses}}{{.lastState.terminated.message}}{{end}}"
Kubernetes retrieves termination messages from the termination message file
specified in the terminationMessagePath field of a Container, which has a default
value of /dev/termination-log. By customizing this field, you can tell Kubernetes
to use a different file. Kubernetes use the contents from the specified file to
populate the Container's status message on both success and failure.
The termination message is intended to be brief final status, such as an assertion failure message.
The kubelet truncates messages that are longer than 4096 bytes. The total message length across all
containers will be limited to 12KiB. The default termination message path is /dev/termination-log.
You cannot set the termination message path after a Pod is launched
In the following example, the container writes termination messages to
/tmp/my-log for Kubernetes to retrieve:
apiVersion: v1
kind: Pod
metadata:
name: msg-path-demo
spec:
containers:
- name: msg-path-demo-container
image: debian
terminationMessagePath: "/tmp/my-log"
Moreover, users can set the terminationMessagePolicy field of a Container for
further customization. This field defaults to "File" which means the termination
messages are retrieved only from the termination message file. By setting the
terminationMessagePolicy to "FallbackToLogsOnError", you can tell Kubernetes
to use the last chunk of container log output if the termination message file
is empty and the container exited with an error. The log output is limited to
2048 bytes or 80 lines, whichever is smaller.
terminationMessagePath field in
Container.Kubernetes applications usually consist of multiple, separate services, each running in its own container. Developing and debugging these services on a remote Kubernetes cluster can be cumbersome, requiring you to get a shell on a running container and running your tools inside the remote shell.
telepresence is a tool to ease the process of developing and debugging services locally, while proxying the service to a remote Kubernetes cluster. Using telepresence allows you to use custom tools, such as a debugger and IDE, for a local service and provides the service full access to ConfigMap, secrets, and the services running on the remote cluster.
This document describes using telepresence to develop and debug services running on a remote cluster locally.
kubectl is configured to communicate with the clusterOpen a terminal and run telepresence with no arguments to get a telepresence shell. This shell runs locally, giving you full access to your local filesystem.
The telepresence shell can be used in a variety of ways. For example, write a shell script on your laptop, and run it directly from the shell in real time. You can do this on a remote shell as well, but you might not be able to use your preferred code editor, and the script is deleted when the container is terminated.
Enter exit to quit and close the shell.
When developing an application on Kubernetes, you typically program or debug a single service. The service might require access to other services for testing and debugging. One option is to use the continuous deployment pipeline, but even the fastest deployment pipeline introduces a delay in the program or debug cycle.
Use the --swap-deployment option to swap an existing deployment with the Telepresence proxy. Swapping allows you to run a service locally and connect to the remote Kubernetes cluster. The services in the remote cluster can now access the locally running instance.
To run telepresence with --swap-deployment, enter:
telepresence --swap-deployment $DEPLOYMENT_NAME
where $DEPLOYMENT_NAME is the name of your existing deployment.
Running this command spawns a shell. In the shell, start your service. You can then make edits to the source code locally, save, and see the changes take effect immediately. You can also run your service in a debugger, or any other local development tool.
If you're interested in a hands-on tutorial, check out this tutorial that walks through locally developing the Guestbook application on Google Kubernetes Engine.
Telepresence has numerous proxying options, depending on your situation.
For further reading, visit the Telepresence website.
This page shows how to use kubectl exec to get a shell to a
running container.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
In this exercise, you create a Pod that has one container. The container runs the nginx image. Here is the configuration file for the Pod:
apiVersion: v1
kind: Pod
metadata:
name: shell-demo
spec:
volumes:
- name: shared-data
emptyDir: {}
containers:
- name: nginx
image: nginx
volumeMounts:
- name: shared-data
mountPath: /usr/share/nginx/html
hostNetwork: true
dnsPolicy: Default
Create the Pod:
kubectl apply -f https://k8s.io/examples/application/shell-demo.yaml
Verify that the container is running:
kubectl get pod shell-demo
Get a shell to the running container:
kubectl exec --stdin --tty shell-demo -- /bin/bash
--) separates the arguments you want to pass to the command from the kubectl arguments.
In your shell, list the root directory:
# Run this inside the container
ls /
In your shell, experiment with other commands. Here are some examples:
# You can run these example commands inside the container
ls /
cat /proc/mounts
cat /proc/1/maps
apt-get update
apt-get install -y tcpdump
tcpdump
apt-get install -y lsof
lsof
apt-get install -y procps
ps aux
ps aux | grep nginx
Look again at the configuration file for your Pod. The Pod
has an emptyDir volume, and the container mounts the volume
at /usr/share/nginx/html.
In your shell, create an index.html file in the /usr/share/nginx/html
directory:
# Run this inside the container
echo 'Hello shell demo' > /usr/share/nginx/html/index.html
In your shell, send a GET request to the nginx server:
# Run this in the shell inside your container
apt-get update
apt-get install curl
curl http://localhost/
The output shows the text that you wrote to the index.html file:
Hello shell demo
When you are finished with your shell, enter exit.
exit # To quit the shell in the container
In an ordinary command window, not your shell, list the environment variables in the running container:
kubectl exec shell-demo env
Experiment with running other commands. Here are some examples:
kubectl exec shell-demo -- ps aux
kubectl exec shell-demo -- ls /
kubectl exec shell-demo -- cat /proc/1/mounts
If a Pod has more than one container, use --container or -c to
specify a container in the kubectl exec command. For example,
suppose you have a Pod named my-pod, and the Pod has two containers
named main-app and helper-app. The following command would open a
shell to the main-app container.
kubectl exec -i -t my-pod --container main-app -- /bin/bash
-i and -t are the same as the long options --stdin and --tty
Node Problem Detector is a daemon for monitoring and reporting about a node's health.
You can run Node Problem Detector as a DaemonSet or as a standalone daemon.
Node Problem Detector collects information about node problems from various daemons
and reports these conditions to the API server as NodeCondition
and Event.
To learn how to install and use Node Problem Detector, see Node Problem Detector project documentation.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Node Problem Detector only supports file based kernel log.
Log tools such as journald are not supported.
Node Problem Detector uses the kernel log format for reporting kernel issues. To learn how to extend the kernel log format, see Add support for another log format.
Some cloud providers enable Node Problem Detector as an Addon.
You can also enable Node Problem Detector with kubectl or by creating an Addon pod.
kubectl provides the most flexible management of Node Problem Detector.
You can overwrite the default configuration to fit it into your environment or
to detect customized node problems. For example:
Create a Node Problem Detector configuration similar to node-problem-detector.yaml:
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: node-problem-detector-v0.1
namespace: kube-system
labels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
spec:
selector:
matchLabels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
template:
metadata:
labels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
spec:
hostNetwork: true
containers:
- name: node-problem-detector
image: k8s.gcr.io/node-problem-detector:v0.1
securityContext:
privileged: true
resources:
limits:
cpu: "200m"
memory: "100Mi"
requests:
cpu: "20m"
memory: "20Mi"
volumeMounts:
- name: log
mountPath: /log
readOnly: true
volumes:
- name: log
hostPath:
path: /var/log/Start node problem detector with kubectl:
kubectl apply -f https://k8s.io/examples/debug/node-problem-detector.yaml
If you are using a custom cluster bootstrap solution and don't need to overwrite the default configuration, you can leverage the Addon pod to further automate the deployment.
Create node-problem-detector.yaml, and save the configuration in the Addon pod's
directory /etc/kubernetes/addons/node-problem-detector on a control plane node.
The default configuration is embedded when building the Docker image of Node Problem Detector.
However, you can use a ConfigMap
to overwrite the configuration:
Change the configuration files in config/
Create the ConfigMap node-problem-detector-config:
kubectl create configmap node-problem-detector-config --from-file=config/
Change the node-problem-detector.yaml to use the ConfigMap:
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: node-problem-detector-v0.1
namespace: kube-system
labels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
spec:
selector:
matchLabels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
template:
metadata:
labels:
k8s-app: node-problem-detector
version: v0.1
kubernetes.io/cluster-service: "true"
spec:
hostNetwork: true
containers:
- name: node-problem-detector
image: k8s.gcr.io/node-problem-detector:v0.1
securityContext:
privileged: true
resources:
limits:
cpu: "200m"
memory: "100Mi"
requests:
cpu: "20m"
memory: "20Mi"
volumeMounts:
- name: log
mountPath: /log
readOnly: true
- name: config # Overwrite the config/ directory with ConfigMap volume
mountPath: /config
readOnly: true
volumes:
- name: log
hostPath:
path: /var/log/
- name: config # Define ConfigMap volume
configMap:
name: node-problem-detector-configRecreate the Node Problem Detector with the new configuration file:
# If you have a node-problem-detector running, delete before recreating
kubectl delete -f https://k8s.io/examples/debug/node-problem-detector.yaml
kubectl apply -f https://k8s.io/examples/debug/node-problem-detector-configmap.yaml
kubectl.
Overwriting a configuration is not supported if a Node Problem Detector runs as a cluster Addon.
The Addon manager does not support ConfigMap.
Kernel Monitor is a system log monitor daemon supported in the Node Problem Detector. Kernel monitor watches the kernel log and detects known kernel issues following predefined rules.
The Kernel Monitor matches kernel issues according to a set of predefined rule list in
config/kernel-monitor.json. The rule list is extensible. You can expand the rule list by overwriting the
configuration.
To support a new NodeCondition, create a condition definition within the conditions field in
config/kernel-monitor.json, for example:
{
"type": "NodeConditionType",
"reason": "CamelCaseDefaultNodeConditionReason",
"message": "arbitrary default node condition message"
}
To detect new problems, you can extend the rules field in config/kernel-monitor.json
with a new rule definition:
{
"type": "temporary/permanent",
"condition": "NodeConditionOfPermanentIssue",
"reason": "CamelCaseShortReason",
"message": "regexp matching the issue in the kernel log"
}
Check your kernel log path location in your operating system (OS) distribution.
The Linux kernel log device is usually presented as /dev/kmsg. However, the log path location varies by OS distribution.
The log field in config/kernel-monitor.json represents the log path inside the container.
You can configure the log field to match the device path as seen by the Node Problem Detector.
Kernel monitor uses the
Translator plugin to translate the internal data structure of the kernel log.
You can implement a new translator for a new log format.
It is recommended to run the Node Problem Detector in your cluster to monitor node health. When running the Node Problem Detector, you can expect extra resource overhead on each node. Usually this is fine, because:
Resource usage metrics, such as container CPU and memory usage,
are available in Kubernetes through the Metrics API. These metrics can be accessed either directly
by the user with the kubectl top command, or by a controller in the cluster, for example
Horizontal Pod Autoscaler, to make decisions.
Through the Metrics API, you can get the amount of resource currently used by a given node or a given pod. This API doesn't store the metric values, so it's not possible, for example, to get the amount of resources used by a given node 10 minutes ago.
The API is no different from any other API:
/apis/metrics.k8s.io/The API is defined in k8s.io/metrics repository. You can find more information about the API there.
CPU is reported as the average usage, in CPU cores, over a period of time. This value is derived by taking a rate over a cumulative CPU counter provided by the kernel (in both Linux and Windows kernels). The kubelet chooses the window for the rate calculation.
Memory is reported as the working set, in bytes, at the instant the metric was collected. In an ideal world, the "working set" is the amount of memory in-use that cannot be freed under memory pressure. However, calculation of the working set varies by host OS, and generally makes heavy use of heuristics to produce an estimate. It includes all anonymous (non-file-backed) memory since Kubernetes does not support swap. The metric typically also includes some cached (file-backed) memory, because the host OS cannot always reclaim such pages.
Metrics Server is a cluster-wide aggregator of resource usage data.
By default, it is deployed in clusters created by kube-up.sh script
as a Deployment object. If you use a different Kubernetes setup mechanism, you can deploy it using the provided
deployment components.yaml file.
Metrics Server collects metrics from the Summary API, exposed by Kubelet on each node, and is registered with the main API server via Kubernetes aggregator.
Learn more about the metrics server in the design doc.
To scale an application and provide a reliable service, you need to understand how the application behaves when it is deployed. You can examine application performance in a Kubernetes cluster by examining the containers, pods, services, and the characteristics of the overall cluster. Kubernetes provides detailed information about an application's resource usage at each of these levels. This information allows you to evaluate your application's performance and where bottlenecks can be removed to improve overall performance.
In Kubernetes, application monitoring does not depend on a single monitoring solution. On new clusters, you can use resource metrics or full metrics pipelines to collect monitoring statistics.
The resource metrics pipeline provides a limited set of metrics related to
cluster components such as the
Horizontal Pod Autoscaler
controller, as well as the kubectl top utility.
These metrics are collected by the lightweight, short-term, in-memory
metrics-server and
are exposed via the metrics.k8s.io API.
metrics-server discovers all nodes on the cluster and
queries each node's
kubelet for CPU and
memory usage. The kubelet acts as a bridge between the Kubernetes master and
the nodes, managing the pods and containers running on a machine. The kubelet
translates each pod into its constituent containers and fetches individual
container usage statistics from the container runtime through the container
runtime interface. The kubelet fetches this information from the integrated
cAdvisor for the legacy Docker integration. It then exposes the aggregated pod
resource usage statistics through the metrics-server Resource Metrics API.
This API is served at /metrics/resource/v1beta1 on the kubelet's authenticated and
read-only ports.
A full metrics pipeline gives you access to richer metrics. Kubernetes can
respond to these metrics by automatically scaling or adapting the cluster
based on its current state, using mechanisms such as the Horizontal Pod
Autoscaler. The monitoring pipeline fetches metrics from the kubelet and
then exposes them to Kubernetes via an adapter by implementing either the
custom.metrics.k8s.io or external.metrics.k8s.io API.
Prometheus, a CNCF project, can natively monitor Kubernetes, nodes, and Prometheus itself. Full metrics pipeline projects that are not part of the CNCF are outside the scope of Kubernetes documentation.
This guide is to help users debug applications that are deployed into Kubernetes and not behaving correctly. This is not a guide for people who want to debug their cluster. For that you should check out this guide.
The first step in troubleshooting is triage. What is the problem? Is it your Pods, your Replication Controller or your Service?
The first step in debugging a Pod is taking a look at it. Check the current state of the Pod and recent events with the following command:
kubectl describe pods ${POD_NAME}
Look at the state of the containers in the pod. Are they all Running? Have there been recent restarts?
Continue debugging depending on the state of the pods.
If a Pod is stuck in Pending it means that it can not be scheduled onto a node. Generally this is because
there are insufficient resources of one type or another that prevent scheduling. Look at the output of the
kubectl describe ... command above. There should be messages from the scheduler about why it can not schedule
your pod. Reasons include:
You don't have enough resources: You may have exhausted the supply of CPU or Memory in your cluster, in this case you need to delete Pods, adjust resource requests, or add new nodes to your cluster. See Compute Resources document for more information.
You are using hostPort: When you bind a Pod to a hostPort there are a limited number of places that pod can be
scheduled. In most cases, hostPort is unnecessary, try using a Service object to expose your Pod. If you do require
hostPort then you can only schedule as many Pods as there are nodes in your Kubernetes cluster.
If a Pod is stuck in the Waiting state, then it has been scheduled to a worker node, but it can't run on that machine.
Again, the information from kubectl describe ... should be informative. The most common cause of Waiting pods is a failure to pull the image. There are three things to check:
docker pull <image> on your machine to see if the image can be pulled.Once your pod has been scheduled, the methods described in Debug Running Pods are available for debugging.
If your pod is not behaving as you expected, it may be that there was an error in your
pod description (e.g. mypod.yaml file on your local machine), and that the error
was silently ignored when you created the pod. Often a section of the pod description
is nested incorrectly, or a key name is typed incorrectly, and so the key is ignored.
For example, if you misspelled command as commnd then the pod will be created but
will not use the command line you intended it to use.
The first thing to do is to delete your pod and try creating it again with the --validate option.
For example, run kubectl apply --validate -f mypod.yaml.
If you misspelled command as commnd then will give an error like this:
I0805 10:43:25.129850 46757 schema.go:126] unknown field: commnd
I0805 10:43:25.129973 46757 schema.go:129] this may be a false alarm, see https://github.com/kubernetes/kubernetes/issues/6842
pods/mypod
The next thing to check is whether the pod on the apiserver
matches the pod you meant to create (e.g. in a yaml file on your local machine).
For example, run kubectl get pods/mypod -o yaml > mypod-on-apiserver.yaml and then
manually compare the original pod description, mypod.yaml with the one you got
back from apiserver, mypod-on-apiserver.yaml. There will typically be some
lines on the "apiserver" version that are not on the original version. This is
expected. However, if there are lines on the original that are not on the apiserver
version, then this may indicate a problem with your pod spec.
Replication controllers are fairly straightforward. They can either create Pods or they can't. If they can't create pods, then please refer to the instructions above to debug your pods.
You can also use kubectl describe rc ${CONTROLLER_NAME} to introspect events related to the replication
controller.
Services provide load balancing across a set of pods. There are several common problems that can make Services not work properly. The following instructions should help debug Service problems.
First, verify that there are endpoints for the service. For every Service object, the apiserver makes an endpoints resource available.
You can view this resource with:
kubectl get endpoints ${SERVICE_NAME}
Make sure that the endpoints match up with the number of pods that you expect to be members of your service. For example, if your Service is for an nginx container with 3 replicas, you would expect to see three different IP addresses in the Service's endpoints.
If you are missing endpoints, try listing pods using the labels that Service uses. Imagine that you have a Service where the labels are:
...
spec:
- selector:
name: nginx
type: frontend
You can use:
kubectl get pods --selector=name=nginx,type=frontend
to list pods that match this selector. Verify that the list matches the Pods that you expect to provide your Service.
Verify that the pod's containerPort matches up with the Service's targetPort
Please see debugging service for more information.
If none of the above solves your problem, follow the instructions in
Debugging Service document
to make sure that your Service is running, has Endpoints, and your Pods are
actually serving; you have DNS working, iptables rules installed, and kube-proxy
does not seem to be misbehaving.
You may also visit troubleshooting document for more information.
This doc is about cluster troubleshooting; we assume you have already ruled out your application as the root cause of the problem you are experiencing. See the application troubleshooting guide for tips on application debugging. You may also visit troubleshooting document for more information.
The first thing to debug in your cluster is if your nodes are all registered correctly.
Run
kubectl get nodes
And verify that all of the nodes you expect to see are present and that they are all in the Ready state.
To get detailed information about the overall health of your cluster, you can run:
kubectl cluster-info dump
For now, digging deeper into the cluster requires logging into the relevant machines. Here are the locations
of the relevant log files. (note that on systemd-based systems, you may need to use journalctl instead)
/var/log/kube-apiserver.log - API Server, responsible for serving the API/var/log/kube-scheduler.log - Scheduler, responsible for making scheduling decisions/var/log/kube-controller-manager.log - Controller that manages replication controllers/var/log/kubelet.log - Kubelet, responsible for running containers on the node/var/log/kube-proxy.log - Kube Proxy, responsible for service load balancingThis is an incomplete list of things that could go wrong, and how to adjust your cluster setup to mitigate the problems.
Action: Use IaaS provider's automatic VM restarting feature for IaaS VMs
Action: Use IaaS providers reliable storage (e.g. GCE PD or AWS EBS volume) for VMs with apiserver+etcd
Action: Use high-availability configuration
Action: Snapshot apiserver PDs/EBS-volumes periodically
Action: use replication controller and services in front of pods
Action: applications (containers) designed to tolerate unexpected restarts
Sometimes things go wrong. This guide is aimed at making them right. It has two sections:
You should also check the known issues for the release you're using.
If your problem isn't answered by any of the guides above, there are variety of ways for you to get help from the Kubernetes community.
The documentation on this site has been structured to provide answers to a wide
range of questions. Concepts explain the Kubernetes
architecture and how each component works, while Setup provides
practical instructions for getting started. Tasks show how to
accomplish commonly used tasks, and Tutorials are more
comprehensive walkthroughs of real-world, industry-specific, or end-to-end
development scenarios. The Reference section provides
detailed documentation on the Kubernetes API
and command-line interfaces (CLIs), such as kubectl.
Someone else from the community may have already asked a similar question or may be able to help with your problem. The Kubernetes team will also monitor posts tagged Kubernetes. If there aren't any existing questions that help, please ask a new one!
Many people from the Kubernetes community hang out on Kubernetes Slack in the #kubernetes-users channel.
Slack requires registration; you can request an invitation,
and registration is open to everyone). Feel free to come and ask any and all questions.
Once registered, access the Kubernetes organisation in Slack
via your web browser or via Slack's own dedicated app.
Once you are registered, browse the growing list of channels for various subjects of
interest. For example, people new to Kubernetes may also want to join the
#kubernetes-novice channel. As another example, developers should join the
#kubernetes-dev channel.
There are also many country specific / local language channels. Feel free to join these channels for localized support and info:
| Country | Channels |
|---|---|
| China | #cn-users, #cn-events |
| Finland | #fi-users |
| France | #fr-users, #fr-events |
| Germany | #de-users, #de-events |
| India | #in-users, #in-events |
| Italy | #it-users, #it-events |
| Japan | #jp-users, #jp-events |
| Korea | #kr-users |
| Netherlands | #nl-users |
| Norway | #norw-users |
| Poland | #pl-users |
| Russia | #ru-users |
| Spain | #es-users |
| Sweden | #se-users |
| Turkey | #tr-users, #tr-events |
You're welcome to join the official Kubernetes Forum: discuss.kubernetes.io.
If you have what looks like a bug, or you would like to make a feature request, please use the GitHub issue tracking system.
Before you file an issue, please search existing issues to see if your issue is already covered.
If filing a bug, please include detailed information about how to reproduce the problem, such as:
kubectl versionConfiguring the aggregation layer allows the Kubernetes apiserver to be extended with additional APIs, which are not part of the core Kubernetes APIs.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Unlike Custom Resource Definitions (CRDs), the Aggregation API involves another server - your Extension apiserver - in addition to the standard Kubernetes apiserver. The Kubernetes apiserver will need to communicate with your extension apiserver, and your extension apiserver will need to communicate with the Kubernetes apiserver. In order for this communication to be secured, the Kubernetes apiserver uses x509 certificates to authenticate itself to the extension apiserver.
This section describes how the authentication and authorization flows work, and how to configure them.
The high-level flow is as follows:
The rest of this section describes these steps in detail.
The flow can be seen in the following diagram.
.
The source for the above swimlanes can be found in the source of this document.
A request to an API path that is served by an extension apiserver begins the same way as all API requests: communication to the Kubernetes apiserver. This path already has been registered with the Kubernetes apiserver by the extension apiserver.
The user communicates with the Kubernetes apiserver, requesting access to the path. The Kubernetes apiserver uses standard authentication and authorization configured with the Kubernetes apiserver to authenticate the user and authorize access to the specific path.
For an overview of authenticating to a Kubernetes cluster, see "Authenticating to a Cluster". For an overview of authorization of access to Kubernetes cluster resources, see "Authorization Overview".
Everything to this point has been standard Kubernetes API requests, authentication and authorization.
The Kubernetes apiserver now is prepared to send the request to the extension apiserver.
The Kubernetes apiserver now will send, or proxy, the request to the extension apiserver that registered to handle the request. In order to do so, it needs to know several things:
In order to provide for these two, you must configure the Kubernetes apiserver using several flags.
The Kubernetes apiserver connects to the extension apiserver over TLS, authenticating itself using a client certificate. You must provide the following to the Kubernetes apiserver upon startup, using the provided flags:
--proxy-client-key-file--proxy-client-cert-file--requestheader-client-ca-file--requestheader-allowed-namesThe Kubernetes apiserver will use the files indicated by --proxy-client-*-file to authenticate to the extension apiserver. In order for the request to be considered valid by a compliant extension apiserver, the following conditions must be met:
--requestheader-client-ca-file.--requestheader-allowed-names.--requestheader-allowed-names="". This will indicate to an extension apiserver that any CN is acceptable.
When started with these options, the Kubernetes apiserver will:
kube-system namespace called extension-apiserver-authentication, in which it will place the CA certificate and the allowed CNs. These in turn can be retrieved by extension apiservers to validate requests.Note that the same client certificate is used by the Kubernetes apiserver to authenticate against all extension apiservers. It does not create a client certificate per extension apiserver, but rather a single one to authenticate as the Kubernetes apiserver. This same one is reused for all extension apiserver requests.
When the Kubernetes apiserver proxies the request to the extension apiserver, it informs the extension apiserver of the username and group with which the original request successfully authenticated. It provides these in http headers of its proxied request. You must inform the Kubernetes apiserver of the names of the headers to be used.
--requestheader-username-headers--requestheader-group-headers--requestheader-extra-headers-prefixThese header names are also placed in the extension-apiserver-authentication configmap, so they can be retrieved and used by extension apiservers.
The extension apiserver, upon receiving a proxied request from the Kubernetes apiserver, must validate that the request actually did come from a valid authenticating proxy, which role the Kubernetes apiserver is fulfilling. The extension apiserver validates it via:
kube-system, as described above:
If the above passes, then the request is a valid proxied request from a legitimate authenticating proxy, in this case the Kubernetes apiserver.
Note that it is the responsibility of the extension apiserver implementation to provide the above. Many do it by default, leveraging the k8s.io/apiserver/ package. Others may provide options to override it using command-line options.
In order to have permission to retrieve the configmap, an extension apiserver requires the appropriate role. There is a default role named extension-apiserver-authentication-reader in the kube-system namespace which can be assigned.
The extension apiserver now can validate that the user/group retrieved from the headers are authorized to execute the given request. It does so by sending a standard SubjectAccessReview request to the Kubernetes apiserver.
In order for the extension apiserver to be authorized itself to submit the SubjectAccessReview request to the Kubernetes apiserver, it needs the correct permissions. Kubernetes includes a default ClusterRole named system:auth-delegator that has the appropriate permissions. It can be granted to the extension apiserver's service account.
If the SubjectAccessReview passes, the extension apiserver executes the request.
Enable the aggregation layer via the following kube-apiserver flags. They may have already been taken care of by your provider.
--requestheader-client-ca-file=<path to aggregator CA cert>
--requestheader-allowed-names=front-proxy-client
--requestheader-extra-headers-prefix=X-Remote-Extra-
--requestheader-group-headers=X-Remote-Group
--requestheader-username-headers=X-Remote-User
--proxy-client-cert-file=<path to aggregator proxy cert>
--proxy-client-key-file=<path to aggregator proxy key>
The Kubernetes apiserver has two client CA options:
--client-ca-file--requestheader-client-ca-fileEach of these functions independently and can conflict with each other, if not used correctly.
--client-ca-file: When a request arrives to the Kubernetes apiserver, if this option is enabled, the Kubernetes apiserver checks the certificate of the request. If it is signed by one of the CA certificates in the file referenced by --client-ca-file, then the request is treated as a legitimate request, and the user is the value of the common name CN=, while the group is the organization O=. See the documentation on TLS authentication.--requestheader-client-ca-file: When a request arrives to the Kubernetes apiserver, if this option is enabled, the Kubernetes apiserver checks the certificate of the request. If it is signed by one of the CA certificates in the file reference by --requestheader-client-ca-file, then the request is treated as a potentially legitimate request. The Kubernetes apiserver then checks if the common name CN= is one of the names in the list provided by --requestheader-allowed-names. If the name is allowed, the request is approved; if it is not, the request is not.If both --client-ca-file and --requestheader-client-ca-file are provided, then the request first checks the --requestheader-client-ca-file CA and then the --client-ca-file. Normally, different CAs, either root CAs or intermediate CAs, are used for each of these options; regular client requests match against --client-ca-file, while aggregation requests match against --requestheader-client-ca-file. However, if both use the same CA, then client requests that normally would pass via --client-ca-file will fail, because the CA will match the CA in --requestheader-client-ca-file, but the common name CN= will not match one of the acceptable common names in --requestheader-allowed-names. This can cause your kubelets and other control plane components, as well as end-users, to be unable to authenticate to the Kubernetes apiserver.
For this reason, use different CA certs for the --client-ca-file option - to authorize control plane components and end-users - and the --requestheader-client-ca-file option - to authorize aggregation apiserver requests.
If you are not running kube-proxy on a host running the API server, then you must make sure that the system is enabled with the following kube-apiserver flag:
--enable-aggregator-routing=true
You can dynamically configure what client requests are proxied to extension apiserver. The following is an example registration:
apiVersion: apiregistration.k8s.io/v1
kind: APIService
metadata:
name: <name of the registration object>
spec:
group: <API group name this extension apiserver hosts>
version: <API version this extension apiserver hosts>
groupPriorityMinimum: <priority this APIService for this group, see API documentation>
versionPriority: <prioritizes ordering of this version within a group, see API documentation>
service:
namespace: <namespace of the extension apiserver service>
name: <name of the extension apiserver service>
caBundle: <pem encoded ca cert that signs the server cert used by the webhook>
The name of an APIService object must be a valid path segment name.
Once the Kubernetes apiserver has determined a request should be sent to an extension apiserver, it needs to know how to contact it.
The service stanza is a reference to the service for an extension apiserver.
The service namespace and name are required. The port is optional and defaults to 443.
Here is an example of an extension apiserver that is configured to be called on port "1234",
and to verify the TLS connection against the ServerName
my-service-name.my-service-namespace.svc using a custom CA bundle.
apiVersion: apiregistration.k8s.io/v1
kind: APIService
...
spec:
...
service:
namespace: my-service-namespace
name: my-service-name
port: 1234
caBundle: "Ci0tLS0tQk...<base64-encoded PEM bundle>...tLS0K"
...
This page shows how to install a custom resource into the Kubernetes API by creating a CustomResourceDefinition.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.16. To check the version, enterkubectl version.
If you are using an older version of Kubernetes that is still supported, switch to
the documentation for that version to see advice that is relevant for your cluster.
When you create a new CustomResourceDefinition (CRD), the Kubernetes API Server
creates a new RESTful resource path for each version you specify. The CRD can be
either namespaced or cluster-scoped, as specified in the CRD's scope field. As
with existing built-in objects, deleting a namespace deletes all custom objects
in that namespace. CustomResourceDefinitions themselves are non-namespaced and
are available to all namespaces.
For example, if you save the following CustomResourceDefinition to resourcedefinition.yaml:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.stable.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: stable.example.com
# list of versions supported by this CustomResourceDefinition
versions:
- name: v1
# Each version can be enabled/disabled by Served flag.
served: true
# One and only one version must be marked as the storage version.
storage: true
schema:
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
image:
type: string
replicas:
type: integer
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the CamelCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
and create it:
kubectl apply -f resourcedefinition.yaml
Then a new namespaced RESTful API endpoint is created at:
/apis/stable.example.com/v1/namespaces/*/crontabs/...
This endpoint URL can then be used to create and manage custom objects.
The kind of these objects will be CronTab from the spec of the
CustomResourceDefinition object you created above.
It might take a few seconds for the endpoint to be created.
You can watch the Established condition of your CustomResourceDefinition
to be true or watch the discovery information of the API server for your
resource to show up.
After the CustomResourceDefinition object has been created, you can create
custom objects. Custom objects can contain custom fields. These fields can
contain arbitrary JSON.
In the following example, the cronSpec and image custom fields are set in a
custom object of kind CronTab. The kind CronTab comes from the spec of the
CustomResourceDefinition object you created above.
If you save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * * */5"
image: my-awesome-cron-image
and create it:
kubectl apply -f my-crontab.yaml
You can then manage your CronTab objects using kubectl. For example:
kubectl get crontab
Should print a list like this:
NAME AGE
my-new-cron-object 6s
Resource names are not case-sensitive when using kubectl, and you can use either the singular or plural forms defined in the CRD, as well as any short names.
You can also view the raw YAML data:
kubectl get ct -o yaml
You should see that it contains the custom cronSpec and image fields
from the YAML you used to create it:
apiVersion: v1
items:
- apiVersion: stable.example.com/v1
kind: CronTab
metadata:
annotations:
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"stable.example.com/v1","kind":"CronTab","metadata":{"annotations":{},"name":"my-new-cron-object","namespace":"default"},"spec":{"cronSpec":"* * * * */5","image":"my-awesome-cron-image"}}
creationTimestamp: "2021-06-20T07:35:27Z"
generation: 1
name: my-new-cron-object
namespace: default
resourceVersion: "1326"
uid: 9aab1d66-628e-41bb-a422-57b8b3b1f5a9
spec:
cronSpec: '* * * * */5'
image: my-awesome-cron-image
kind: List
metadata:
resourceVersion: ""
selfLink: ""
When you delete a CustomResourceDefinition, the server will uninstall the RESTful API endpoint and delete all custom objects stored in it.
kubectl delete -f resourcedefinition.yaml
kubectl get crontabs
Error from server (NotFound): Unable to list {"stable.example.com" "v1" "crontabs"}: the server could not find the requested resource (get crontabs.stable.example.com)
If you later recreate the same CustomResourceDefinition, it will start out empty.
CustomResources store structured data in custom fields (alongside the built-in
fields apiVersion, kind and metadata, which the API server validates
implicitly). With OpenAPI v3.0 validation a schema can be
specified, which is validated during creation and updates, compare below for
details and limits of such a schema.
With apiextensions.k8s.io/v1 the definition of a structural schema is
mandatory for CustomResourceDefinitions. In the beta version of
CustomResourceDefinition, the structural schema was optional.
A structural schema is an OpenAPI v3.0 validation schema which:
type in OpenAPI) for the root, for each specified field of an object node (via properties or additionalProperties in OpenAPI) and for each item in an array node (via items in OpenAPI), with the exception of:
x-kubernetes-int-or-string: truex-kubernetes-preserve-unknown-fields: trueallOf, anyOf, oneOf or not, the schema also specifies the field/item outside of those logical junctors (compare example 1 and 2).description, type, default, additionalProperties, nullable within an allOf, anyOf, oneOf or not, with the exception of the two pattern for x-kubernetes-int-or-string: true (see below).metadata is specified, then only restrictions on metadata.name and metadata.generateName are allowed.Non-structural example 1:
allOf:
- properties:
foo:
...
conflicts with rule 2. The following would be correct:
properties:
foo:
...
allOf:
- properties:
foo:
...
Non-structural example 2:
allOf:
- items:
properties:
foo:
...
conflicts with rule 2. The following would be correct:
items:
properties:
foo:
...
allOf:
- items:
properties:
foo:
...
Non-structural example 3:
properties:
foo:
pattern: "abc"
metadata:
type: object
properties:
name:
type: string
pattern: "^a"
finalizers:
type: array
items:
type: string
pattern: "my-finalizer"
anyOf:
- properties:
bar:
type: integer
minimum: 42
required: ["bar"]
description: "foo bar object"
is not a structural schema because of the following violations:
foo is missing (rule 1).bar inside of anyOf is not specified outside (rule 2).bar's type is within anyOf (rule 3).anyOf (rule 3).metadata.finalizers might not be restricted (rule 4).In contrast, the following, corresponding schema is structural:
type: object
description: "foo bar object"
properties:
foo:
type: string
pattern: "abc"
bar:
type: integer
metadata:
type: object
properties:
name:
type: string
pattern: "^a"
anyOf:
- properties:
bar:
minimum: 42
required: ["bar"]
Violations of the structural schema rules are reported in the NonStructural condition in the CustomResourceDefinition.
CustomResourceDefinitions store validated resource data in the cluster's persistence store, etcd. As with native Kubernetes resources such as ConfigMap, if you specify a field that the API server does not recognize, the unknown field is pruned (removed) before being persisted.
CRDs converted from apiextensions.k8s.io/v1beta1 to apiextensions.k8s.io/v1 might lack structural schemas, and spec.preserveUnknownFields might be true.
For legacy CustomResourceDefinition objects created as
apiextensions.k8s.io/v1beta1 with spec.preserveUnknownFields set to
true, the following is also true:
For compatibility with apiextensions.k8s.io/v1, update your custom
resource definitions to:
spec.preserveUnknownFields to false.If you save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * * */5"
image: my-awesome-cron-image
someRandomField: 42
and create it:
kubectl create --validate=false -f my-crontab.yaml -o yaml
your output is similar to:
apiVersion: stable.example.com/v1
kind: CronTab
metadata:
creationTimestamp: 2017-05-31T12:56:35Z
generation: 1
name: my-new-cron-object
namespace: default
resourceVersion: "285"
uid: 9423255b-4600-11e7-af6a-28d2447dc82b
spec:
cronSpec: '* * * * */5'
image: my-awesome-cron-image
Notice that the field someRandomField was pruned.
This example turned off client-side validation to demonstrate the API server's behavior, by adding the --validate=false command line option.
Because the OpenAPI validation schemas are also published
to clients, kubectl also checks for unknown fields and rejects those objects well before they would be sent to the API server.
By default, all unspecified fields for a custom resource, across all versions, are pruned. It is possible though to opt-out of that for specifc sub-trees of fields by adding x-kubernetes-preserve-unknown-fields: true in the structural OpenAPI v3 validation schema.
For example:
type: object
properties:
json:
x-kubernetes-preserve-unknown-fields: true
The field json can store any JSON value, without anything being pruned.
You can also partially specify the permitted JSON; for example:
type: object
properties:
json:
x-kubernetes-preserve-unknown-fields: true
type: object
description: this is arbitrary JSON
With this, only object type values are allowed.
Pruning is enabled again for each specified property (or additionalProperties):
type: object
properties:
json:
x-kubernetes-preserve-unknown-fields: true
type: object
properties:
spec:
type: object
properties:
foo:
type: string
bar:
type: string
With this, the value:
json:
spec:
foo: abc
bar: def
something: x
status:
something: x
is pruned to:
json:
spec:
foo: abc
bar: def
status:
something: x
This means that the something field in the specified spec object is pruned, but everything outside is not.
Nodes in a schema with x-kubernetes-int-or-string: true are excluded from rule 1, such that the following is structural:
type: object
properties:
foo:
x-kubernetes-int-or-string: true
Also those nodes are partially excluded from rule 3 in the sense that the following two patterns are allowed (exactly those, without variations in order to additional fields):
x-kubernetes-int-or-string: true
anyOf:
- type: integer
- type: string
...
and
x-kubernetes-int-or-string: true
allOf:
- anyOf:
- type: integer
- type: string
- ... # zero or more
...
With one of those specification, both an integer and a string validate.
In Validation Schema Publishing,
x-kubernetes-int-or-string: true is unfolded to one of the two patterns shown above.
RawExtensions (as in runtime.RawExtension defined in
k8s.io/apimachinery)
holds complete Kubernetes objects, i.e. with apiVersion and kind fields.
It is possible to specify those embedded objects (both completely without constraints or partially specified) by setting x-kubernetes-embedded-resource: true. For example:
type: object
properties:
foo:
x-kubernetes-embedded-resource: true
x-kubernetes-preserve-unknown-fields: true
Here, the field foo holds a complete object, e.g.:
foo:
apiVersion: v1
kind: Pod
spec:
...
Because x-kubernetes-preserve-unknown-fields: true is specified alongside, nothing is pruned. The use of x-kubernetes-preserve-unknown-fields: true is optional though.
With x-kubernetes-embedded-resource: true, the apiVersion, kind and metadata are implicitly specified and validated.
See Custom resource definition versioning for more information about serving multiple versions of your CustomResourceDefinition and migrating your objects from one version to another.
Finalizers allow controllers to implement asynchronous pre-delete hooks. Custom objects support finalizers similar to built-in objects.
You can add a finalizer to a custom object like this:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
finalizers:
- stable.example.com/finalizer
Identifiers of custom finalizers consist of a domain name, a forward slash and the name of the finalizer. Any controller can add a finalizer to any object's list of finalizers.
The first delete request on an object with finalizers sets a value for the
metadata.deletionTimestamp field but does not delete it. Once this value is set,
entries in the finalizers list can only be removed. While any finalizers remain it is also
impossible to force the deletion of an object.
When the metadata.deletionTimestamp field is set, controllers watching the object execute any
finalizers they handle and remove the finalizer from the list after they are done. It is the
responsibility of each controller to remove its finalizer from the list.
The value of metadata.deletionGracePeriodSeconds controls the interval between polling updates.
Once the list of finalizers is empty, meaning all finalizers have been executed, the resource is deleted by Kubernetes.
Custom resources are validated via OpenAPI v3 schemas and you can add additional validation using admission webhooks.
Additionally, the following restrictions are applied to the schema:
definitions,dependencies,deprecated,discriminator,id,patternProperties,readOnly,writeOnly,xml,$ref.uniqueItems cannot be set to true.additionalProperties cannot be set to false.additionalProperties is mutually exclusive with properties.The default field can be set when the Defaulting feature is enabled,
which is the case with apiextensions.k8s.io/v1 CustomResourceDefinitions.
Defaulting is in GA since 1.17 (beta since 1.16 with the CustomResourceDefaulting
feature gate
enabled, which is the case automatically for many clusters for beta features).
Refer to the structural schemas section for other restrictions and CustomResourceDefinition features.
The schema is defined in the CustomResourceDefinition. In the following example, the CustomResourceDefinition applies the following validations on the custom object:
spec.cronSpec must be a string and must be of the form described by the regular expression.spec.replicas must be an integer and must have a minimum value of 1 and a maximum value of 10.Save the CustomResourceDefinition to resourcedefinition.yaml:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: crontabs.stable.example.com
spec:
group: stable.example.com
versions:
- name: v1
served: true
storage: true
schema:
# openAPIV3Schema is the schema for validating custom objects.
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
pattern: '^(\d+|\*)(/\d+)?(\s+(\d+|\*)(/\d+)?){4}$'
image:
type: string
replicas:
type: integer
minimum: 1
maximum: 10
scope: Namespaced
names:
plural: crontabs
singular: crontab
kind: CronTab
shortNames:
- ct
and create it:
kubectl apply -f resourcedefinition.yaml
A request to create a custom object of kind CronTab is rejected if there are invalid values in its fields. In the following example, the custom object contains fields with invalid values:
spec.cronSpec does not match the regular expression.spec.replicas is greater than 10.If you save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * *"
image: my-awesome-cron-image
replicas: 15
and attempt to create it:
kubectl apply -f my-crontab.yaml
then you get an error:
The CronTab "my-new-cron-object" is invalid: []: Invalid value: map[string]interface {}{"apiVersion":"stable.example.com/v1", "kind":"CronTab", "metadata":map[string]interface {}{"name":"my-new-cron-object", "namespace":"default", "deletionTimestamp":interface {}(nil), "deletionGracePeriodSeconds":(*int64)(nil), "creationTimestamp":"2017-09-05T05:20:07Z", "uid":"e14d79e7-91f9-11e7-a598-f0761cb232d1", "clusterName":""}, "spec":map[string]interface {}{"cronSpec":"* * * *", "image":"my-awesome-cron-image", "replicas":15}}:
validation failure list:
spec.cronSpec in body should match '^(\d+|\*)(/\d+)?(\s+(\d+|\*)(/\d+)?){4}$'
spec.replicas in body should be less than or equal to 10
If the fields contain valid values, the object creation request is accepted.
Save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * * */5"
image: my-awesome-cron-image
replicas: 5
And create it:
kubectl apply -f my-crontab.yaml
crontab "my-new-cron-object" created
apiextensions.k8s.io/v1.
Defaulting allows to specify default values in the OpenAPI v3 validation schema:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: crontabs.stable.example.com
spec:
group: stable.example.com
versions:
- name: v1
served: true
storage: true
schema:
# openAPIV3Schema is the schema for validating custom objects.
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
pattern: '^(\d+|\*)(/\d+)?(\s+(\d+|\*)(/\d+)?){4}$'
default: "5 0 * * *"
image:
type: string
replicas:
type: integer
minimum: 1
maximum: 10
default: 1
scope: Namespaced
names:
plural: crontabs
singular: crontab
kind: CronTab
shortNames:
- ct
With this both cronSpec and replicas are defaulted:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
image: my-awesome-cron-image
leads to
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "5 0 * * *"
image: my-awesome-cron-image
replicas: 1
Defaulting happens on the object
Defaults applied when reading data from etcd are not automatically written back to etcd. An update request via the API is required to persist those defaults back into etcd.
Default values must be pruned (with the exception of defaults for metadata fields) and must validate against a provided schema.
Default values for metadata fields of x-kubernetes-embedded-resources: true nodes (or parts of a default value covering metadata) are not pruned during CustomResourceDefinition creation, but through the pruning step during handling of requests.
New in 1.20: null values for fields that either don't specify the nullable flag, or give it a false value, will be pruned before defaulting happens. If a default is present, it will be applied. When nullable is true, null values will be conserved and won't be defaulted.
For example, given the OpenAPI schema below:
type: object
properties:
spec:
type: object
properties:
foo:
type: string
nullable: false
default: "default"
bar:
type: string
nullable: true
baz:
type: string
creating an object with null values for foo and bar and baz
spec:
foo: null
bar: null
baz: null
leads to
spec:
foo: "default"
bar: null
with foo pruned and defaulted because the field is non-nullable, bar maintaining the null value due to nullable: true, and baz pruned because the field is non-nullable and has no default.
CustomResourceDefinition OpenAPI v3 validation schemas which are structural and enable pruning are published as part of the OpenAPI v2 spec from Kubernetes API server.
The kubectl command-line tool consumes the published schema to perform client-side validation (kubectl create and kubectl apply), schema explanation (kubectl explain) on custom resources. The published schema can be consumed for other purposes as well, like client generation or documentation.
The OpenAPI v3 validation schema is converted to OpenAPI v2 schema, and
show up in definitions and paths fields in the OpenAPI v2 spec.
The following modifications are applied during the conversion to keep backwards compatibility with kubectl in previous 1.13 version. These modifications prevent kubectl from being over-strict and rejecting valid OpenAPI schemas that it doesn't understand. The conversion won't modify the validation schema defined in CRD, and therefore won't affect validation in the API server.
allOf, anyOf, oneOf and not are removednullable: true is set, we drop type, nullable, items and properties because OpenAPI v2 is not able to express nullable. To avoid kubectl to reject good objects, this is necessary.The kubectl tool relies on server-side output formatting. Your cluster's API server decides which
columns are shown by the kubectl get command. You can customize these columns for a
CustomResourceDefinition. The following example adds the Spec, Replicas, and Age
columns.
Save the CustomResourceDefinition to resourcedefinition.yaml:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: crontabs.stable.example.com
spec:
group: stable.example.com
scope: Namespaced
names:
plural: crontabs
singular: crontab
kind: CronTab
shortNames:
- ct
versions:
- name: v1
served: true
storage: true
schema:
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
image:
type: string
replicas:
type: integer
additionalPrinterColumns:
- name: Spec
type: string
description: The cron spec defining the interval a CronJob is run
jsonPath: .spec.cronSpec
- name: Replicas
type: integer
description: The number of jobs launched by the CronJob
jsonPath: .spec.replicas
- name: Age
type: date
jsonPath: .metadata.creationTimestamp
Create the CustomResourceDefinition:
kubectl apply -f resourcedefinition.yaml
Create an instance using the my-crontab.yaml from the previous section.
Invoke the server-side printing:
kubectl get crontab my-new-cron-object
Notice the NAME, SPEC, REPLICAS, and AGE columns in the output:
NAME SPEC REPLICAS AGE
my-new-cron-object * * * * * 1 7s
NAME column is implicit and does not need to be defined in the CustomResourceDefinition.
Each column includes a priority field. Currently, the priority
differentiates between columns shown in standard view or wide view (using the -o wide flag).
0 are shown in standard view.0 are shown only in wide view.A column's type field can be any of the following (compare OpenAPI v3 data types):
integer – non-floating-point numbersnumber – floating point numbersstring – stringsboolean – true or falsedate – rendered differentially as time since this timestamp.If the value inside a CustomResource does not match the type specified for the column, the value is omitted. Use CustomResource validation to ensure that the value types are correct.
A column's format field can be any of the following:
int32int64floatdoublebytedatedate-timepasswordThe column's format controls the style used when kubectl prints the value.
Custom resources support /status and /scale subresources.
The status and scale subresources can be optionally enabled by defining them in the CustomResourceDefinition.
When the status subresource is enabled, the /status subresource for the custom resource is exposed.
The status and the spec stanzas are represented by the .status and .spec JSONPaths respectively inside of a custom resource.
PUT requests to the /status subresource take a custom resource object and ignore changes to anything except the status stanza.
PUT requests to the /status subresource only validate the status stanza of the custom resource.
PUT/POST/PATCH requests to the custom resource ignore changes to the status stanza.
The .metadata.generation value is incremented for all changes, except for changes to .metadata or .status.
Only the following constructs are allowed at the root of the CRD OpenAPI validation schema:
descriptionexampleexclusiveMaximumexclusiveMinimumexternalDocsformatitemsmaximummaxItemsmaxLengthminimumminItemsminLengthmultipleOfpatternpropertiesrequiredtitletypeuniqueItemsWhen the scale subresource is enabled, the /scale subresource for the custom resource is exposed.
The autoscaling/v1.Scale object is sent as the payload for /scale.
To enable the scale subresource, the following fields are defined in the CustomResourceDefinition.
specReplicasPath defines the JSONPath inside of a custom resource that corresponds to scale.spec.replicas.
.spec and with the dot notation are allowed.specReplicasPath in the custom resource,
the /scale subresource will return an error on GET.statusReplicasPath defines the JSONPath inside of a custom resource that corresponds to scale.status.replicas.
.status and with the dot notation are allowed.statusReplicasPath in the custom resource,
the status replica value in the /scale subresource will default to 0.labelSelectorPath defines the JSONPath inside of a custom resource that corresponds to Scale.Status.Selector.
.status or .spec and with the dot notation are allowed.labelSelectorPath in the custom resource,
the status selector value in the /scale subresource will default to the empty string.In the following example, both status and scale subresources are enabled.
Save the CustomResourceDefinition to resourcedefinition.yaml:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: crontabs.stable.example.com
spec:
group: stable.example.com
versions:
- name: v1
served: true
storage: true
schema:
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
image:
type: string
replicas:
type: integer
status:
type: object
properties:
replicas:
type: integer
labelSelector:
type: string
# subresources describes the subresources for custom resources.
subresources:
# status enables the status subresource.
status: {}
# scale enables the scale subresource.
scale:
# specReplicasPath defines the JSONPath inside of a custom resource that corresponds to Scale.Spec.Replicas.
specReplicasPath: .spec.replicas
# statusReplicasPath defines the JSONPath inside of a custom resource that corresponds to Scale.Status.Replicas.
statusReplicasPath: .status.replicas
# labelSelectorPath defines the JSONPath inside of a custom resource that corresponds to Scale.Status.Selector.
labelSelectorPath: .status.labelSelector
scope: Namespaced
names:
plural: crontabs
singular: crontab
kind: CronTab
shortNames:
- ct
And create it:
kubectl apply -f resourcedefinition.yaml
After the CustomResourceDefinition object has been created, you can create custom objects.
If you save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * * */5"
image: my-awesome-cron-image
replicas: 3
and create it:
kubectl apply -f my-crontab.yaml
Then new namespaced RESTful API endpoints are created at:
/apis/stable.example.com/v1/namespaces/*/crontabs/status
and
/apis/stable.example.com/v1/namespaces/*/crontabs/scale
A custom resource can be scaled using the kubectl scale command.
For example, the following command sets .spec.replicas of the
custom resource created above to 5:
kubectl scale --replicas=5 crontabs/my-new-cron-object
crontabs "my-new-cron-object" scaled
kubectl get crontabs my-new-cron-object -o jsonpath='{.spec.replicas}'
5
You can use a PodDisruptionBudget to protect custom resources that have the scale subresource enabled.
Categories is a list of grouped resources the custom resource belongs to (eg. all).
You can use kubectl get <category-name> to list the resources belonging to the category.
The following example adds all in the list of categories in the CustomResourceDefinition
and illustrates how to output the custom resource using kubectl get all.
Save the following CustomResourceDefinition to resourcedefinition.yaml:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
name: crontabs.stable.example.com
spec:
group: stable.example.com
versions:
- name: v1
served: true
storage: true
schema:
openAPIV3Schema:
type: object
properties:
spec:
type: object
properties:
cronSpec:
type: string
image:
type: string
replicas:
type: integer
scope: Namespaced
names:
plural: crontabs
singular: crontab
kind: CronTab
shortNames:
- ct
# categories is a list of grouped resources the custom resource belongs to.
categories:
- all
and create it:
kubectl apply -f resourcedefinition.yaml
After the CustomResourceDefinition object has been created, you can create custom objects.
Save the following YAML to my-crontab.yaml:
apiVersion: "stable.example.com/v1"
kind: CronTab
metadata:
name: my-new-cron-object
spec:
cronSpec: "* * * * */5"
image: my-awesome-cron-image
and create it:
kubectl apply -f my-crontab.yaml
You can specify the category when using kubectl get:
kubectl get all
and it will include the custom resources of kind CronTab:
NAME AGE
crontabs/my-new-cron-object 3s
Read about custom resources.
Serve multiple versions of a CustomResourceDefinition.
This page explains how to add versioning information to CustomResourceDefinitions, to indicate the stability level of your CustomResourceDefinitions or advance your API to a new version with conversion between API representations. It also describes how to upgrade an object from one version to another.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
You should have a initial understanding of custom resources.
Your Kubernetes server must be at or later than version v1.16. To check the version, enterkubectl version.
The CustomResourceDefinition API provides a workflow for introducing and upgrading to new versions of a CustomResourceDefinition.
When a CustomResourceDefinition is created, the first version is set in the
CustomResourceDefinition spec.versions list to an appropriate stability level
and a version number. For example v1beta1 would indicate that the first
version is not yet stable. All custom resource objects will initially be stored
at this version.
Once the CustomResourceDefinition is created, clients may begin using the
v1beta1 API.
Later it might be necessary to add new version such as v1.
Adding a new version:
None
conversion strategy may be used and only the apiVersion field will be
modified when serving different versions.spec.versions list with served:true. Also, set spec.conversion field
to the selected conversion strategy. If using a conversion webhook, configure
spec.conversion.webhookClientConfig field to call the webhook.Once the new version is added, clients may incrementally migrate to the new version. It is perfectly safe for some clients to use the old version while others use the new version.
Migrate stored objects to the new version:
It is safe for clients to use both the old and new version before, during and after upgrading the objects to a new stored version.
Removing an old version:
served to false for the old version in the spec.versions list. If
any clients are still unexpectedly using the old version they may begin reporting
errors attempting to access the custom resource objects at the old version.
If this occurs, switch back to using served:true on the old version, migrate the
remaining clients to the new version and repeat this step.storage is set to true for the new version in the spec.versions list in the CustomResourceDefinition.status.storedVersions.spec.versions list.The CustomResourceDefinition API versions field can be used to support multiple versions of custom resources that you
have developed. Versions can have different schemas, and conversion webhooks can convert custom resources between versions.
Webhook conversions should follow the Kubernetes API conventions wherever applicable.
Specifically, See the API change documentation for a set of useful gotchas and suggestions.
apiextensions.k8s.io/v1beta1, there was a version field instead of versions. The
version field is deprecated and optional, but if it is not empty, it must
match the first item in the versions field.
This example shows a CustomResourceDefinition with two versions. For the first example, the assumption is all versions share the same schema with no conversion between them. The comments in the YAML provide more context.
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: example.com
# list of versions supported by this CustomResourceDefinition
versions:
- name: v1beta1
# Each version can be enabled/disabled by Served flag.
served: true
# One and only one version must be marked as the storage version.
storage: true
# A schema is required
schema:
openAPIV3Schema:
type: object
properties:
host:
type: string
port:
type: string
- name: v1
served: true
storage: false
schema:
openAPIV3Schema:
type: object
properties:
host:
type: string
port:
type: string
# The conversion section is introduced in Kubernetes 1.13+ with a default value of
# None conversion (strategy sub-field set to None).
conversion:
# None conversion assumes the same schema for all versions and only sets the apiVersion
# field of custom resources to the proper value
strategy: None
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the CamelCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: example.com
# list of versions supported by this CustomResourceDefinition
versions:
- name: v1beta1
# Each version can be enabled/disabled by Served flag.
served: true
# One and only one version must be marked as the storage version.
storage: true
- name: v1
served: true
storage: false
validation:
openAPIV3Schema:
type: object
properties:
host:
type: string
port:
type: string
# The conversion section is introduced in Kubernetes 1.13+ with a default value of
# None conversion (strategy sub-field set to None).
conversion:
# None conversion assumes the same schema for all versions and only sets the apiVersion
# field of custom resources to the proper value
strategy: None
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the PascalCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
You can save the CustomResourceDefinition in a YAML file, then use
kubectl apply to create it.
kubectl apply -f my-versioned-crontab.yaml
After creation, the API server starts to serve each enabled version at an HTTP
REST endpoint. In the above example, the API versions are available at
/apis/example.com/v1beta1 and /apis/example.com/v1.
Regardless of the order in which versions are defined in a CustomResourceDefinition, the version with the highest priority is used by kubectl as the default version to access objects. The priority is determined by parsing the name field to determine the version number, the stability (GA, Beta, or Alpha), and the sequence within that stability level.
The algorithm used for sorting the versions is designed to sort versions in the
same way that the Kubernetes project sorts Kubernetes versions. Versions start with a
v followed by a number, an optional beta or alpha designation, and
optional additional numeric versioning information. Broadly, a version string might look
like v2 or v2beta1. Versions are sorted using the following algorithm:
beta or alpha follow the first numeric portion, they sorted
in that order, after the equivalent string without the beta or alpha
suffix (which is presumed to be the GA version).beta, or alpha, those numbers are also
sorted from largest to smallest.foo1 is sorted above foo10. This is different from the sorting of the
numeric portion of entries that do follow the Kubernetes version patterns.This might make sense if you look at the following sorted version list:
- v10
- v2
- v1
- v11beta2
- v10beta3
- v3beta1
- v12alpha1
- v11alpha2
- foo1
- foo10
For the example in Specify multiple versions, the
version sort order is v1, followed by v1beta1. This causes the kubectl
command to use v1 as the default version unless the provided object specifies
the version.
Kubernetes v1.19 [stable]
Starting in v1.19, a CustomResourceDefinition can indicate a particular version of the resource it defines is deprecated. When API requests to a deprecated version of that resource are made, a warning message is returned in the API response as a header. The warning message for each deprecated version of the resource can be customized if desired.
A customized warning message should indicate the deprecated API group, version, and kind, and should indicate what API group, version, and kind should be used instead, if applicable.
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
name: crontabs.example.com
spec:
group: example.com
names:
plural: crontabs
singular: crontab
kind: CronTab
scope: Namespaced
versions:
- name: v1alpha1
served: true
# This indicates the v1alpha1 version of the custom resource is deprecated.
# API requests to this version receive a warning header in the server response.
deprecated: true
# This overrides the default warning returned to API clients making v1alpha1 API requests.
deprecationWarning: "example.com/v1alpha1 CronTab is deprecated; see http://example.com/v1alpha1-v1 for instructions to migrate to example.com/v1 CronTab"
schema: ...
- name: v1beta1
served: true
# This indicates the v1beta1 version of the custom resource is deprecated.
# API requests to this version receive a warning header in the server response.
# A default warning message is returned for this version.
deprecated: true
schema: ...
- name: v1
served: true
storage: true
schema: ...
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
metadata:
name: crontabs.example.com
spec:
group: example.com
names:
plural: crontabs
singular: crontab
kind: CronTab
scope: Namespaced
validation: ...
versions:
- name: v1alpha1
served: true
# This indicates the v1alpha1 version of the custom resource is deprecated.
# API requests to this version receive a warning header in the server response.
deprecated: true
# This overrides the default warning returned to API clients making v1alpha1 API requests.
deprecationWarning: "example.com/v1alpha1 CronTab is deprecated; see http://example.com/v1alpha1-v1 for instructions to migrate to example.com/v1 CronTab"
- name: v1beta1
served: true
# This indicates the v1beta1 version of the custom resource is deprecated.
# API requests to this version receive a warning header in the server response.
# A default warning message is returned for this version.
deprecated: true
- name: v1
served: true
storage: true
Kubernetes v1.16 [stable]
CustomResourceWebhookConversion feature must be enabled, which is the case automatically for many clusters for beta features. Please refer to the feature gate documentation for more information.
The above example has a None conversion between versions which only sets the apiVersion field
on conversion and does not change the rest of the object. The API server also supports webhook
conversions that call an external service in case a conversion is required. For example when:
To cover all of these cases and to optimize conversion by the API server, the conversion requests may contain multiple objects in order to minimize the external calls. The webhook should perform these conversions independently.
Please refer to the implementation of the custom resource conversion webhook
server
that is validated in a Kubernetes e2e test. The webhook handles the
ConversionReview requests sent by the API servers, and sends back conversion
results wrapped in ConversionResponse. Note that the request
contains a list of custom resources that need to be converted independently without
changing the order of objects.
The example server is organized in a way to be reused for other conversions.
Most of the common code are located in the
framework file
that leaves only
one function
to be implemented for different conversions.
ClientAuth field
empty,
which defaults to NoClientCert. This means that the webhook server does not
authenticate the identity of the clients, supposedly API servers. If you need
mutual TLS or other ways to authenticate the clients, see
how to authenticate API servers.
A conversion webhook must not mutate anything inside of metadata of the converted object
other than labels and annotations.
Attempted changes to name, UID and namespace are rejected and fail the request
which caused the conversion. All other changes are ignored.
Documentation for deploying the conversion webhook is the same as for the
admission webhook example service.
The assumption for next sections is that the conversion webhook server is deployed to a service
named example-conversion-webhook-server in default namespace and serving traffic on path /crdconvert.
The None conversion example can be extended to use the conversion webhook by modifying conversion
section of the spec:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: example.com
# list of versions supported by this CustomResourceDefinition
versions:
- name: v1beta1
# Each version can be enabled/disabled by Served flag.
served: true
# One and only one version must be marked as the storage version.
storage: true
# Each version can define it's own schema when there is no top-level
# schema is defined.
schema:
openAPIV3Schema:
type: object
properties:
hostPort:
type: string
- name: v1
served: true
storage: false
schema:
openAPIV3Schema:
type: object
properties:
host:
type: string
port:
type: string
conversion:
# a Webhook strategy instruct API server to call an external webhook for any conversion between custom resources.
strategy: Webhook
# webhook is required when strategy is `Webhook` and it configures the webhook endpoint to be called by API server.
webhook:
# conversionReviewVersions indicates what ConversionReview versions are understood/preferred by the webhook.
# The first version in the list understood by the API server is sent to the webhook.
# The webhook must respond with a ConversionReview object in the same version it received.
conversionReviewVersions: ["v1","v1beta1"]
clientConfig:
service:
namespace: default
name: example-conversion-webhook-server
path: /crdconvert
caBundle: "Ci0tLS0tQk...<base64-encoded PEM bundle>...tLS0K"
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the CamelCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
metadata:
# name must match the spec fields below, and be in the form: <plural>.<group>
name: crontabs.example.com
spec:
# group name to use for REST API: /apis/<group>/<version>
group: example.com
# prunes object fields that are not specified in OpenAPI schemas below.
preserveUnknownFields: false
# list of versions supported by this CustomResourceDefinition
versions:
- name: v1beta1
# Each version can be enabled/disabled by Served flag.
served: true
# One and only one version must be marked as the storage version.
storage: true
# Each version can define it's own schema when there is no top-level
# schema is defined.
schema:
openAPIV3Schema:
type: object
properties:
hostPort:
type: string
- name: v1
served: true
storage: false
schema:
openAPIV3Schema:
type: object
properties:
host:
type: string
port:
type: string
conversion:
# a Webhook strategy instruct API server to call an external webhook for any conversion between custom resources.
strategy: Webhook
# webhookClientConfig is required when strategy is `Webhook` and it configures the webhook endpoint to be called by API server.
webhookClientConfig:
service:
namespace: default
name: example-conversion-webhook-server
path: /crdconvert
caBundle: "Ci0tLS0tQk...<base64-encoded PEM bundle>...tLS0K"
# either Namespaced or Cluster
scope: Namespaced
names:
# plural name to be used in the URL: /apis/<group>/<version>/<plural>
plural: crontabs
# singular name to be used as an alias on the CLI and for display
singular: crontab
# kind is normally the CamelCased singular type. Your resource manifests use this.
kind: CronTab
# shortNames allow shorter string to match your resource on the CLI
shortNames:
- ct
You can save the CustomResourceDefinition in a YAML file, then use
kubectl apply to apply it.
kubectl apply -f my-versioned-crontab-with-conversion.yaml
Make sure the conversion service is up and running before applying new changes.
Once the API server has determined a request should be sent to a conversion webhook,
it needs to know how to contact the webhook. This is specified in the webhookClientConfig
stanza of the webhook configuration.
Conversion webhooks can either be called via a URL or a service reference, and can optionally include a custom CA bundle to use to verify the TLS connection.
url gives the location of the webhook, in standard URL form
(scheme://host:port/path).
The host should not refer to a service running in the cluster; use
a service reference by specifying the service field instead.
The host might be resolved via external DNS in some apiservers
(i.e., kube-apiserver cannot resolve in-cluster DNS as that would
be a layering violation). host may also be an IP address.
Please note that using localhost or 127.0.0.1 as a host is
risky unless you take great care to run this webhook on all hosts
which run an apiserver which might need to make calls to this
webhook. Such installations are likely to be non-portable or not readily run in a new cluster.
The scheme must be "https"; the URL must begin with "https://".
Attempting to use a user or basic auth (for example "user:password@") is not allowed. Fragments ("#...") and query parameters ("?...") are also not allowed.
Here is an example of a conversion webhook configured to call a URL (and expects the TLS certificate to be verified using system trust roots, so does not specify a caBundle):
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
webhook:
clientConfig:
url: "https://my-webhook.example.com:9443/my-webhook-path"
...
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
webhookClientConfig:
url: "https://my-webhook.example.com:9443/my-webhook-path"
...
The service stanza inside webhookClientConfig is a reference to the service for a conversion webhook.
If the webhook is running within the cluster, then you should use service instead of url.
The service namespace and name are required. The port is optional and defaults to 443.
The path is optional and defaults to "/".
Here is an example of a webhook that is configured to call a service on port "1234"
at the subpath "/my-path", and to verify the TLS connection against the ServerName
my-service-name.my-service-namespace.svc using a custom CA bundle.
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
webhook:
clientConfig:
service:
namespace: my-service-namespace
name: my-service-name
path: /my-path
port: 1234
caBundle: "Ci0tLS0tQk...<base64-encoded PEM bundle>...tLS0K"
...
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
webhookClientConfig:
service:
namespace: my-service-namespace
name: my-service-name
path: /my-path
port: 1234
caBundle: "Ci0tLS0tQk...<base64-encoded PEM bundle>...tLS0K"
...
Webhooks are sent a POST request, with Content-Type: application/json,
with a ConversionReview API object in the apiextensions.k8s.io API group
serialized to JSON as the body.
Webhooks can specify what versions of ConversionReview objects they accept
with the conversionReviewVersions field in their CustomResourceDefinition:
apiVersion: apiextensions.k8s.io/v1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
webhook:
conversionReviewVersions: ["v1", "v1beta1"]
...
conversionReviewVersions is a required field when creating
apiextensions.k8s.io/v1 custom resource definitions.
Webhooks are required to support at least one ConversionReview
version understood by the current and previous API server.
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
apiVersion: apiextensions.k8s.io/v1beta1
kind: CustomResourceDefinition
...
spec:
...
conversion:
strategy: Webhook
conversionReviewVersions: ["v1", "v1beta1"]
...
If no conversionReviewVersions are specified, the default when creating
apiextensions.k8s.io/v1beta1 custom resource definitions is v1beta1.
API servers send the first ConversionReview version in the conversionReviewVersions list they support.
If none of the versions in the list are supported by the API server, the custom resource definition will not be allowed to be created.
If an API server encounters a conversion webhook configuration that was previously created and does not support any of the ConversionReview
versions the API server knows how to send, attempts to call to the webhook will fail.
This example shows the data contained in an ConversionReview object
for a request to convert CronTab objects to example.com/v1:
{
"apiVersion": "apiextensions.k8s.io/v1",
"kind": "ConversionReview",
"request": {
# Random uid uniquely identifying this conversion call
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002",
# The API group and version the objects should be converted to
"desiredAPIVersion": "example.com/v1",
# The list of objects to convert.
# May contain one or more objects, in one or more versions.
"objects": [
{
"kind": "CronTab",
"apiVersion": "example.com/v1beta1",
"metadata": {
"creationTimestamp": "2019-09-04T14:03:02Z",
"name": "local-crontab",
"namespace": "default",
"resourceVersion": "143",
"uid": "3415a7fc-162b-4300-b5da-fd6083580d66"
},
"hostPort": "localhost:1234"
},
{
"kind": "CronTab",
"apiVersion": "example.com/v1beta1",
"metadata": {
"creationTimestamp": "2019-09-03T13:02:01Z",
"name": "remote-crontab",
"resourceVersion": "12893",
"uid": "359a83ec-b575-460d-b553-d859cedde8a0"
},
"hostPort": "example.com:2345"
}
]
}
}
{
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
"apiVersion": "apiextensions.k8s.io/v1beta1",
"kind": "ConversionReview",
"request": {
# Random uid uniquely identifying this conversion call
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002",
# The API group and version the objects should be converted to
"desiredAPIVersion": "example.com/v1",
# The list of objects to convert.
# May contain one or more objects, in one or more versions.
"objects": [
{
"kind": "CronTab",
"apiVersion": "example.com/v1beta1",
"metadata": {
"creationTimestamp": "2019-09-04T14:03:02Z",
"name": "local-crontab",
"namespace": "default",
"resourceVersion": "143",
"uid": "3415a7fc-162b-4300-b5da-fd6083580d66"
},
"hostPort": "localhost:1234"
},
{
"kind": "CronTab",
"apiVersion": "example.com/v1beta1",
"metadata": {
"creationTimestamp": "2019-09-03T13:02:01Z",
"name": "remote-crontab",
"resourceVersion": "12893",
"uid": "359a83ec-b575-460d-b553-d859cedde8a0"
},
"hostPort": "example.com:2345"
}
]
}
}
Webhooks respond with a 200 HTTP status code, Content-Type: application/json,
and a body containing a ConversionReview object (in the same version they were sent),
with the response stanza populated, serialized to JSON.
If conversion succeeds, a webhook should return a response stanza containing the following fields:
uid, copied from the request.uid sent to the webhookresult, set to {"status":"Success"}convertedObjects, containing all of the objects from request.objects, converted to request.desiredVersionExample of a minimal successful response from a webhook:
{
"apiVersion": "apiextensions.k8s.io/v1",
"kind": "ConversionReview",
"response": {
# must match <request.uid>
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002",
"result": {
"status": "Success"
},
# Objects must match the order of request.objects, and have apiVersion set to <request.desiredAPIVersion>.
# kind, metadata.uid, metadata.name, and metadata.namespace fields must not be changed by the webhook.
# metadata.labels and metadata.annotations fields may be changed by the webhook.
# All other changes to metadata fields by the webhook are ignored.
"convertedObjects": [
{
"kind": "CronTab",
"apiVersion": "example.com/v1",
"metadata": {
"creationTimestamp": "2019-09-04T14:03:02Z",
"name": "local-crontab",
"namespace": "default",
"resourceVersion": "143",
"uid": "3415a7fc-162b-4300-b5da-fd6083580d66"
},
"host": "localhost",
"port": "1234"
},
{
"kind": "CronTab",
"apiVersion": "example.com/v1",
"metadata": {
"creationTimestamp": "2019-09-03T13:02:01Z",
"name": "remote-crontab",
"resourceVersion": "12893",
"uid": "359a83ec-b575-460d-b553-d859cedde8a0"
},
"host": "example.com",
"port": "2345"
}
]
}
}
{
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
"apiVersion": "apiextensions.k8s.io/v1beta1",
"kind": "ConversionReview",
"response": {
# must match <request.uid>
"uid": "705ab4f5-6393-11e8-b7cc-42010a800002",
"result": {
"status": "Failed"
},
# Objects must match the order of request.objects, and have apiVersion set to <request.desiredAPIVersion>.
# kind, metadata.uid, metadata.name, and metadata.namespace fields must not be changed by the webhook.
# metadata.labels and metadata.annotations fields may be changed by the webhook.
# All other changes to metadata fields by the webhook are ignored.
"convertedObjects": [
{
"kind": "CronTab",
"apiVersion": "example.com/v1",
"metadata": {
"creationTimestamp": "2019-09-04T14:03:02Z",
"name": "local-crontab",
"namespace": "default",
"resourceVersion": "143",
"uid": "3415a7fc-162b-4300-b5da-fd6083580d66"
},
"host": "localhost",
"port": "1234"
},
{
"kind": "CronTab",
"apiVersion": "example.com/v1",
"metadata": {
"creationTimestamp": "2019-09-03T13:02:01Z",
"name": "remote-crontab",
"resourceVersion": "12893",
"uid": "359a83ec-b575-460d-b553-d859cedde8a0"
},
"host": "example.com",
"port": "2345"
}
]
}
}
If conversion fails, a webhook should return a response stanza containing the following fields:
uid, copied from the request.uid sent to the webhookresult, set to {"status":"Failed"}Example of a response from a webhook indicating a conversion request failed, with an optional message:
{
"apiVersion": "apiextensions.k8s.io/v1",
"kind": "ConversionReview",
"response": {
"uid": "<value from request.uid>",
"result": {
"status": "Failed",
"message": "hostPort could not be parsed into a separate host and port"
}
}
}
{
# Deprecated in v1.16 in favor of apiextensions.k8s.io/v1
"apiVersion": "apiextensions.k8s.io/v1beta1",
"kind": "ConversionReview",
"response": {
"uid": "<value from request.uid>",
"result": {
"status": "Failed",
"message": "hostPort could not be parsed into a separate host and port"
}
}
}
When an object is written, it is persisted at the version designated as the storage version at the time of the write. If the storage version changes, existing objects are never converted automatically. However, newly-created or updated objects are written at the new storage version. It is possible for an object to have been written at a version that is no longer served.
When you read an object, you specify the version as part of the path. If you
specify a version that is different from the object's persisted version,
Kubernetes returns the object to you at the version you requested, but the
persisted object is neither changed on disk, nor converted in any way
(other than changing the apiVersion string) while serving the request.
You can request an object at any version that is currently served.
If you update an existing object, it is rewritten at the version that is currently the storage version. This is the only way that objects can change from one version to another.
To illustrate this, consider the following hypothetical series of events:
v1beta1. You create an object. It is persisted in
storage at version v1beta1v1 to your CustomResourceDefinition and designate it as
the storage version.v1beta1, then you read the object again at
version v1. Both returned objects are identical except for the apiVersion
field.v1. You now
have two objects, one of which is at v1beta1, and the other of which is at
v1.v1 since that
is the current storage version.The API server records each version which has ever been marked as the storage
version in the status field storedVersions. Objects may have been persisted
at any version that has ever been designated as a storage version. No objects
can exist in storage at a version that has never been a storage version.
When deprecating versions and dropping support, select a storage upgrade procedure.
Option 1: Use the Storage Version Migrator
status.storedVersions field.Option 2: Manually upgrade the existing objects to a new stored version
The following is an example procedure to upgrade from v1beta1 to v1.
v1 as the storage in the CustomResourceDefinition file and apply it
using kubectl. The storedVersions is now v1beta1, v1.v1.v1beta1 from the CustomResourceDefinition status.storedVersions field.Setting up an extension API server to work with the aggregation layer allows the Kubernetes apiserver to be extended with additional APIs, which are not part of the core Kubernetes APIs.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
The following steps describe how to set up an extension-apiserver at a high level. These steps apply regardless if you're using YAML configs or using APIs. An attempt is made to specifically identify any differences between the two. For a concrete example of how they can be implemented using YAML configs, you can look at the sample-apiserver in the Kubernetes repo.
Alternatively, you can use an existing 3rd party solution, such as apiserver-builder, which should generate a skeleton and automate all of the following steps for you.
--runtime-config). It should be on by default, unless it's been deliberately turned off in your cluster.<service name>.<service name namespace>.svcsystem:auth-delegator cluster role to delegate auth decisions to the Kubernetes core API server.extension-apiserver-authentication-reader role. This allows your extension api-server to access the extension-apiserver-authentication configmap.Kubernetes ships with a default scheduler that is described here. If the default scheduler does not suit your needs you can implement your own scheduler. Moreover, you can even run multiple schedulers simultaneously alongside the default scheduler and instruct Kubernetes what scheduler to use for each of your pods. Let's learn how to run multiple schedulers in Kubernetes with an example.
A detailed description of how to implement a scheduler is outside the scope of this document. Please refer to the kube-scheduler implementation in pkg/scheduler in the Kubernetes source directory for a canonical example.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Package your scheduler binary into a container image. For the purposes of this example, you can use the default scheduler (kube-scheduler) as your second scheduler. Clone the Kubernetes source code from GitHub and build the source.
git clone https://github.com/kubernetes/kubernetes.git
cd kubernetes
make
Create a container image containing the kube-scheduler binary. Here is the Dockerfile
to build the image:
FROM busybox
ADD ./_output/local/bin/linux/amd64/kube-scheduler /usr/local/bin/kube-scheduler
Save the file as Dockerfile, build the image and push it to a registry. This example
pushes the image to
Google Container Registry (GCR).
For more details, please read the GCR
documentation.
docker build -t gcr.io/my-gcp-project/my-kube-scheduler:1.0 .
gcloud docker -- push gcr.io/my-gcp-project/my-kube-scheduler:1.0
Now that you have your scheduler in a container image, create a pod
configuration for it and run it in your Kubernetes cluster. But instead of creating a pod
directly in the cluster, you can use a Deployment
for this example. A Deployment manages a
Replica Set which in turn manages the pods,
thereby making the scheduler resilient to failures. Here is the deployment
config. Save it as my-scheduler.yaml:
apiVersion: v1
kind: ServiceAccount
metadata:
name: my-scheduler
namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: my-scheduler-as-kube-scheduler
subjects:
- kind: ServiceAccount
name: my-scheduler
namespace: kube-system
roleRef:
kind: ClusterRole
name: system:kube-scheduler
apiGroup: rbac.authorization.k8s.io
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: my-scheduler-as-volume-scheduler
subjects:
- kind: ServiceAccount
name: my-scheduler
namespace: kube-system
roleRef:
kind: ClusterRole
name: system:volume-scheduler
apiGroup: rbac.authorization.k8s.io
---
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
component: scheduler
tier: control-plane
name: my-scheduler
namespace: kube-system
spec:
selector:
matchLabels:
component: scheduler
tier: control-plane
replicas: 1
template:
metadata:
labels:
component: scheduler
tier: control-plane
version: second
spec:
serviceAccountName: my-scheduler
containers:
- command:
- /usr/local/bin/kube-scheduler
- --address=0.0.0.0
- --leader-elect=false
- --scheduler-name=my-scheduler
image: gcr.io/my-gcp-project/my-kube-scheduler:1.0
livenessProbe:
httpGet:
path: /healthz
port: 10251
initialDelaySeconds: 15
name: kube-second-scheduler
readinessProbe:
httpGet:
path: /healthz
port: 10251
resources:
requests:
cpu: '0.1'
securityContext:
privileged: false
volumeMounts: []
hostNetwork: false
hostPID: false
volumes: []
An important thing to note here is that the name of the scheduler specified as an
argument to the scheduler command in the container spec should be unique. This is the name that is matched against the value of the optional spec.schedulerName on pods, to determine whether this scheduler is responsible for scheduling a particular pod.
Note also that we created a dedicated service account my-scheduler and bind the cluster role
system:kube-scheduler to it so that it can acquire the same privileges as kube-scheduler.
Please see the kube-scheduler documentation for detailed description of other command line arguments.
In order to run your scheduler in a Kubernetes cluster, create the deployment specified in the config above in a Kubernetes cluster:
kubectl create -f my-scheduler.yaml
Verify that the scheduler pod is running:
kubectl get pods --namespace=kube-system
NAME READY STATUS RESTARTS AGE
....
my-scheduler-lnf4s-4744f 1/1 Running 0 2m
...
You should see a "Running" my-scheduler pod, in addition to the default kube-scheduler pod in this list.
To run multiple-scheduler with leader election enabled, you must do the following:
First, update the following fields in your YAML file:
--leader-elect=true--lock-object-namespace=<lock-object-namespace>--lock-object-name=<lock-object-name>kube-system namespace.
If RBAC is enabled on your cluster, you must update the system:kube-scheduler cluster role.
Add your scheduler name to the resourceNames of the rule applied for endpoints and leases resources, as in the following example:
kubectl edit clusterrole system:kube-scheduler
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
annotations:
rbac.authorization.kubernetes.io/autoupdate: "true"
labels:
kubernetes.io/bootstrapping: rbac-defaults
name: system:kube-scheduler
rules:
- apiGroups:
- coordination.k8s.io
resources:
- leases
verbs:
- create
- apiGroups:
- coordination.k8s.io
resourceNames:
- kube-scheduler
- my-scheduler
resources:
- leases
verbs:
- get
- update
- apiGroups:
- ""
resourceNames:
- kube-scheduler
- my-scheduler
resources:
- endpoints
verbs:
- delete
- get
- patch
- update
Now that your second scheduler is running, create some pods, and direct them to be scheduled by either the default scheduler or the one you deployed. In order to schedule a given pod using a specific scheduler, specify the name of the scheduler in that pod spec. Let's look at three examples.
Pod spec without any scheduler name
apiVersion: v1
kind: Pod
metadata:
name: no-annotation
labels:
name: multischeduler-example
spec:
containers:
- name: pod-with-no-annotation-container
image: k8s.gcr.io/pause:2.0When no scheduler name is supplied, the pod is automatically scheduled using the default-scheduler.
Save this file as pod1.yaml and submit it to the Kubernetes cluster.
kubectl create -f pod1.yaml
Pod spec with default-scheduler
apiVersion: v1
kind: Pod
metadata:
name: annotation-default-scheduler
labels:
name: multischeduler-example
spec:
schedulerName: default-scheduler
containers:
- name: pod-with-default-annotation-container
image: k8s.gcr.io/pause:2.0
A scheduler is specified by supplying the scheduler name as a value to spec.schedulerName. In this case, we supply the name of the
default scheduler which is default-scheduler.
Save this file as pod2.yaml and submit it to the Kubernetes cluster.
kubectl create -f pod2.yaml
Pod spec with my-scheduler
apiVersion: v1
kind: Pod
metadata:
name: annotation-second-scheduler
labels:
name: multischeduler-example
spec:
schedulerName: my-scheduler
containers:
- name: pod-with-second-annotation-container
image: k8s.gcr.io/pause:2.0
In this case, we specify that this pod should be scheduled using the scheduler that we
deployed - my-scheduler. Note that the value of spec.schedulerName should match the name supplied to the scheduler
command as an argument in the deployment config for the scheduler.
Save this file as pod3.yaml and submit it to the Kubernetes cluster.
kubectl create -f pod3.yaml
Verify that all three pods are running.
kubectl get pods
In order to make it easier to work through these examples, we did not verify that the
pods were actually scheduled using the desired schedulers. We can verify that by
changing the order of pod and deployment config submissions above. If we submit all the
pod configs to a Kubernetes cluster before submitting the scheduler deployment config,
we see that the pod annotation-second-scheduler remains in "Pending" state forever
while the other two pods get scheduled. Once we submit the scheduler deployment config
and our new scheduler starts running, the annotation-second-scheduler pod gets
scheduled as well.
Alternatively, you can look at the "Scheduled" entries in the event logs to verify that the pods were scheduled by the desired schedulers.
kubectl get events
You can also use a custom scheduler configuration or a custom container image for the cluster's main scheduler by modifying its static pod manifest on the relevant control plane nodes.
This page shows how to use an HTTP proxy to access the Kubernetes API.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
If you do not already have an application running in your cluster, start a Hello world application by entering this command:
kubectl create deployment node-hello --image=gcr.io/google-samples/node-hello:1.0 --port=8080
This command starts a proxy to the Kubernetes API server:
kubectl proxy --port=8080
When the proxy server is running, you can explore the API using curl, wget,
or a browser.
Get the API versions:
curl http://localhost:8080/api/
The output should look similar to this:
{
"kind": "APIVersions",
"versions": [
"v1"
],
"serverAddressByClientCIDRs": [
{
"clientCIDR": "0.0.0.0/0",
"serverAddress": "10.0.2.15:8443"
}
]
}
Get a list of pods:
curl http://localhost:8080/api/v1/namespaces/default/pods
The output should look similar to this:
{
"kind": "PodList",
"apiVersion": "v1",
"metadata": {
"resourceVersion": "33074"
},
"items": [
{
"metadata": {
"name": "kubernetes-bootcamp-2321272333-ix8pt",
"generateName": "kubernetes-bootcamp-2321272333-",
"namespace": "default",
"uid": "ba21457c-6b1d-11e6-85f7-1ef9f1dab92b",
"resourceVersion": "33003",
"creationTimestamp": "2016-08-25T23:43:30Z",
"labels": {
"pod-template-hash": "2321272333",
"run": "kubernetes-bootcamp"
},
...
}
Learn more about kubectl proxy.
The Konnectivity service provides a TCP level proxy for the control plane to cluster communication.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
The following steps require an egress configuration, for example:
apiVersion: apiserver.k8s.io/v1beta1
kind: EgressSelectorConfiguration
egressSelections:
# Since we want to control the egress traffic to the cluster, we use the
# "cluster" as the name. Other supported values are "etcd", and "master".
- name: cluster
connection:
# This controls the protocol between the API Server and the Konnectivity
# server. Supported values are "GRPC" and "HTTPConnect". There is no
# end user visible difference between the two modes. You need to set the
# Konnectivity server to work in the same mode.
proxyProtocol: GRPC
transport:
# This controls what transport the API Server uses to communicate with the
# Konnectivity server. UDS is recommended if the Konnectivity server
# locates on the same machine as the API Server. You need to configure the
# Konnectivity server to listen on the same UDS socket.
# The other supported transport is "tcp". You will need to set up TLS
# config to secure the TCP transport.
uds:
udsName: /etc/kubernetes/konnectivity-server/konnectivity-server.socket
You need to configure the API Server to use the Konnectivity service and direct the network traffic to the cluster nodes:
ServiceAccountTokenVolumeProjection feature gate
is enabled. You can enable
service account token volume protection
by providing the following flags to the kube-apiserver:
--service-account-issuer=api
--service-account-signing-key-file=/etc/kubernetes/pki/sa.key
--api-audiences=system:konnectivity-server
admin/konnectivity/egress-selector-configuration.yaml.--egress-selector-config-file flag of the API Server to the path of
your API Server egress configuration file.spec:
containers:
volumeMounts:
- name: konnectivity-uds
mountPath: /etc/kubernetes/konnectivity-server
readOnly: false
volumes:
- name: konnectivity-uds
hostPath:
path: /etc/kubernetes/konnectivity-server
type: DirectoryOrCreate
Generate or obtain a certificate and kubeconfig for konnectivity-server.
For example, you can use the OpenSSL command line tool to issue a X.509 certificate,
using the cluster CA certificate /etc/kubernetes/pki/ca.crt from a control-plane host.
openssl req -subj "/CN=system:konnectivity-server" -new -newkey rsa:2048 -nodes -out konnectivity.csr -keyout konnectivity.key -out konnectivity.csr
openssl x509 -req -in konnectivity.csr -CA /etc/kubernetes/pki/ca.crt -CAkey /etc/kubernetes/pki/ca.key -CAcreateserial -out konnectivity.crt -days 375 -sha256
SERVER=$(kubectl config view -o jsonpath='{.clusters..server}')
kubectl --kubeconfig /etc/kubernetes/konnectivity-server.conf config set-credentials system:konnectivity-server --client-certificate konnectivity.crt --client-key konnectivity.key --embed-certs=true
kubectl --kubeconfig /etc/kubernetes/konnectivity-server.conf config set-cluster kubernetes --server "$SERVER" --certificate-authority /etc/kubernetes/pki/ca.crt --embed-certs=true
kubectl --kubeconfig /etc/kubernetes/konnectivity-server.conf config set-context system:konnectivity-server@kubernetes --cluster kubernetes --user system:konnectivity-server
kubectl --kubeconfig /etc/kubernetes/konnectivity-server.conf config use-context system:konnectivity-server@kubernetes
rm -f konnectivity.crt konnectivity.key konnectivity.csr
Next, you need to deploy the Konnectivity server and agents. kubernetes-sigs/apiserver-network-proxy is a reference implementation.
Deploy the Konnectivity server on your control plane node. The provided
konnectivity-server.yaml manifest assumes
that the Kubernetes components are deployed as a static Pod in your cluster. If not, you can deploy the Konnectivity
server as a DaemonSet.
apiVersion: v1
kind: Pod
metadata:
name: konnectivity-server
namespace: kube-system
spec:
priorityClassName: system-cluster-critical
hostNetwork: true
containers:
- name: konnectivity-server-container
image: us.gcr.io/k8s-artifacts-prod/kas-network-proxy/proxy-server:v0.0.16
command: ["/proxy-server"]
args: [
"--logtostderr=true",
# This needs to be consistent with the value set in egressSelectorConfiguration.
"--uds-name=/etc/kubernetes/konnectivity-server/konnectivity-server.socket",
# The following two lines assume the Konnectivity server is
# deployed on the same machine as the apiserver, and the certs and
# key of the API Server are at the specified location.
"--cluster-cert=/etc/kubernetes/pki/apiserver.crt",
"--cluster-key=/etc/kubernetes/pki/apiserver.key",
# This needs to be consistent with the value set in egressSelectorConfiguration.
"--mode=grpc",
"--server-port=0",
"--agent-port=8132",
"--admin-port=8133",
"--health-port=8134",
"--agent-namespace=kube-system",
"--agent-service-account=konnectivity-agent",
"--kubeconfig=/etc/kubernetes/konnectivity-server.conf",
"--authentication-audience=system:konnectivity-server"
]
livenessProbe:
httpGet:
scheme: HTTP
host: 127.0.0.1
port: 8134
path: /healthz
initialDelaySeconds: 30
timeoutSeconds: 60
ports:
- name: agentport
containerPort: 8132
hostPort: 8132
- name: adminport
containerPort: 8133
hostPort: 8133
- name: healthport
containerPort: 8134
hostPort: 8134
volumeMounts:
- name: k8s-certs
mountPath: /etc/kubernetes/pki
readOnly: true
- name: kubeconfig
mountPath: /etc/kubernetes/konnectivity-server.conf
readOnly: true
- name: konnectivity-uds
mountPath: /etc/kubernetes/konnectivity-server
readOnly: false
volumes:
- name: k8s-certs
hostPath:
path: /etc/kubernetes/pki
- name: kubeconfig
hostPath:
path: /etc/kubernetes/konnectivity-server.conf
type: FileOrCreate
- name: konnectivity-uds
hostPath:
path: /etc/kubernetes/konnectivity-server
type: DirectoryOrCreate
Then deploy the Konnectivity agents in your cluster:
apiVersion: apps/v1
# Alternatively, you can deploy the agents as Deployments. It is not necessary
# to have an agent on each node.
kind: DaemonSet
metadata:
labels:
addonmanager.kubernetes.io/mode: Reconcile
k8s-app: konnectivity-agent
namespace: kube-system
name: konnectivity-agent
spec:
selector:
matchLabels:
k8s-app: konnectivity-agent
template:
metadata:
labels:
k8s-app: konnectivity-agent
spec:
priorityClassName: system-cluster-critical
tolerations:
- key: "CriticalAddonsOnly"
operator: "Exists"
containers:
- image: us.gcr.io/k8s-artifacts-prod/kas-network-proxy/proxy-agent:v0.0.16
name: konnectivity-agent
command: ["/proxy-agent"]
args: [
"--logtostderr=true",
"--ca-cert=/var/run/secrets/kubernetes.io/serviceaccount/ca.crt",
# Since the konnectivity server runs with hostNetwork=true,
# this is the IP address of the master machine.
"--proxy-server-host=35.225.206.7",
"--proxy-server-port=8132",
"--admin-server-port=8133",
"--health-server-port=8134",
"--service-account-token-path=/var/run/secrets/tokens/konnectivity-agent-token"
]
volumeMounts:
- mountPath: /var/run/secrets/tokens
name: konnectivity-agent-token
livenessProbe:
httpGet:
port: 8134
path: /healthz
initialDelaySeconds: 15
timeoutSeconds: 15
serviceAccountName: konnectivity-agent
volumes:
- name: konnectivity-agent-token
projected:
sources:
- serviceAccountToken:
path: konnectivity-agent-token
audience: system:konnectivity-server
Last, if RBAC is enabled in your cluster, create the relevant RBAC rules:
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: system:konnectivity-server
labels:
kubernetes.io/cluster-service: "true"
addonmanager.kubernetes.io/mode: Reconcile
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: system:auth-delegator
subjects:
- apiGroup: rbac.authorization.k8s.io
kind: User
name: system:konnectivity-server
---
apiVersion: v1
kind: ServiceAccount
metadata:
name: konnectivity-agent
namespace: kube-system
labels:
kubernetes.io/cluster-service: "true"
addonmanager.kubernetes.io/mode: Reconcile
This page shows how to enable and configure certificate rotation for the kubelet.
Kubernetes v1.19 [stable]
The kubelet uses certificates for authenticating to the Kubernetes API. By default, these certificates are issued with one year expiration so that they do not need to be renewed too frequently.
Kubernetes contains kubelet certificate rotation, that will automatically generate a new key and request a new certificate from the Kubernetes API as the current certificate approaches expiration. Once the new certificate is available, it will be used for authenticating connections to the Kubernetes API.
The kubelet process accepts an argument --rotate-certificates that controls
if the kubelet will automatically request a new certificate as the expiration of
the certificate currently in use approaches.
The kube-controller-manager process accepts an argument
--cluster-signing-duration (--experimental-cluster-signing-duration prior to 1.19)
that controls how long certificates will be issued for.
When a kubelet starts up, if it is configured to bootstrap (using the
--bootstrap-kubeconfig flag), it will use its initial certificate to connect
to the Kubernetes API and issue a certificate signing request. You can view the
status of certificate signing requests using:
kubectl get csr
Initially a certificate signing request from the kubelet on a node will have a
status of Pending. If the certificate signing requests meets specific
criteria, it will be auto approved by the controller manager, then it will have
a status of Approved. Next, the controller manager will sign a certificate,
issued for the duration specified by the
--cluster-signing-duration parameter, and the signed certificate
will be attached to the certificate signing request.
The kubelet will retrieve the signed certificate from the Kubernetes API and
write that to disk, in the location specified by --cert-dir. Then the kubelet
will use the new certificate to connect to the Kubernetes API.
As the expiration of the signed certificate approaches, the kubelet will automatically issue a new certificate signing request, using the Kubernetes API. This can happen at any point between 30% and 10% of the time remaining on the certificate. Again, the controller manager will automatically approve the certificate request and attach a signed certificate to the certificate signing request. The kubelet will retrieve the new signed certificate from the Kubernetes API and write that to disk. Then it will update the connections it has to the Kubernetes API to reconnect using the new certificate.
Kubernetes provides a certificates.k8s.io API, which lets you provision TLS
certificates signed by a Certificate Authority (CA) that you control. These CA
and certificates can be used by your workloads to establish trust.
certificates.k8s.io API uses a protocol that is similar to the ACME
draft.
certificates.k8s.io API are signed by a
dedicated CA. It is possible to configure your cluster to use the cluster root
CA for this purpose, but you should never rely on this. Do not assume that
these certificates will validate against the cluster root CA.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
To check the version, enterkubectl version.
Trusting the custom CA from an application running as a pod usually requires
some extra application configuration. You will need to add the CA certificate
bundle to the list of CA certificates that the TLS client or server trusts. For
example, you would do this with a golang TLS config by parsing the certificate
chain and adding the parsed certificates to the RootCAs field in the
tls.Config struct.
You can distribute the CA certificate as a ConfigMap that your pods have access to use.
The following section demonstrates how to create a TLS certificate for a Kubernetes service accessed through DNS.
The cfssl tools used in this example can be downloaded at https://github.com/cloudflare/cfssl/releases.
Generate a private key and certificate signing request (or CSR) by running the following command:
cat <<EOF | cfssl genkey - | cfssljson -bare server
{
"hosts": [
"my-svc.my-namespace.svc.cluster.local",
"my-pod.my-namespace.pod.cluster.local",
"192.0.2.24",
"10.0.34.2"
],
"CN": "system:node:my-pod.my-namespace.pod.cluster.local",
"key": {
"algo": "ecdsa",
"size": 256
},
"names": [
{
"O": "system:nodes"
}
]
}
EOF
Where 192.0.2.24 is the service's cluster IP,
my-svc.my-namespace.svc.cluster.local is the service's DNS name,
10.0.34.2 is the pod's IP and my-pod.my-namespace.pod.cluster.local
is the pod's DNS name. You should see the following output:
2017/03/21 06:48:17 [INFO] generate received request
2017/03/21 06:48:17 [INFO] received CSR
2017/03/21 06:48:17 [INFO] generating key: ecdsa-256
2017/03/21 06:48:17 [INFO] encoded CSR
This command generates two files; it generates server.csr containing the PEM
encoded pkcs#10 certification request,
and server-key.pem containing the PEM encoded key to the certificate that
is still to be created.
Generate a CSR yaml blob and send it to the apiserver by running the following command:
cat <<EOF | kubectl apply -f -
apiVersion: certificates.k8s.io/v1
kind: CertificateSigningRequest
metadata:
name: my-svc.my-namespace
spec:
request: $(cat server.csr | base64 | tr -d '\n')
signerName: kubernetes.io/kubelet-serving
usages:
- digital signature
- key encipherment
- server auth
EOF
Notice that the server.csr file created in step 1 is base64 encoded
and stashed in the .spec.request field. We are also requesting a
certificate with the "digital signature", "key encipherment", and "server
auth" key usages, signed by the kubernetes.io/kubelet-serving signer.
A specific signerName must be requested.
View documentation for supported signer names
for more information.
The CSR should now be visible from the API in a Pending state. You can see it by running:
kubectl describe csr my-svc.my-namespace
Name: my-svc.my-namespace
Labels: <none>
Annotations: <none>
CreationTimestamp: Tue, 21 Mar 2017 07:03:51 -0700
Requesting User: yourname@example.com
Status: Pending
Subject:
Common Name: my-svc.my-namespace.svc.cluster.local
Serial Number:
Subject Alternative Names:
DNS Names: my-svc.my-namespace.svc.cluster.local
IP Addresses: 192.0.2.24
10.0.34.2
Events: <none>
Approving the certificate signing request is either done by an automated approval process or on a one off basis by a cluster administrator. More information on what this involves is covered below.
Once the CSR is signed and approved you should see the following:
kubectl get csr
NAME AGE REQUESTOR CONDITION
my-svc.my-namespace 10m yourname@example.com Approved,Issued
You can download the issued certificate and save it to a server.crt file
by running the following:
kubectl get csr my-svc.my-namespace -o jsonpath='{.status.certificate}' \
| base64 --decode > server.crt
Now you can use server.crt and server-key.pem as the keypair to start
your HTTPS server.
A Kubernetes administrator (with appropriate permissions) can manually approve
(or deny) Certificate Signing Requests by using the kubectl certificate approve and kubectl certificate deny commands. However if you intend
to make heavy usage of this API, you might consider writing an automated
certificates controller.
Whether a machine or a human using kubectl as above, the role of the approver is to verify that the CSR satisfies two requirements:
If and only if these two requirements are met, the approver should approve the CSR and otherwise should deny the CSR.
The ability to approve CSRs decides who trusts whom within your environment. The ability to approve CSRs should not be granted broadly or lightly. The requirements of the challenge noted in the previous section and the repercussions of issuing a specific certificate should be fully understood before granting this permission.
This tutorial assumes that a signer is setup to serve the certificates API. The
Kubernetes controller manager provides a default implementation of a signer. To
enable it, pass the --cluster-signing-cert-file and
--cluster-signing-key-file parameters to the controller manager with paths to
your Certificate Authority's keypair.
This page shows how to manually rotate the certificate authority (CA) certificates.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version v1.13. To check the version, enterkubectl version.
Make sure to back up your certificate directory along with configuration files and any other necessary files.
This approach assumes operation of the Kubernetes control plane in a HA configuration with multiple API servers. Graceful termination of the API server is also assumed so clients can cleanly disconnect from one API server and reconnect to another.
Configurations with a single API server will experience unavailability while the API server is being restarted.
Distribute the new CA certificates and private keys
(ex: ca.crt, ca.key, front-proxy-ca.crt, and front-proxy-ca.key)
to all your control plane nodes in the Kubernetes certificates directory.
Update kube-controller-manager's --root-ca-file to
include both old and new CA. Then restart the component.
Any service account created after this point will get secrets that include both old and new CAs.
--client-ca-file and --cluster-signing-cert-file
cannot be CA bundles. If these flags and --root-ca-file point to the same ca.crt file which is now a
bundle (includes both old and new CA) you will face an error. To workaround this problem you can copy the new CA to a separate
file and make the flags --client-ca-file and --cluster-signing-cert-file point to the copy. Once ca.crt is no longer
a bundle you can restore the problem flags to point to ca.crt and delete the copy.
Update all service account tokens to include both old and new CA certificates.
If any pods are started before new CA is used by API servers, they will get this update and trust both old and new CAs.
base64_encoded_ca="$(base64 -w0 <path to file containing both old and new CAs>)"
for namespace in $(kubectl get ns --no-headers | awk '{print $1}'); do
for token in $(kubectl get secrets --namespace "$namespace" --field-selector type=kubernetes.io/service-account-token -o name); do
kubectl get $token --namespace "$namespace" -o yaml | \
/bin/sed "s/\(ca.crt:\).*/\1 ${base64_encoded_ca}/" | \
kubectl apply -f -
done
done
Restart all pods using in-cluster configs (ex: kube-proxy, coredns, etc) so they can use the updated certificate authority data from ServiceAccount secrets.
Append the both old and new CA to the file against --client-ca-file and --kubelet-certificate-authority flag in the kube-apiserver configuration.
Append the both old and new CA to the file against --client-ca-file flag in the kube-scheduler configuration.
Update certificates for user accounts by replacing the content of client-certificate-data and client-key-data respectively.
For information about creating certificates for individual user accounts, see Configure certificates for user accounts.
Additionally, update the certificate-authority-data section in the kubeconfig files,
respectively with Base64-encoded old and new certificate authority data
Follow below steps in a rolling fashion.
Restart any other aggregated api servers or webhook handlers to trust the new CA certificates.
Restart the kubelet by update the file against clientCAFile in kubelet configuration and
certificate-authority-data in kubelet.conf to use both the old and new CA on all nodes.
If your kubelet is not using client certificate rotation update client-certificate-data and
client-key-data in kubelet.conf on all nodes along with the kubelet client certificate file
usually found in /var/lib/kubelet/pki.
Restart API servers with the certificates (apiserver.crt, apiserver-kubelet-client.crt and
front-proxy-client.crt) signed by new CA.
You can use the existing private keys or new private keys.
If you changed the private keys then update these in the Kubernetes certificates directory as well.
Since the pod trusts both old and new CAs, there will be a momentarily disconnection after which the pod's kube client will reconnect to the new API server that uses the certificate signed by the new CA.
Restart Scheduler to use the new CAs.
Make sure control plane components logs no TLS errors.
openssl command line tool, see Certificates (openssl).
You can also use cfssl.
Annotate any Daemonsets and Deployments to trigger pod replacement in a safer rolling fashion.
Example:
for namespace in $(kubectl get namespace -o jsonpath='{.items[*].metadata.name}'); do
for name in $(kubectl get deployments -n $namespace -o jsonpath='{.items[*].metadata.name}'); do
kubectl patch deployment -n ${namespace} ${name} -p '{"spec":{"template":{"metadata":{"annotations":{"ca-rotation": "1"}}}}}';
done
for name in $(kubectl get daemonset -n $namespace -o jsonpath='{.items[*].metadata.name}'); do
kubectl patch daemonset -n ${namespace} ${name} -p '{"spec":{"template":{"metadata":{"annotations":{"ca-rotation": "1"}}}}}';
done
done
If your cluster is using bootstrap tokens to join nodes, update the ConfigMap cluster-info in the kube-public namespace with new CA.
base64_encoded_ca="$(base64 -w0 /etc/kubernetes/pki/ca.crt)"
kubectl get cm/cluster-info --namespace kube-public -o yaml | \
/bin/sed "s/\(certificate-authority-data:\).*/\1 ${base64_encoded_ca}/" | \
kubectl apply -f -
Verify the cluster functionality.
Validate the logs from control plane components, along with the kubelet and the kube-proxy are not throwing any tls errors, see looking at the logs.
Validate logs from any aggregated api servers and pods using in-cluster config.
Once the cluster functionality is successfully verified:
Update all service account tokens to include new CA certificate only.
Restart the control plane components by removing the old CA from the kubeconfig files and the files against --client-ca-file, --root-ca-file flags resp.
Restart kubelet by removing the old CA from file against the clientCAFile flag and kubelet kubeconfig file.
This page shows how to perform a rolling update on a DaemonSet.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
DaemonSet has two update strategy types:
OnDelete: With OnDelete update strategy, after you update a DaemonSet template, new
DaemonSet pods will only be created when you manually delete old DaemonSet
pods. This is the same behavior of DaemonSet in Kubernetes version 1.5 or
before.RollingUpdate: This is the default update strategy.RollingUpdate update strategy, after you update a
DaemonSet template, old DaemonSet pods will be killed, and new DaemonSet pods
will be created automatically, in a controlled fashion. At most one pod of
the DaemonSet will be running on each node during the whole update process.To enable the rolling update feature of a DaemonSet, you must set its
.spec.updateStrategy.type to RollingUpdate.
You may want to set
.spec.updateStrategy.rollingUpdate.maxUnavailable
(default to 1),
.spec.minReadySeconds
(default to 0) and
.spec.maxSurge
(a beta feature and defaults to 25%) as well.
RollingUpdate update strategyThis YAML file specifies a DaemonSet with an update strategy as 'RollingUpdate'
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: fluentd-elasticsearch
namespace: kube-system
labels:
k8s-app: fluentd-logging
spec:
selector:
matchLabels:
name: fluentd-elasticsearch
updateStrategy:
type: RollingUpdate
rollingUpdate:
maxUnavailable: 1
template:
metadata:
labels:
name: fluentd-elasticsearch
spec:
tolerations:
# this toleration is to have the daemonset runnable on master nodes
# remove it if your masters can't run pods
- key: node-role.kubernetes.io/master
effect: NoSchedule
containers:
- name: fluentd-elasticsearch
image: quay.io/fluentd_elasticsearch/fluentd:v2.5.2
volumeMounts:
- name: varlog
mountPath: /var/log
- name: varlibdockercontainers
mountPath: /var/lib/docker/containers
readOnly: true
terminationGracePeriodSeconds: 30
volumes:
- name: varlog
hostPath:
path: /var/log
- name: varlibdockercontainers
hostPath:
path: /var/lib/docker/containers
After verifying the update strategy of the DaemonSet manifest, create the DaemonSet:
kubectl create -f https://k8s.io/examples/controllers/fluentd-daemonset.yaml
Alternatively, use kubectl apply to create the same DaemonSet if you plan to
update the DaemonSet with kubectl apply.
kubectl apply -f https://k8s.io/examples/controllers/fluentd-daemonset.yaml
RollingUpdate update strategyCheck the update strategy of your DaemonSet, and make sure it's set to
RollingUpdate:
kubectl get ds/fluentd-elasticsearch -o go-template='{{.spec.updateStrategy.type}}{{"\n"}}' -n kube-system
If you haven't created the DaemonSet in the system, check your DaemonSet manifest with the following command instead:
kubectl apply -f https://k8s.io/examples/controllers/fluentd-daemonset.yaml --dry-run=client -o go-template='{{.spec.updateStrategy.type}}{{"\n"}}'
The output from both commands should be:
RollingUpdate
If the output isn't RollingUpdate, go back and modify the DaemonSet object or
manifest accordingly.
Any updates to a RollingUpdate DaemonSet .spec.template will trigger a rolling
update. Let's update the DaemonSet by applying a new YAML file. This can be done with several different kubectl commands.
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: fluentd-elasticsearch
namespace: kube-system
labels:
k8s-app: fluentd-logging
spec:
selector:
matchLabels:
name: fluentd-elasticsearch
updateStrategy:
type: RollingUpdate
rollingUpdate:
maxUnavailable: 1
template:
metadata:
labels:
name: fluentd-elasticsearch
spec:
tolerations:
# this toleration is to have the daemonset runnable on master nodes
# remove it if your masters can't run pods
- key: node-role.kubernetes.io/master
effect: NoSchedule
containers:
- name: fluentd-elasticsearch
image: quay.io/fluentd_elasticsearch/fluentd:v2.5.2
resources:
limits:
memory: 200Mi
requests:
cpu: 100m
memory: 200Mi
volumeMounts:
- name: varlog
mountPath: /var/log
- name: varlibdockercontainers
mountPath: /var/lib/docker/containers
readOnly: true
terminationGracePeriodSeconds: 30
volumes:
- name: varlog
hostPath:
path: /var/log
- name: varlibdockercontainers
hostPath:
path: /var/lib/docker/containers
If you update DaemonSets using
configuration files,
use kubectl apply:
kubectl apply -f https://k8s.io/examples/controllers/fluentd-daemonset-update.yaml
If you update DaemonSets using
imperative commands,
use kubectl edit :
kubectl edit ds/fluentd-elasticsearch -n kube-system
If you only need to update the container image in the DaemonSet template, i.e.
.spec.template.spec.containers[*].image, use kubectl set image:
kubectl set image ds/fluentd-elasticsearch fluentd-elasticsearch=quay.io/fluentd_elasticsearch/fluentd:v2.6.0 -n kube-system
Finally, watch the rollout status of the latest DaemonSet rolling update:
kubectl rollout status ds/fluentd-elasticsearch -n kube-system
When the rollout is complete, the output is similar to this:
daemonset "fluentd-elasticsearch" successfully rolled out
Sometimes, a DaemonSet rolling update may be stuck. Here are some possible causes:
The rollout is stuck because new DaemonSet pods can't be scheduled on at least one node. This is possible when the node is running out of resources.
When this happens, find the nodes that don't have the DaemonSet pods scheduled on
by comparing the output of kubectl get nodes and the output of:
kubectl get pods -l name=fluentd-elasticsearch -o wide -n kube-system
Once you've found those nodes, delete some non-DaemonSet pods from the node to make room for new DaemonSet pods.
If the recent DaemonSet template update is broken, for example, the container is crash looping, or the container image doesn't exist (often due to a typo), DaemonSet rollout won't progress.
To fix this, update the DaemonSet template again. New rollout won't be blocked by previous unhealthy rollouts.
If .spec.minReadySeconds is specified in the DaemonSet, clock skew between
master and nodes will make DaemonSet unable to detect the right rollout
progress.
Delete DaemonSet from a namespace :
kubectl delete ds fluentd-elasticsearch -n kube-system
This page shows how to perform a rollback on a DaemonSet.
You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:
Your Kubernetes server must be at or later than version 1.7. To check the version, enterkubectl version.
You should already know how to perform a rolling update on a DaemonSet.
You can skip this step if you only want to roll back to the last revision.
List all revisions of a DaemonSet:
kubectl rollout history daemonset <daemonset-name>
This returns a list of DaemonSet revisions:
daemonsets "<daemonset-name>"
REVISION CHANGE-CAUSE
1 ...
2 ...
...
kubernetes.io/change-cause
to its revisions upon creation. You may specify --record=true in kubectl
to record the command executed in the change cause annotation.To see the details of a specific revision:
kubectl rollout history daemonset <daemonset-name> --revision=1
This returns the details of that revision:
daemonsets "<daemonset-name>" with revision #1
Pod Template:
Labels: foo=bar
Containers:
app:
Image: ...
Port: ...
Environment: ...
Mounts: ...
Volumes: ...
# Specify the revision number you get from Step 1 in --to-revision
kubectl rollout undo daemonset <daemonset-name> --to-revision=<revision>
If it succeeds, the command returns:
daemonset "<daemonset-name>" rolled back
--to-revision flag is not specified, kubectl picks the most recent revision.
kubectl rollout undo daemonset tells the server to start rolling back the
DaemonSet. The real rollback is done asynchronously inside the cluster
control plane.
To watch the progress of the rollback:
kubectl rollout status ds/<daemonset-name>
When the rollback is complete, the output is similar to:
daemonset "<daemonset-name>" successfully rolled out
In the previous kubectl rollout history step, you got a list of DaemonSet
revisions. Each revision is stored in a resource named ControllerRevision.
To see what is stored in each revision, find the DaemonSet revision raw resources:
kubectl get controllerrevision -l <daemonset-selector-key>=<daemonset-selector-value>
This returns a list of ControllerRevisions:
NAME CONTROLLER REVISION AGE
<daemonset-name>-<revision-hash> DaemonSet/<daemonset-name> 1 1h
<daemonset-name>-<revision-hash> DaemonSet/<daemonset-name> 2 1h
Each ControllerRevision stores the annotations and template of a DaemonSet revision.
kubectl rollout undo takes a specific ControllerRevision and replaces
DaemonSet template with the template stored in the ControllerRevision.
kubectl rollout undo is equivalent to updating DaemonSet template to a
previous revision through other commands, such as kubectl edit or kubectl apply.
.revision field) of the
ControllerRevision being rolled back to will advance. For example, if you
have revision 1 and 2 in the system, and roll back from revision 2 to revision
1, the ControllerRevision with .revision: 1 will become .revision: 3.
Service Catalog is an extension API that enables applications running in Kubernetes clusters to easily use external managed software offerings, such as a datastore service offered by a cloud provider.
It provides a way to list, provision, and bind with external Managed Services from Service Brokers without needing detailed knowledge about how those services are created or managed.
Use Helm to install Service Catalog on your Kubernetes cluster. Up to date information on this process can be found at the kubernetes-sigs/service-catalog repo.
hack/local-up-cluster.sh, ensure that the KUBE_ENABLE_CLUSTER_DNS environment variable is set, then run the install script.helm init to install Tiller, the server-side component of Helm.Once Helm is installed, add the service-catalog Helm repository to your local machine by executing the following command:
helm repo add svc-cat https://kubernetes-sigs.github.io/service-catalog
Check to make sure that it installed successfully by executing the following command:
helm search repo service-catalog
If the installation was successful, the command should output the following:
NAME CHART VERSION APP VERSION DESCRIPTION
svc-cat/catalog 0.2.1 service-catalog API server and controller-manager helm chart
svc-cat/catalog-v0.2 0.2.2 service-catalog API server and controller-manager helm chart
Your Kubernetes cluster must have RBAC enabled, which requires your Tiller Pod(s) to have cluster-admin access.
When using Minikube v0.25 or older, you must run Minikube with RBAC explicitly enabled:
minikube start --extra-config=apiserver.Authorization.Mode=RBAC
When using Minikube v0.26+, run:
minikube start
With Minikube v0.26+, do not specify --extra-config. The flag has since been changed to --extra-config=apiserver.authorization-mode and Minikube now uses RBAC by default. Specifying the older flag may cause the start command to hang.
If you are using hack/local-up-cluster.sh, set the AUTHORIZATION_MODE environment variable with the following values:
AUTHORIZATION_MODE=Node,RBAC hack/local-up-cluster.sh -O
By default, helm init installs the Tiller Pod into the kube-system namespace, with Tiller configured to use the default service account.
--tiller-namespace or --service-account flags when running helm init, the --serviceaccount flag in the following command needs to be adjusted to reference the appropriate namespace and ServiceAccount name.
Configure Tiller to have cluster-admin access:
kubectl create clusterrolebinding tiller-cluster-admin \
--clusterrole=cluster-admin \
--serviceaccount=kube-system:default
Install Service Catalog from the root of the Helm repository using the following command:
helm install catalog svc-cat/catalog --namespace catalog
helm install svc-cat/catalog --name catalog --namespace catalog
Service Catalog is an extension API that enables applications running in Kubernetes clusters to easily use external managed software offerings, such as a datastore service offered by a cloud provider.
It provides a way to list, provision, and bind with external Managed Services from Service Brokers without needing detailed knowledge about how those services are created or managed.
You can use the GCP Service Catalog Installer tool to easily install or uninstall Service Catalog on your Kubernetes cluster, linking it to Google Cloud projects.
Service Catalog can work with any kind of managed service, not only Google Cloud.
Understand the key concepts of Service Catalog.
Install Go 1.6+ and set the GOPATH.
Install the cfssl tool needed for generating SSL artifacts.
Service Catalog requires Kubernetes version 1.7+.
Install and setup kubectl so that it is configured to connect to a Kubernetes v1.7+ cluster.
The kubectl user must be bound to the cluster-admin role for it to install Service Catalog. To ensure that this is true, run the following command:
kubectl create clusterrolebinding cluster-admin-binding --clusterrole=cluster-admin --user=<user-name>
sc in your local environmentThe installer runs on your local computer as a CLI tool named sc.
Install using go get:
go get github.com/GoogleCloudPlatform/k8s-service-catalog/installer/cmd/sc
sc should now be installed in your GOPATH/bin directory.
First, verify that all dependencies have been installed. Run:
sc check
If the check is successful, it should return:
Dependency check passed. You are good to go.
Next, run the install command and specify the storageclass that you want to use for the backup:
sc install --etcd-backup-storageclass "standard"
If you would like to uninstall Service Catalog from your Kubernetes cluster using the sc tool, run:
sc uninstall
Adding entries to a Pod's /etc/hosts file provides Pod-level override of hostname resolution when DNS and other options are not applicable. You can add these custom entries with the HostAliases field in PodSpec.
Modification not using HostAliases is not suggested because the file is managed by the kubelet and can be overwritten on during Pod creation/restart.
Start an Nginx Pod which is assigned a Pod IP:
kubectl run nginx --image nginx
pod/nginx created
Examine a Pod IP:
kubectl get pods --output=wide
NAME READY STATUS RESTARTS AGE IP NODE
nginx 1/1 Running 0 13s 10.200.0.4 worker0
The hosts file content would look like this:
kubectl exec nginx -- cat /etc/hosts
# Kubernetes-managed hosts file.
127.0.0.1 localhost
::1 localhost ip6-localhost ip6-loopback
fe00::0 ip6-localnet
fe00::0 ip6-mcastprefix
fe00::1 ip6-allnodes
fe00::2 ip6-allrouters
10.200.0.4 nginx
By default, the hosts file only includes IPv4 and IPv6 boilerplates like
localhost and its own hostname.
In addition to the default boilerplate, you can add additional entries to the
hosts file.
For example: to resolve foo.local, bar.local to 127.0.0.1 and foo.remote,
bar.remote to 10.1.2.3, you can configure HostAliases for a Pod under
.spec.hostAliases:
apiVersion: v1
kind: Pod
metadata:
name: hostaliases-pod
spec:
restartPolicy: Never
hostAliases:
- ip: "127.0.0.1"
hostnames:
- "foo.local"
- "bar.local"
- ip: "10.1.2.3"
hostnames:
- "foo.remote"
- "bar.remote"
containers:
- name: cat-hosts
image: busybox
command:
- cat
args:
- "/etc/hosts"
You can start a Pod with that configuration by running:
kubectl apply -f https://k8s.io/examples/service/networking/hostaliases-pod.yaml
pod/hostaliases-pod created
Examine a Pod's details to see its IPv4 address and its status:
kubectl get pod --output=wide
NAME READY STATUS RESTARTS AGE IP NODE
hostaliases-pod 0/1 Completed 0 6s 10.200.0.5 worker0
The hosts file content looks like this:
kubectl logs hostaliases-pod
# Kubernetes-managed hosts file.
127.0.0.1 localhost
::1 localhost ip6-localhost ip6-loopback
fe00::0 ip6-localnet
fe00::0 ip6-mcastprefix
fe00::1 ip6-allnodes
fe00::2 ip6-allrouters
10.200.0.5 hostaliases-pod
# Entries added by HostAliases.
127.0.0.1 foo.local bar.local
10.1.2.3 foo.remote bar.remote
with the additional entries specified at the bottom.
The kubelet manages the
hosts file for each container of the Pod to prevent Docker from
modifying the file after the
containers have already been started.
Avoid making manual changes to the hosts file inside a container.
If you make manual changes to the hosts file, those changes are lost when the container exits.
This document shares how to validate IPv4/IPv6 dual-stack enabled Kubernetes clusters.
kubectl version.
Each dual-stack Node should have a single IPv4 block and a single IPv6 block allocated. Validate that IPv4/IPv6 Pod address ranges are configured by running the following command. Replace the sample node name with a valid dual-stack Node from your cluster. In this example, the Node's name is k8s-linuxpool1-34450317-0:
kubectl get nodes k8s-linuxpool1-34450317-0 -o go-template --template='{{range .spec.podCIDRs}}{{printf "%s\n" .}}{{end}}'
10.244.1.0/24
a00:100::/24
There should be one IPv4 block and one IPv6 block allocated.
Validate that the node has an IPv4 and IPv6 interface detected. Replace node name with a valid node from the cluster. In this example the node name is k8s-linuxpool1-34450317-0:
kubectl get nodes k8s-linuxpool1-34450317-0 -o go-template --template='{{range .status.addresses}}{{printf "%s: %s\n" .type .address}}{{end}}'
Hostname: k8s-linuxpool1-34450317-0
InternalIP: 10.240.0.5
InternalIP: 2001:1234:5678:9abc::5
Validate that a Pod has an IPv4 and IPv6 address assigned. Replace the Pod name with a valid Pod in your cluster. In this example the Pod name is pod01:
kubectl get pods pod01 -o go-template --template='{{range .status.podIPs}}{{printf "%s\n" .ip}}{{end}}'
10.244.1.4
a00:100::4
You can also validate Pod IPs using the Downward API via the status.podIPs fieldPath. The following snippet demonstrates how you can expose the Pod IPs via an environment variable called MY_POD_IPS within a container.
env:
- name: MY_POD_IPS
valueFrom:
fieldRef:
fieldPath: status.podIPs
The following command prints the value of the MY_POD_IPS environment variable from within a container. The value is a comma separated list that corresponds to the Pod's IPv4 and IPv6 addresses.
kubectl exec -it pod01 -- set | grep MY_POD_IPS
MY_POD_IPS=10.244.1.4,a00:100::4
The Pod's IP addresses will also be written to /etc/hosts within a container. The following command executes a cat on /etc/hosts on a dual stack Pod. From the output you can verify both the IPv4 and IPv6 IP address for the Pod.
kubectl exec -it pod01 -- cat /etc/hosts
# Kubernetes-managed hosts file.
127.0.0.1 localhost
::1 localhost ip6-localhost ip6-loopback
fe00::0 ip6-localnet
fe00::0 ip6-mcastprefix
fe00::1 ip6-allnodes
fe00::2 ip6-allrouters
10.244.1.4 pod01
a00:100::4 pod01
Create the following Service that does not explicitly define .spec.ipFamilyPolicy. Kubernetes will assign a cluster IP for the Service from the first configured service-cluster-ip-range and set the .spec.ipFamilyPolicy to SingleStack.
apiVersion: v1
kind: Service
metadata:
name: my-service
labels:
app: MyApp
spec:
selector:
app: MyApp
ports:
- protocol: TCP
port: 80
Use kubectl to view the YAML for the Service.
kubectl get svc my-service -o yaml
The Service has .spec.ipFamilyPolicy set to SingleStack and .spec.clusterIP set to an IPv4 address from the first configured range set via --service-cluster-ip-range flag on kube-controller-manager.
apiVersion: v1
kind: Service
metadata:
name: my-service
namespace: default
spec:
clusterIP: 10.0.217.164
clusterIPs:
- 10.0.217.164
ipFamilies:
- IPv4
ipFamilyPolicy: SingleStack
ports:
- port: 80
protocol: TCP
targetPort: 9376
selector:
app: MyApp
sessionAffinity: None
type: ClusterIP
status:
loadBalancer: {}
Create the following Service that explicitly defines IPv6 as the first array element in .spec.ipFamilies. Kubernetes will assign a cluster IP for the Service from the IPv6 range configured service-cluster-ip-range and set the .spec.ipFamilyPolicy to SingleStack.
apiVersion: v1
kind: Service
metadata:
name: my-service
labels:
app: MyApp
spec:
ipFamilies:
- IPv6
selector:
app: MyApp
ports:
- protocol: TCP
port: 80
Use kubectl to view the YAML for the Service.
kubectl get svc my-service -o yaml
The Service has .spec.ipFamilyPolicy set to SingleStack and .spec.clusterIP set to an IPv6 address from the IPv6 range set via --service-cluster-ip-range flag on kube-controller-manager.
apiVersion: v1
kind: Service
metadata:
labels:
app: MyApp
name: my-service
spec:
clusterIP: fd00::5118
clusterIPs:
- fd00::5118
ipFamilies:
- IPv6
ipFamilyPolicy: SingleStack
ports:
- port: 80
protocol: TCP
targetPort: 80
selector:
app: MyApp
sessionAffinity: None
type: ClusterIP
status:
loadBalancer: {}
Create the following Service that explicitly defines PreferDualStack in .spec.ipFamilyPolicy. Kubernetes will assign both IPv4 and IPv6 addresses (as this cluster has dual-stack enabled) and select the .spec.ClusterIP from the list of .spec.ClusterIPs based on the address family of the first element in the .spec.ipFamilies array.
apiVersion: v1
kind: Service
metadata:
name: my-service
labels:
app: MyApp
spec:
ipFamilyPolicy: PreferDualStack
selector:
app: MyApp
ports:
- protocol: TCP
port: 80
The kubectl get svc command will only show the primary IP in the CLUSTER-IP field.
kubectl get svc -l app=MyApp
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-service ClusterIP 10.0.216.242 <none> 80/TCP 5s
Validate that the Service gets cluster IPs from the IPv4 and IPv6 address blocks using kubectl describe. You may then validate access to the service via the IPs and ports.
kubectl describe svc -l app=MyApp
Name: my-service
Namespace: default
Labels: app=MyApp
Annotations: <none>
Selector: app=MyApp
Type: ClusterIP
IP Family Policy: PreferDualStack
IP Families: IPv4,IPv6
IP: 10.0.216.242
IPs: 10.0.216.242,fd00::af55
Port: <unset> 80/TCP
TargetPort: 9376/TCP
Endpoints: <none>
Session Affinity: None
Events: <none>
If the cloud provider supports the provisioning of IPv6 enabled external load balancers, create the following Service with PreferDualStack in .spec.ipFamilyPolicy, IPv6 as the first element of the .spec.ipFamilies array and the type field set to LoadBalancer.
apiVersion: v1
kind: Service
metadata:
name: my-service
labels:
app: MyApp
spec:
ipFamilyPolicy: PreferDualStack
ipFamilies:
- IPv6
type: LoadBalancer
selector:
app: MyApp
ports:
- protocol: TCP
port: 80
Check the Service:
kubectl get svc -l app=MyApp
Validate that the Service receives a CLUSTER-IP address from the IPv6 address block along with an EXTERNAL-IP. You may then validate access to the service via the IP and port.
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-service LoadBalancer fd00::7ebc 2603:1030:805::5 80:30790/TCP 35s
Kubernetes v1.20 [alpha]
Starting from Kubernetes v1.20, the kubelet can dynamically retrieve credentials for a container image registry using exec plugins. The kubelet and the exec plugin communicate through stdio (stdin, stdout, and stderr) using Kubernetes versioned APIs. These plugins allow the kubelet to request credentials for a container registry dynamically as opposed to storing static credentials on disk. For example, the plugin may talk to a local metadata server to retrieve short-lived credentials for an image that is being pulled by the kubelet.
You may be interested in using this capability if any of the below are true:
This guide demonstrates how to configure the kubelet's image credential provider plugin mechanism.
KubeletCredentialProviders must be enabled on only the kubelet for the feature to work.A credential provider plugin is an executable binary that will be run by the kubelet. Ensure that the plugin binary exists on every node in your cluster and stored in a known directory. The directory will be required later when configuring kubelet flags.
In order to use this feature, the kubelet expects two flags to be set:
--image-credential-provider-config - the path to the credential provider plugin config file.--image-credential-provider-bin-dir - the path to the directory where credential provider plugin binaries are located.The configuration file passed into --image-credential-provider-config is read by the kubelet to determine which exec plugins
should be invoked for which container images. Here's an example configuration file you may end up using if you are using the ECR-based plugin:
kind: CredentialProviderConfig
apiVersion: kubelet.config.k8s.io/v1alpha1
# providers is a list of credential provider plugins that will be enabled by the kubelet.
# Multiple providers may match against a single image, in which case credentials
# from all providers will be returned to the kubelet. If multiple providers are called
# for a single image, the results are combined. If providers return overlapping
# auth keys, the value from the provider earlier in this list is used.
providers:
# name is the required name of the credential provider. It must match the name of the
# provider executable as seen by the kubelet. The executable must be in the kubelet's
# bin directory (set by the --image-credential-provider-bin-dir flag).
- name: ecr
# matchImages is a required list of strings used to match against images in order to
# determine if this provider should be invoked. If one of the strings matches the
# requested image from the kubelet, the plugin will be invoked and given a chance
# to provide credentials. Images are expected to contain the registry domain
# and URL path.
#
# Each entry in matchImages is a pattern which can optionally contain a port and a path.
# Globs can be used in the domain, but not in the port or the path. Globs are supported
# as subdomains like '*.k8s.io' or 'k8s.*.io', and top-level-domains such as 'k8s.*'.
# Matching partial subdomains like 'app*.k8s.io' is also supported. Each glob can only match
# a single subdomain segment, so *.io does not match *.k8s.io.
#
# A match exists between an image and a matchImage when all of the below are true:
# - Both contain the same number of domain parts and each part matches.
# - The URL path of an imageMatch must be a prefix of the target image URL path.
# - If the imageMatch contains a port, then the port must match in the image as well.
#
# Example values of matchImages:
# - 123456789.dkr.ecr.us-east-1.amazonaws.com
# - *.azurecr.io
# - gcr.io
# - *.*.registry.io
# - registry.io:8080/path
matchImages:
- "*.dkr.ecr.*.amazonaws.com"
- "*.dkr.ecr.*.amazonaws.cn"
- "*.dkr.ecr-fips.*.amazonaws.com"
- "*.dkr.ecr.us-iso-east-1.c2s.ic.gov"
- "*.dkr.ecr.us-isob-east-1.sc2s.sgov.gov"
# defaultCacheDuration is the default duration the plugin will cache credentials in-memory
# if a cache duration is not provided in the plugin response. This field is required.
defaultCacheDuration: "12h"
# Required input version of the exec CredentialProviderRequest. The returned CredentialProviderResponse
# MUST use the same encoding version as the input. Current supported values are:
# - credentialprovider.kubelet.k8s.io/v1alpha1
apiVersion: credentialprovider.kubelet.k8s.io/v1alpha1
# Arguments to pass to the command when executing it.
# +optional
args:
- get-credentials
# Env defines additional environment variables to expose to the process. These
# are unioned with the host's environment, as well as variables client-go uses
# to pass argument to the plugin.
# +optional
env:
- name: AWS_PROFILE
value: example_profile
The providers field is a list of enabled plugins used by the kubelet. Each entry has a few required fields:
name: the name of the plugin which MUST match the name of the executable binary that exists in the directory passed into --image-credential-provider-bin-dir.matchImages: a list of strings used to match against images in order to determine if this provider should be invoked. More on this below.defaultCacheDuration: the default duration the kubelet will cache credentials in-memory if a cache duration was not specified by the plugin.apiVersion: the api version that the kubelet and the exec plugin will use when communicating.Each credential provider can also be given optional args and environment variables as well. Consult the plugin implementors to determine what set of arguments and environment variables are required for a given plugin.
The matchImages field for each credential provider is used by the kubelet to determine whether a plugin should be invoked
for a given image that a Pod is using. Each entry in matchImages is an image pattern which can optionally contain a port and a path.
Globs can be used in the domain, but not in the port or the path. Globs are supported as subdomains like *.k8s.io or k8s.*.io,
and top-level domains such as k8s.*. Matching partial subdomains like app*.k8s.io is also supported. Each glob can only match
a single subdomain segment, so *.io does NOT match *.k8s.io.
A match exists between an image name and a matchImage entry when all of the below are true:
Some example values of matchImages patterns are:
123456789.dkr.ecr.us-east-1.amazonaws.com*.azurecr.iogcr.io*.*.registry.iofoo.registry.io:8080/pathThis guide demonstrates how to install and write extensions for kubectl. By thinking of core kubectl commands as essential building blocks for interacting with a Kubernetes cluster, a cluster administrator can think
of plugins as a means of utilizing these building blocks to create more complex behavior. Plugins extend kubectl with new sub-commands, allowing for new and custom features not included in the main distribution of kubectl.
You need to have a working kubectl binary installed.
A plugin is a standalone executable file, whose name begins with kubectl-. To install a plugin, move its executable file to anywhere on your PATH.
You can also discover and install kubectl plugins available in the open source using Krew. Krew is a plugin manager maintained by the Kubernetes SIG CLI community.
kubectl provides a command kubectl plugin list that searches your PATH for valid plugin executables.
Executing this command causes a traversal of all files in your PATH. Any files that are executable, and begin with kubectl- will show up in the order in which they are present in your PATH in this command's output.
A warning will be included for any files beginning with kubectl- that are not executable.
A warning will also be included for any valid plugin files that overlap each other's name.
You can use Krew to discover and install kubectl
plugins from a community-curated
plugin index.
It is currently not possible to create plugins that overwrite existing kubectl commands. For example, creating a plugin kubectl-version will cause that plugin to never be executed, as the existing kubectl version command will always take precedence over it. Due to this limitation, it is also not possible to use plugins to add new subcommands to existing kubectl commands. For example, adding a subcommand kubectl create foo by naming your plugin kubectl-create-foo will cause that plugin to be ignored.
kubectl plugin list shows warnings for any valid plugins that attempt to do this.
You can write a plugin in any programming language or script that allows you to write command-line commands.
There is no plugin installation or pre-loading required. Plugin executables receive
the inherited environment from the kubectl binary.
A plugin determines which command path it wishes to implement based on its name.
For example, a plugin named kubectl-foo provides a command kubectl foo. You must
install the plugin executable somewhere in your PATH.
#!/bin/bash
# optional argument handling
if [[ "$1" == "version" ]]
then
echo "1.0.0"
exit 0
fi
# optional argument handling
if [[ "$1" == "config" ]]
then
echo "$KUBECONFIG"
exit 0
fi
echo "I am a plugin named kubectl-foo"
To use a plugin, make the plugin executable:
sudo chmod +x ./kubectl-foo
and place it anywhere in your PATH:
sudo mv ./kubectl-foo /usr/local/bin
You may now invoke your plugin as a kubectl command:
kubectl foo
I am a plugin named kubectl-foo
All args and flags are passed as-is to the executable:
kubectl foo version
1.0.0
All environment variables are also passed as-is to the executable:
export KUBECONFIG=~/.kube/config
kubectl foo config
/home/<user>/.kube/config
KUBECONFIG=/etc/kube/config kubectl foo config
/etc/kube/config
Additionally, the first argument that is passed to a plugin will always be the full path to the location where it was invoked ($0 would equal /usr/local/bin/kubectl-foo in the example above).
As seen in the example above, a plugin determines the command path that it will implement based on its filename. Every sub-command in the command path that a plugin targets, is separated by a dash (-).
For example, a plugin that wishes to be invoked whenever the command kubectl foo bar baz is invoked by the user, would have the filename of kubectl-foo-bar-baz.
The plugin mechanism does not create any custom, plugin-specific values or environment variables for a plugin process.
An older kubectl plugin mechanism provided environment variables such as KUBECTL_PLUGINS_CURRENT_NAMESPACE; that no longer happens.
kubectl plugins must parse and validate all of the arguments passed to them. See using the command line runtime package for details of a Go library aimed at plugin authors.
Here are some additional cases where users invoke your plugin while providing additional flags and arguments. This builds upon the kubectl-foo-bar-baz plugin from the scenario above.
If you run kubectl foo bar baz arg1 --flag=value arg2, kubectl's plugin mechanism will first try to find the plugin with the longest possible name, which in this case
would be kubectl-foo-bar-baz-arg1. Upon not finding that plugin, kubectl then treats the last dash-separated value as an argument (arg1 in this case), and attempts to find the next longest possible name, kubectl-foo-bar-baz.
Upon having found a plugin with this name, kubectl then invokes that plugin, passing all args and flags after the plugin's name as arguments to the plugin process.
Example:
# create a plugin
echo -e '#!/bin/bash\n\necho "My first command-line argument was $1"' > kubectl-foo-bar-baz
sudo chmod +x ./kubectl-foo-bar-baz
# "install" your plugin by moving it to a directory in your $PATH
sudo mv ./kubectl-foo-bar-baz /usr/local/bin
# check that kubectl recognizes your plugin
kubectl plugin list
The following kubectl-compatible plugins are available:
/usr/local/bin/kubectl-foo-bar-baz
# test that calling your plugin via a "kubectl" command works
# even when additional arguments and flags are passed to your
# plugin executable by the user.
kubectl foo bar baz arg1 --meaningless-flag=true
My first command-line argument was arg1
As you can see, your plugin was found based on the kubectl command specified by a user, and all extra arguments and flags were passed as-is to the plugin executable once it was found.
Although the kubectl plugin mechanism uses the dash (-) in plugin filenames to separate the sequence of sub-commands processed by the plugin, it is still possible to create a plugin
command containing dashes in its commandline invocation by using underscores (_) in its filename.
Example:
# create a plugin containing an underscore in its filename
echo -e '#!/bin/bash\n\necho "I am a plugin with a dash in my name"' > ./kubectl-foo_bar
sudo chmod +x ./kubectl-foo_bar
# move the plugin into your $PATH
sudo mv ./kubectl-foo_bar /usr/local/bin
# You can now invoke your plugin via kubectl:
kubectl foo-bar
I am a plugin with a dash in my name
Note that the introduction of underscores to a plugin filename does not prevent you from having commands such as kubectl foo_bar.
The command from the above example, can be invoked using either a dash (-) or an underscore (_):
# You can invoke your custom command with a dash
kubectl foo-bar
I am a plugin with a dash in my name
# You can also invoke your custom command with an underscore
kubectl foo_bar
I am a plugin with a dash in my name
It is possible to have multiple plugins with the same filename in different locations throughout your PATH.
For example, given a PATH with the following value: PATH=/usr/local/bin/plugins:/usr/local/bin/moreplugins, a copy of plugin kubectl-foo could exist in /usr/local/bin/plugins and /usr/local/bin/moreplugins,
such that the output of the kubectl plugin list command is:
PATH=/usr/local/bin/plugins:/usr/local/bin/moreplugins kubectl plugin list
The following kubectl-compatible plugins are available:
/usr/local/bin/plugins/kubectl-foo
/usr/local/bin/moreplugins/kubectl-foo
- warning: /usr/local/bin/moreplugins/kubectl-foo is overshadowed by a similarly named plugin: /usr/local/bin/plugins/kubectl-foo
error: one plugin warning was found
In the above scenario, the warning under /usr/local/bin/moreplugins/kubectl-foo tells you that this plugin will never be executed. Instead, the executable that appears first in your PATH, /usr/local/bin/plugins/kubectl-foo, will always be found and executed first by the kubectl plugin mechanism.
A way to resolve this issue is to ensure that the location of the plugin that you wish to use with kubectl always comes first in your PATH. For example, if you want to always use /usr/local/bin/moreplugins/kubectl-foo anytime that the kubectl command kubectl foo was invoked, change the value of your PATH to be /usr/local/bin/moreplugins:/usr/local/bin/plugins.
There is another kind of overshadowing that can occur with plugin filenames. Given two plugins present in a user's PATH: kubectl-foo-bar and kubectl-foo-bar-baz, the kubectl plugin mechanism will always choose the longest possible plugin name for a given user command. Some examples below, clarify this further:
# for a given kubectl command, the plugin with the longest possible filename will always be preferred
kubectl foo bar baz
Plugin kubectl-foo-bar-baz is executed
kubectl foo bar
Plugin kubectl-foo-bar is executed
kubectl foo bar baz buz
Plugin kubectl-foo-bar-baz is executed, with "buz" as its first argument
kubectl foo bar buz
Plugin kubectl-foo-bar is executed, with "buz" as its first argument
This design choice ensures that plugin sub-commands can be implemented across multiple files, if needed, and that these sub-commands can be nested under a "parent" plugin command:
ls ./plugin_command_tree
kubectl-parent
kubectl-parent-subcommand
kubectl-parent-subcommand-subsubcommand
You can use the aforementioned kubectl plugin list command to ensure that your plugin is visible by kubectl, and verify that there are no warnings preventing it from being called as a kubectl command.
kubectl plugin list
The following kubectl-compatible plugins are available:
test/fixtures/pkg/kubectl/plugins/kubectl-foo
/usr/local/bin/kubectl-foo
- warning: /usr/local/bin/kubectl-foo is overshadowed by a similarly named plugin: test/fixtures/pkg/kubectl/plugins/kubectl-foo
plugins/kubectl-invalid
- warning: plugins/kubectl-invalid identified as a kubectl plugin, but it is not executable
error: 2 plugin warnings were found
If you're writing a plugin for kubectl and you're using Go, you can make use of the cli-runtime utility libraries.
These libraries provide helpers for parsing or updating a user's kubeconfig file, for making REST-style requests to the API server, or to bind flags associated with configuration and printing.
See the Sample CLI Plugin for an example usage of the tools provided in the CLI Runtime repo.
If you have developed a plugin for others to use, you should consider how you package it, distribute it and deliver updates to your users.
Krew offers a cross-platform way to package and distribute your plugins. This way, you use a single packaging format for all target platforms (Linux, Windows, macOS etc) and deliver updates to your users. Krew also maintains a plugin index so that other people can discover your plugin and install it.
Alternatively, you can use traditional package managers such as, apt or yum
on Linux, Chocolatey on Windows, and Homebrew on macOS. Any package
manager will be suitable if it can place new executables placed somewhere
in the user's PATH.
As a plugin author, if you pick this option then you also have the burden
of updating your kubectl plugin's distribution package across multiple
platforms for each release.
You can publish the source code; for example, as a Git repository. If you choose this option, someone who wants to use that plugin must fetch the code, set up a build environment (if it needs compiling), and deploy the plugin. If you also make compiled packages available, or use Krew, that will make installs easier.
Kubernetes v1.22 [stable]
Kubernetes supports the allocation and consumption of pre-allocated huge pages by applications in a Pod. This page describes how users can consume huge pages.
The nodes will automatically discover and report all huge page resources as schedulable resources.
Huge pages can be consumed via container level resource requirements using the
resource name hugepages-<size>, where <size> is the most compact binary
notation using integer values supported on a particular node. For example, if a
node supports 2048KiB and 1048576KiB page sizes, it will expose a schedulable
resources hugepages-2Mi and hugepages-1Gi. Unlike CPU or memory, huge pages
do not support overcommit. Note that when requesting hugepage resources, either
memory or CPU resources must be requested as well.
A pod may consume multiple huge page sizes in a single pod spec. In this case it
must use medium: HugePages-<hugepagesize> notation for all volume mounts.
apiVersion: v1
kind: Pod
metadata:
name: huge-pages-example
spec:
containers:
- name: example
image: fedora:latest
command:
- sleep
- inf
volumeMounts:
- mountPath: /hugepages-2Mi
name: hugepage-2mi
- mountPath: /hugepages-1Gi
name: hugepage-1gi
resources:
limits:
hugepages-2Mi: 100Mi
hugepages-1Gi: 2Gi
memory: 100Mi
requests:
memory: 100Mi
volumes:
- name: hugepage-2mi
emptyDir:
medium: HugePages-2Mi
- name: hugepage-1gi
emptyDir:
medium: HugePages-1Gi
A pod may use medium: HugePages only if it requests huge pages of one size.
apiVersion: v1
kind: Pod
metadata:
name: huge-pages-example
spec:
containers:
- name: example
image: fedora:latest
command:
- sleep
- inf
volumeMounts:
- mountPath: /hugepages
name: hugepage
resources:
limits:
hugepages-2Mi: 100Mi
memory: 100Mi
requests:
memory: 100Mi
volumes:
- name: hugepage
emptyDir:
medium: HugePages
shmget() with SHM_HUGETLB must
run with a supplemental group that matches proc/sys/vm/hugetlb_shm_group.cpu or memory using the hugepages-<size>
token.Kubernetes v1.10 [beta]
Kubernetes includes experimental support for managing AMD and NVIDIA GPUs (graphical processing units) across several nodes.
This page describes how users can consume GPUs across different Kubernetes versions and the current limitations.
Kubernetes implements Device Plugins to let Pods access specialized hardware features such as GPUs.
As an administrator, you have to install GPU drivers from the corresponding hardware vendor on the nodes and run the corresponding device plugin from the GPU vendor:
When the above conditions are true, Kubernetes will expose amd.com/gpu or
nvidia.com/gpu as a schedulable resource.
You can consume these GPUs from your containers by requesting
<vendor>.com/gpu the same way you request cpu or memory.
However, there are some limitations in how you specify the resource requirements
when using GPUs:
limits section, which means:
limits without specifying requests because
Kubernetes will use the limit as the request value by default.limits and requests but these two values
must be equal.requests without specifying limits.Here's an example:
apiVersion: v1
kind: Pod
metadata:
name: cuda-vector-add
spec:
restartPolicy: OnFailure
containers:
- name: cuda-vector-add
# https://github.com/kubernetes/kubernetes/blob/v1.7.11/test/images/nvidia-cuda/Dockerfile
image: "k8s.gcr.io/cuda-vector-add:v0.1"
resources:
limits:
nvidia.com/gpu: 1 # requesting 1 GPU
The official AMD GPU device plugin has the following requirements:
To deploy the AMD device plugin once your cluster is running and the above requirements are satisfied:
kubectl create -f https://raw.githubusercontent.com/RadeonOpenCompute/k8s-device-plugin/v1.10/k8s-ds-amdgpu-dp.yaml
You can report issues with this third-party device plugin by logging an issue in RadeonOpenCompute/k8s-device-plugin.
There are currently two device plugin implementations for NVIDIA GPUs:
The official NVIDIA GPU device plugin has the following requirements:
nvidia-container-runtime must be configured as the default runtime
for Docker, instead of runc.To deploy the NVIDIA device plugin once your cluster is running and the above requirements are satisfied:
kubectl create -f https://raw.githubusercontent.com/NVIDIA/k8s-device-plugin/1.0.0-beta4/nvidia-device-plugin.yml
You can report issues with this third-party device plugin by logging an issue in NVIDIA/k8s-device-plugin.
The NVIDIA GPU device plugin used by GCE doesn't require using nvidia-docker and should work with any container runtime that is compatible with the Kubernetes Container Runtime Interface (CRI). It's tested on Container-Optimized OS and has experimental code for Ubuntu from 1.9 onwards.
You can use the following commands to install the NVIDIA drivers and device plugin:
# Install NVIDIA drivers on Container-Optimized OS:
kubectl create -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/stable/daemonset.yaml
# Install NVIDIA drivers on Ubuntu (experimental):
kubectl create -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/stable/nvidia-driver-installer/ubuntu/daemonset.yaml
# Install the device plugin:
kubectl create -f https://raw.githubusercontent.com/kubernetes/kubernetes/release-1.14/cluster/addons/device-plugins/nvidia-gpu/daemonset.yaml
You can report issues with using or deploying this third-party device plugin by logging an issue in GoogleCloudPlatform/container-engine-accelerators.
Google publishes its own instructions for using NVIDIA GPUs on GKE .
If different nodes in your cluster have different types of GPUs, then you can use Node Labels and Node Selectors to schedule pods to appropriate nodes.
For example:
# Label your nodes with the accelerator type they have.
kubectl label nodes <node-with-k80> accelerator=nvidia-tesla-k80
kubectl label nodes <node-with-p100> accelerator=nvidia-tesla-p100
If you're using AMD GPU devices, you can deploy Node Labeller. Node Labeller is a controller that automatically labels your nodes with GPU device properties.
At the moment, that controller can add labels for:
kubectl describe node cluster-node-23
Name: cluster-node-23
Roles: <none>
Labels: beta.amd.com/gpu.cu-count.64=1
beta.amd.com/gpu.device-id.6860=1
beta.amd.com/gpu.family.AI=1
beta.amd.com/gpu.simd-count.256=1
beta.amd.com/gpu.vram.16G=1
beta.kubernetes.io/arch=amd64
beta.kubernetes.io/os=linux
kubernetes.io/hostname=cluster-node-23
Annotations: kubeadm.alpha.kubernetes.io/cri-socket: /var/run/dockershim.sock
node.alpha.kubernetes.io/ttl: 0
…
With the Node Labeller in use, you can specify the GPU type in the Pod spec:
apiVersion: v1
kind: Pod
metadata:
name: cuda-vector-add
spec:
restartPolicy: OnFailure
containers:
- name: cuda-vector-add
# https://github.com/kubernetes/kubernetes/blob/v1.7.11/test/images/nvidia-cuda/Dockerfile
image: "k8s.gcr.io/cuda-vector-add:v0.1"
resources:
limits:
nvidia.com/gpu: 1
nodeSelector:
accelerator: nvidia-tesla-p100 # or nvidia-tesla-k80 etc.
This will ensure that the Pod will be scheduled to a node that has the GPU type you specified.