Using multiple config files with kubectl and other Kubernetes tools

If you want to point tools like kubectl to a config file other than ~/.kube/config you can set the environment variable KUBECONFIG. But did you know that KUBECONFIG behaves sort of like a path, and kubectl will load all the config files it finds?

Let’s say my ~/.kube/config file has config for one cluster.

$ kubectl config get-contexts
CURRENT   NAME                 CLUSTER              AUTHINFO             NAMESPACE
*         my-awesome-cluster   my-awesome-cluster   my-awesome-cluster

If I am provided with another config file for another cluster it might not be clear where to put it. I could just put it in my ~/.kube/ directory and call it ~/.kube/other.config and point my KUBECONFIG to it.

$ KUBECONFIG=~/.kube/other.config kubectl config get-contexts
CURRENT   NAME                 CLUSTER              AUTHINFO             NAMESPACE
*         some-other-cluster   some-other-cluster   some-other-cluster

I could merge it with my existing config. But what if this new config is going to change from time to time and I may need to redownload it occasionally. I don’t want to have to pull apart my config or delete all the entries from it just to merge a new one in.

I can also update my KUBECONFIG environment variable to be a colon : separated list and then kubectl will load both.

$ KUBECONFIG=~/.kube/config:~/.kube/other.config kubectl config get-contexts
CURRENT   NAME                 CLUSTER              AUTHINFO             NAMESPACE
*         my-awesome-cluster   my-awesome-cluster   my-awesome-cluster   
          some-other-cluster   some-other-cluster   some-other-cluster

If you have lots of config files that you use regularly you could update your .bashrc/.zshrc to do this all the time.

# .bashrc
# ...

# Explicitly set the default location
export KUBECONFIG="$HOME/.kube/config"

# Add more configs in a way that feels like PATH=/foo/bin:$PATH
export KUBECONFIG="$KUBECONFIG:$HOME/.kube/other.config"

When specifying multiple configs any write operations like kubectl config use-context foo will be applied to the first config in the list. So it’s best that this stays set to ~/.kube/config unless you explicitly want to modify a different config.

For more information see the Kubernetes documentation.

I love hearing feedback on my posts. You should head over to Bluesky and let me know what you think!

Using multiple config files with kubectl and other Kubernetes tools

Have thoughts?