Kubernetes CLI: kubectl #

The CLI which we use to interact with Kubernetes is called kubectl. How to pronounce the name of the CLI is splitting waters. Some of the most common ones are

• kubeCTL
• kubecuddle
• kubecontrol

I don’t care what you call it, personally I usually go with “kubeCTL”.

Kube Config #

The configuration file for the kubectl CLI is commonly reffered to as “kube config”. By default the file is located at ~/.kube/config.

If you at some point are provided with a “kube config” which you want to use instead of the what you have in ~/.kube/config, you can do one of three things.

• Merge it with the contents of ~/.kube/config. This can be a bit tricky but is certainly doable, as kubectl supports multiple contexts (more on this in next section).
• Set the KUBECONFIG environment variable to the file path of the kube config file. This will make kubectl use that instead of the default.
• Set the --kubeconfig flag to the file path of the kube config file. This will make kubectl use that instead of the default.

To view your current kube config run

kubectl config view

Kube Context #

The k8s CLI has a context system, which let’s us store multiple k8s cluster connections. Why might we need this? Different environments are often isolated to different clusters. When we use the kubectl CLI, the commands will reach the Kubernetes cluster which our context currently points to.

To list your contexts run

kubectl config get-contexts

The context which is currently active will be marked by an asterisk in the CURRENT column. If your current context isn’t the one you expected you can change it using

kubectl config use-context <context-name>

You view the kube config only showing what’s relevant to your current context

kubectl config view --minify