Managing Multiple Config Files

How to easily manage and switch between multiple Kubernetes environments.


If you frequently switch between multiple Kuberentes clusters from different cloud providers, it can be tedious to manage your Kubernetes configuration.

In this document we offer some helpful hints for configuring and switching between multiple Kubernetes clusters.

Where Configurations are Stored

Main Config File

By default, Kubernetes creates a configuration file stored at ~/.kube/config.

This file is referenced by an environment variable named KUBECONFIG.

The kubeconfig file can contain contexts for multiple Kubernetes clusters. It is often the case when using the command line tools provided by cloud vendors (for example, aws from Amazon, az from Microsoft Azure, and gcloud from Google) that the kubeconfig file will be automatically updated and new configuration from newly created contexts can be merged into it by the CLI.

It is convenient when multiple configurations are stored in a single file, because it means you don’t have to update your KUBECONFIG environment variable to connect to a new cluster.

Alternate Config Files

Unfortunately, not every cloud vendor provides a command line tool that can retrieve the configuration and merge it into the ~/.kube/config file.

Another common method that cloud vendors give to retrieve the cluster configuration is to download it from their web portal. In this case, you need to store the configuration as a separate file. It helps to keep all of these alternate configuration files together in a single folder.

We suggest creating a folder at ~/.kube/configs and placing kubeconfig files there. You then need to update your KUBECONFIG environment variable to reference every file that is in ~/.kube/configs.

Keeping KUBECONFIG Updated

Here is a script you can run that will refresh your KUBECONFIG environment variable to reference all the files contained within ~/.kube/configs.

#!/usr/bin/env bash
# If there's already a kubeconfig file in ~/.kube/config it will import that too and all the contexts
# Your additional kubeconfig files should be inside ~/.kube/configs
for kubeconfigFile in `find "${ADD_KUBECONFIG_FILES}" -type f -name "*.yml" -o -name "*.yaml"`
    export KUBECONFIG="$kubeconfigFile:$KUBECONFIG"

Create a file ending in the .sh extension and place the file in a folder in your PATH.

After that, whenever you add a new file to your ~/.kube/configs folder, you can run this script to refresh your KUBECONFIG environment variable.

Setting KUBECONFIG at Shell Startup

If you’re using Bash (the default terminal shell for most Linux, Mac, and Windows Subsystem for Linux installations), set your KUBECONFIG file by running the script in your .bashrc or .bash_profile files.

For example, if you named your script and placed it in your ~/scripts folder, you can call it from your shell loading script with the line:

source ~/scripts/

Aliasing kubectl

In many of our videos, you may see us shorten the typing of kubectl and instead simply type k. To create a similar alias for yourself, add an alias command to your .bashrc or .bash_profile file.

alias k='kubectl'

Using krew to Manage kubectl Plugins

There are many kubectl plugins available that can extend the command line functionality and make it easier to perform tasks. Plugins are managed using krew which itself is a kubectl plugin.

To install krew, follow the instructions here.

Once you’ve installed krew, you can add a plugin with the command:

kubectl krew install plugin-name-here

For a list of plugins, click here.

Switching Contexts Using the ctx Plugin

The ctx plugin can speed up the way you switch Kubernetes contexts. See here for more information.

To install ctx do:

kubectl krew install ctx

Similarly, the ns plugin makes it faster to switch namespaces. To install ns do:

kubectl krew install ns

To get a list of current contexts with ctx do:

kubectl ctx

To switch to a different context, do:

kubectl ctx context-name-here