Accessing a k8s Cluster

After obtaining a kubeconfig for a cluster it’s time to get setup and access that cluster. This is done mostly via the command line using the kubectl commands.

Specify Kubeconfig

There’s a few different ways that a kubeconfig can be specified. The k8s documentation here goes in to more detail on those methods.

Note

Access to k8s resources is also controlled via kubeconfigs. In the examples below you may not have the applicable permissions to view the resources specified. If you get a response that contains Error from server (Forbidden) you are communicating with the cluster, but you do not have permission to view the requested resources.

Default kubeconfig

By default kubectl commands look to the ~/.kube/config.yml file. If you do not specify a kubeconfig to use explicitly kubectl by is configured to use this file by default. You can use any kubeconfig files to overwrite this file and then you don’t have to specify what file to use, this is not recommended. Instead you should use one of the methods below

Specify inline

When you run any kubectl commands you can specify the kubeconfig to use inline by adding the --kubeconfig flag like in the following example:

kubectl get nodes --kubeconfig ~/Downloads/siodine.yml

Export to KUBECONFIG variable

Another method to specify the kubeconfig is to specify the file location in an environment variable, specifically KUBECONFIG. The following example shows how to export the variable (Unix) from the command line:

export KUBECONFIG=/home/user/Downloads/siodine.yml
kubectl get nodes

Note

The export command does not persist after a reboot. If you want the export to persist it should be added to the ~/.bashrc file or the equivalent for your environment. Remember to also run source ~/.bashrc, or your environment equivalent, after to apply any changes. If you are unsure how to create persistent variables for your operating system please search the web on how to accomplish this.