Kubectl
Accessing the kubeconfig file
kubectl is used to access the kubernetes API server on your cluster from the command line. Unless you previously installed kubectl, then after provisioning the K3s cluster, the kubectl command probably references the kubectl provided by K3s. Check this with which kubectl.
By default, the K3s-provided kubectl references the cluster configuration provided at /etc/rancher/K3s/K3s.yaml which requires sudo privileges to run. This is deliberate, as it makes sure only users with superuser access privileges can manipulate the cluster. However, entering a sudo password quickly becomes annoying (convenience is the enemy of security).
Instead, the site playbook drops the cluster config file into the default kubectl configuration directory at ~/.kube/config (this is a file, not a directory). To get the K3s kubectl to reference this config file, either use the --kubeconfig flag or set the KUBECONFIG environment variable for your current shell.
If you followed along in the Installation section, then you probably have direnv set up, which can be used to load the KUBECONFIG env variable automatically when you cd into your local copy of the GitOps directory:
echo "export KUBECONFIG=$HOME/.kube/local-config" > .envrc
direnv allow
Created: June 13, 2024 16:06:52