Update CA tool reports leader changed error is temporary, and user can run the tool again later to try again.

Problem

Update CA tool reports leader changed error when running kubectl commands to operate cluster and update CA certificate. For example:

update_ca[CRITICAL]: failed to update airgap ca cert to tkc/kcp/kct on mgmt cluster [mc210] with errors: RuntimeError('run "kubectl --kubeconfig /opt/vmware/k8s-bootstrapper/a1d9dcc8-d43f-4278-909e-0a22019a21db/kubeconfig --request-timeout 30s get ValidatingWebhookConfiguration tca-kubecluster-operator-validating-webhook-configuration -o yaml" failed, err: Error from server: etcdserver: leader changed\n',) 

Cause

Update CA tool will modify kubecontrolplane CR of cluster, which will lead to redeployment of control plane node , and migration of endpoint IP to new node. Therefore, the kubenetes API service may be unavailable during the endpoint IP floating to another node.

Solution

Rerun the update-all command after some time.