Documentation URL: https://docs.openshift.com/container-platform/4.16/backup_and_restore/control_plane_backup_and_restore/replacing-unhealthy-etcd-member.html#restore-replace-stopped-etcd-member_replacing-unhealthy-etcd-member

During the resolution of OCPBUGS-16663, it was made the wrong assumption that all the environments have controlplanemachinesets and that, in all environments, if you delete a master machine, it gets re-created.

However, this is wrong. Controlplanemachinesets become available on each environment at different versions, so during several OCP versions there is (and there will be) the possibility that some clusters have controlplanemachinesets and some don't.

So the step 5 must be fixed like this:

• The user must check if a controlplanemachineset exists as per oc -n openshift-machine-api get controlplanemachineset output.
• If the controlplanemachineset exists, then they have to follow the current steps.
• If the controlplanemachineset doesn't exist, then they have to follow the steps documented before OCPBUGS-16663 PR was merged.