In order to restore an etcd backup the documentation says (https://kubernetes.io/docs/tasks/administer-cluster/configure-upgrade-etcd/#restoring-an-etcd-cluster):
stop all API server instances
restore state in all etcd instances
restart all API server instances
How can I do it? Deleting the api server pod will immediately start it again.
Kube API Server runs as static Pod. The manifest of api-server (kube-apiserver.yaml) is present in /etc/kubernetes/manifests directory. Moving this file to a different folder will result in complete termination of api-server pod.