Add capability to restore an existing multi-node etcd cluster

[unmarshall]

How to categorize this issue?

/area control-plane
/area ops-productivity
/kind enhancement

What would you like to be added:
Add an ability to for an end-user to inform etcd-druid to restore an etcd cluster from existing backups.

Why is this needed:
To restore a multi-node etcd cluster, it is essential that you always go from 0->1 replicas, do the restoration from an existing backup and then only move from 1->n replicas.

If there is an existing mutli-node etcd cluster and its been scaled down and its PV's have not be preserved then there is a need to restore this cluster from a backup (if backup has been enabled).
In gardener context there are 2 use cases for this:

• Wake from Hibernation. Today we preserve the PVs and there is no way an end-user can influence this decision. However with Optimize costs on cluster hibernation #859 we will provide an ability to also remove the PVs after taking a full snapshot to ensure that no data is lost when the cluster is woken up from hibernation.
• Control plane migration. After the backups are copied from source bucket to target bucket, etcd cluster in the target control plane has to be restored from the backup.