[Feature] Modularize and enhance status management according to multi-node ETCD proposal

[amshuman-kr]

Feature (What you would like to be added):

1. Enhance the Etcd resource status structure according to the changes proposed here
2. Enhance etcd-backup-restore and etcd-druid to update the corresponding member's health status in the Etcd resource status. This need not include the task of cutting off traffic in case of backup failure yet as the evaluation/decision on that is pending for the scenario of multi-node ETCD with ephemeral persistence. The etcd-backup-restore needs to consider the following scenarios in the implementation.
   • etcd-druid can maintain the AllMembersReady and Ready conditions as well as the following transitions for the member status where the member's etcd-backup-restore is unable to update its own status. Probably this has to be done in etcd-druid by enhancing the etcd status controller (custodian).
     1. From stale health to Unknown
     2. From Unknown to NotReady
     3. From not ready pod to NotReady status.

Motivation (Why is this needed?):

Pick individually executable pieces of the multi-node proposal.

Approach/Hint to the implement solution (optional):

Also, it would be preferable to use StatusWriter.Patch() to avoid race-conditions.

[amshuman-kr]

Split the updating of the Etcd resource status structure out to #168.

[amshuman-kr]

etcd-backup-restore can maintain the BackupReady and the Members status.

Moved this to "[Feature] Make etcd-backup-restore Kubernetes-aware keep member and backup status up-to-date #333".

[timuthy]

/assign