[storwize svc] In _update_replication_group On failure it should not leave the volume relation in stopped state and RCCG in stopped state

[Preeti duhan]

Storage Backend: Storwize svc

Scenario 1:

1)Create 2 bootable volumes. Try to create clone of the volumes.

2)Rccg group and volumes will be in inconsistent copying state.

3)Add another volume to the group, now the _update_replication_grp will fail since rccg will be in inconsistent_stopped state and volume will be in consistent stopped state.

4)And on failure , it leaves group into consistent stopped for forever.

Expectation: On failure it should not leave the volume relation in stopped state and RCCG in stopped state

Issue 2:

Update Group should not mark into ERROR, when any validation failure

Issue 3:
While adding multiple volumes, we should not start rccg for each volume addition. We can avoid multiple calls.

Issue 4:
Group API should have an interface to return CG live status.(just to monitor progress)