chore(upgrade): modify check for rebuilding

[sinhaashish]

No description provided.

[abhilashshetty04]

@sinhaashish, Lets add nexus state validation before cordoning the node. Not necessary to list all nexuses in cluster. Only nexuses of the node about to be cordoned is fine. Since rebuild_wait doesnt account for Faulted state we might switchover nexus and trigger unnecessary Full rebuild.

@tiagolobocastro @avishnu , During every node upgrade, wait_for_rebuild lists all volumes and checks rebuild_progress in all childs. This will not return if any child has rebuild_progress. I think we should wait even if child is Faulted as eventually same or different child will rebuild if Faulted and then we will have rebuild_pregress populated. It might happen on the next node upgrade.

[sinhaashish]

Closing this PR as a consolidated #314 is raised to resolve this issue.