Data mover restore enhancement for WaitForFirstConsumer volume

[Lyndon-Li]

For a WaitForFirstConsumer volume, the PV should not be provisioned until the pod consuming it is scheduled. This conflicts with the generic restore process for data mover.
As the current behavior, the generic restore always wait there until the pod to be restored is scheduled.

This is good enough to make the generic restore works appropriately with WaitForFirstConsumer volumes, but it has some drawbacks:

1. This couples the data mover restore with the resource restore (workload pod restore), if the pod is not scheduled for any reason, the data mover restore will get wait timeout and fail
2. This doesn't allow the data mover restore to support restoring PVC/PV only without pod. While it is already supported for Immediate volumes

Actually, users must know their topology constrains if they have that, or they must know in which nodes their restored workload pods can be scheduled.
Therefore, we can add a parameter for each volume to be restored for users to specify the candidate nodes. Consequently, generic restore will not wait for the workload pod to be scheduled and get the node, but use the candidate nodes to schedule the restorePod.

[github-actions]

This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 14 days. If a Velero team member has requested log or more information, please provide the output of the shared commands.

[sseago]

not stale