As an OpenShift administrator managing a cluster with OpenShift Virtualization deployed, I want to back up my VirtualMachines with virtual disk quiesce.

While discussing the requirements for backing up VMs we identified some work items that the OADP team can assist with:

• Provide a mechanism to prioritize VirtualMachine object backup over other cluster resources such as PVCs and Pods. This will allow a kubevirt velero plugin to trigger a quiesce via the pre-backup and post-backup hooks. The quiesce must happen before the PVCs associated with the VM are backed up.
• Double-check velero behavior to ensure that when a plugin includes resources to be backed up that these resources are backed up right away. We want to minimize the time that elapses between the pre-backup hook and the post-backup hook since IO inside the VM will be paused during that time.

This should be supported by Velero's existing configuration around resource priority, but need to confirm the workflow with hooks is possible.

Test VM definition: Centos 9 stream
https://github.com/kaovilai/cnv-manifests/blob/main/vm.yaml
OADP kubevirt supported version 1.0.2