Hi - we have been trying to test out some failover functionality within Harvester/Longhorn and while isolating a host, our VMs fail over fine and things seem OK. This was with just a basic VM deployment from Rancher to prepare for some more in-depth testing. When we started deploying our application from Rancher, we are getting an error related to PVCs.

AttachVolume.Attach failed for volume "pvc-45518d36-1a4e-4d00-b819-0bc7f244e080" : rpc error: code = DeadlineExceeded desc = Failed to wait the volume pvc-45518d36-1a4e-4d00-b819-0bc7f244e080 status to settled

Longhorn looks healthy, but shows volumes detached despite all nodes being schedulable with no errors. Same thing within Harvester.