Description of problem:

Destination VM is observing XFS filesystem corruption after migrating using VMware VM to OpenShift Virtualization using MTV.

This has already been reported in MTV-1656 , but in addition we still can experience some reproductions even with the candidate fix for MTV-1656.

 
As per https://developer.broadcom.com/xapis/virtual-infrastructure-json-api/8.0.1.0/sdk/vim25/release/VirtualMachineSnapshot/moId/RemoveSnapshot_Task/post/ , RemoveSnapshot API returns 200 OK and a task to be monitored.

MTV code:
https://github.com/kubev2v/forklift/blob/b75f9abba8b5193da199b0fdc76783196f101690/pkg/controller/plan/adapter/vsphere/client.go#L384-L388
is currently ignoring that task and immediately trying to create a new snapshot and indeed in the logs we see that

2024-11-18 11:19:57.715	Removing snapshot
2024-11-18 11:19:57.775	Creating snapshot
2024-11-18 11:26:37.014	Created snapshot
2024-11-18 11:26:42.268	RemoveSnapshot
2024-11-18 11:26:42.268	Removing snapshot
2024-11-18 11:26:42.311	Creating snapshot
2024-11-18 11:34:49.845	Created snapshot

Removing snapshot and Creating snapshot happens at only about 50 msec of difference which is obviously not enough for it.

Version-Release number of selected component (if applicable):

Migration Toolkit for Virtualization Operator   2.7.3

OpenShift Virtualization   4.17.0

Guest is RHEL 9.4

How reproducible:

???

Actual results:

XFS filesystem corruption after warm migration of VM from VMware

Expected results:

No corruption