Did you have a power outage? Is there anything in the SMART information of the drives? Anything in the system logs since the previous scrub?
Removing the snapshots might not be enough. I would suggest removing the VM and restoring from backup.
How can i test this?Is it possible that there is a RAM error so all operation on storage is inconsistent and error prone?
memtest if you are in legacy mode? Some ILO or remote access console of the server with logs (generic or specific ECC memory)? Fix the UPS problem (could also be bad power coming to the server) or bypass it completely? I've no more ideas, sorry, and is just a shot in the dark.How can i test this?