My suggestion is to not let them off the hook so easily. If the problem is
hardware, they should be able to provide concrete proof that it is. That
means they should be able to walk through every other component on the
system and prove that it is operating correctly and configured correctly.
You still have a non-functioning cluster that has to be fixed and doing the
"it's not my problem, it's their problem" doesn't help to get it fixed.
While it is running on VMWare, which isn't a supported environment, I have
seen this exact behavior where a file written to a shared drive on one node
is not visible to the other node upon failover.
Solid Quality Learning
"More than just Training"
SQL Server MVP