+1

The only problem may be that even the VM doesn't actually know the reason (it could just trigger the delete at vm-level which will just return a failure state).

Peter