Past weekend (as usual O_o) we have experienced the issue in our XCP 1.6 production pool.
Shared iSCSI storage was shutted down due to misconfigured UPS settings while XCP servers continued to work.
When storage was returned to working state and reconnected to pool most VM did not boot with the same message - VDI is not available.
Googling give me mentioned above method - forgot and reconnect VDI.
Result was even worser - the whole SR become unusable.
Storage rescan gazered lot of errors like bad header on LVM and many other.
Finally i've disconnect failed SR from pool, connect it back and SR become healthy (it looks so). But anyone VM was not start with disk from this SR and freeze during startup.
I did not find solution and restored most VMs from backup (long live VMPP!)
So, i just wanna say - be highly careful with VDI on shared storage repository in production environment)
_______________________________________________