Hi Ann,
is this the fc or i incarnation of the D2D4106? Or in other words, is the VTL connected through Fibre Channel or iSCSI? In case of iSCSI, the checkbox is essential (as with any reboot of the MA host, drives are changing device numbers wildly, the S/N is the only constant). I don't know how bad that is with FC, but I always check this box on any medium - it almost always helps more than it could ever create issues.
That being said, if you use the iSCSI incarnation, there are several things that can go wrong:
- D2D rebooted
- MA host rebooted when D2D was unreachable
- General network issues in the iSCSI broadcast domain
These are all badly enhanced by Windows iSCSI initiator being unable to cleanly pick up a device that disappeared and returned to the living. You simply cannot tell it to reconnect to any permanent iSCSI targets, not by clicking some GUI, not by using the CLI and not even by restarting the iSCSI initiator service. The only way to force a proper reconnect to all permanent targets is to reboot the machine. After that, the devices should be connected again, and have randomized device names, but that is helped by the above mechanism to redetect where the devices turned up this time.
There should not be an ongoing issue like you describe. However, in practice, I've already seen one that looked quite similar - in that case, somebody upgraded a firewall device in the iSCSI broadcast domain in question, and due to a mishap, that device started to proxy ARP for the D2D's address. You can imagine how iSCSI goes bonkers in such a situation. So ongoing issues might point to a network problem, specifically repeated loss of connectivity or a duplicate IP address.
HTH,
Andre.