We have a 7 node cluster, NW6.5/SP5. From time to time we can no longer
do a CIFS mapping to volumes on a particular node. We can still map to
local volumes on the node and even map to the cluster volumes if we use
the physical server name in the UNC path rather than the virtual server.
However attempting to map to a virtual server fails. CIFSSTOP/CIFSSTART
will not help. Migrating the volumes to another node will fix it (at
least for the migrated volumes). Rebooting the server will fix it.

Any ideas on this one? Thanks,