I have been working in a lab environment and came across one of several
strange issues but this one being with CIFS.

The three cluster nodes are all NetWare 6.5 SP5.

When I created a new volume both through iManager and NSSMU e.g. DATA1
and brought it online I can get it to work on the server on which the
pool and volume was created.

If I then migrate the DATA1 resource to another server in the cluster,
the pool mounts and the CIFS ADD command is run but there is an error
regarding CIFS not being able to read attributes for the volume and the
volume is deactivated.

If I then edit the cluster resource script and REM out the CIFS ADD
command and take the resource offline and bring it back on again the
volume can mount and be migrated happily between cluster nodes.

Then, if I remove the REM from the CIFS ADD line and offline and online
the resource CIFS loads and the volume loads and I can migrate the
volume with CIFS loaded between all nodes.

Strange behaviour.