On Thu, Apr 22, 2004 at 11:03:57AM -0500, Karl.Rossing AT Federated DOT CA 
wrote:
> Solaris NB 5.0MP1 server running nb 5.0MP1
> Solaris NB 5.0GA Solaris clients work fine **apparently mp1 has problems 
> with exclude lists. 
> Windows2k NB 5.0GA clients work but...

As for Solaris 5.0GA, there are issues with include/exclude lists
As for Solaris 5.0MP1, there are different issues with include/exclude
lists.  There is a workaround for this issue - contact Veritas if your
include/exclude lists are being ignored.

FWIW, there are also issues with scheduling Windows jobs on a Solaris
5.0MP1 server.  We're still awaiting a resolution to this from Veritas -
another customer had already escalated it to engineering before we did.

> If i submit the backup using the admin console they work. If i submit the 
> backup using bpbackup on the NB server, they fail... eventually.
> 
> What seems to happen is they get a status 25(cannot connect on socket), 
> then all the queued jobs seem to get 96 (unable to allocate new media for 
> backup, storage unit has none available)
> 
> What seems more bizar, is if one backup goes south, all the others follow.
> 
> Maybe it's because i have two autoloaders, each with one tape in them... 
> But the number of tapes doesn't seem to mater with our unix clients.

Windows and Unix can't share the same storage unit since you can't mix
Windows and Unix data on the same tape (type 13 vs type 0).  So, your
Windows storage unit likely has no free tapes which is what will give
you the 96.  When you get an error 25, is the current tape frozen?  If
so, that would explain the 96 errors since you're out of tapes.
-- 
Ed Wilts, Mounds View, MN, USA
mailto:ewilts AT ewilts DOT org