I work with Veritas Netbackup 3.4  and STK 9714 libraries with DLT 7000 drives.
          I am not sure your problem originates from the Breecehill library.
          I experience very similar problems to what you described, so I will write down freely some points I can think of.
 
          1.  Did the problems start when you upgraded to 3.4 , or have you experienced such problems prior to using 3.4 as well ?
          2.  Drives are brought down if they experience 3 i/o errors within 12 hours. Check the following places to see if this occur
               1. Media log report ( invoke it from xbpadm->reports )
                2. /usr/openv/netbackup/logs/bptm/log.<date>
                3. /usr/openv/netbackup/logs/bpbrm/log.<date>
                4. /var/adm/messages

                In all these log files grep for DOWN or Down - some lines before or after it will describe the reason for bringing the drives down

          3.  If you change passowrd on the netapps you have to report this change to Veritas NetBackup using /usr/openv/volmgr/bin/set_ndmp_attr command. If you fail to do it the drives go down.

           Hope this helps
           Please let me know.

           Best Regards
           Nachshon Yeshurun
           Motorola Semiconductor
 

 

mati

---

Subject: Veritas Netbackup 3.4 for Solaris & Breecehill Q6.210 DLT Library
Date: Tue, 27 Feb 2001 09:44:47 MST
From: Douglas Ritschel <Douglas.Ritschel@tddny.fujitsu.com>
To: "toasters@mathworks.com" <toasters@mathworks.com>

Does anyone have experience with Veritas Netbackup 3.4 for Solaris and a
Breecehill Q6.210 DLT Library?

We have a Breecehill Q6.210 with 6 Quantum DLT 7000 drives in it. The
first two drives are daisy chained to a NetApp 740. The second two are
daisy chained to the robot, which is connected to a Sun 450, running
Solaris 2.6. The third pair of drives are daisy chained to a NetApp 760. 

The problem, is that when I come in every Monday morning, after the our
Full backups are scheduled to run, 4 of the 6 drives are reported as
being down in Veritas Device Monitor. The 4 drives that are down are not
the same every week. To get the backups going again, I have to shutdown
the Veritas daemons, reboot the Library, reinventory the Library and
Verites, then restart the backups that did not complete. 

The following day, after some of the backups completed, some of the
drives are down again, and I have to repeat the process.