Hung Duong wrote:

Thank for reply to my message. I found the solution to this problem. It was the drive mapping in Networker. Since my tape drives connect to host via a FC Bridge, host discovers tape devices in a random order. This sometimes results in device mapping in the wrong order. On Solaris once device handles have been created the mapping of device names to physical SCSI address will not change. So, we must ensure the tapes devices have correct device files associated the them for use in Networker.

- Solution: for host to map devices in the right order, we need to turn on one device at the time. The process is as follows: 1. Remove the tape device files associated with the new drives from /dev/rmt/* 2. # sync; sync; reboot -- -rv

      3. Turn on first physical drive from the top.
      4. Ensure there are right amount of files in /dev/rmt/ directory.
      5. # sync; sync; reboot -- -rv
      6. Turn on second physical drive from the top.
      7. Ensure there are right amount of files in /dev/rmt/ directory.
      8. Continue until all drives had been added.

- Once the device files had been created, you can add new devices to Networker using jbedit. This way Networker will see drives in the same order as the host assign device files.

Regards,

Lee Stecklov wrote:

Hung Duong wrote:

We recently add two new SuperDLT drives to our existing Quantum P4000

Library. The new drives are connected to Solaris machine using FC HBA. OS

able to detect the drives and use it. I tested with mt and tar command

without any error. However, whenever I try to label a new tape (SDLT II) it

craps out with following message:

....

Jukebox handle being opened at port [email protected]
Slot `604' is being loaded into drive `/dev/rmt/6cbn'
box_display: `-' for drive `/dev/rmt/6cbn'
Jukebox handle being closed at port [email protected]
load sleep for 5 seconds
nsrjb: About to read label on /dev/rmt/6cbn
nsrjb: About to write label CIN600 on /dev/rmt/6cbn
nsrjb: About to eject volume from /dev/rmt/6cbn
slot number 604, write open error, I/O error
eject sleep for 60 seconds
Jukebox handle being opened at port [email protected]
Unloading jukebox: drive `/dev/rmt/6cbn' into slot `604'
Jukebox handle being closed at port [email protected]
unload sleep for 60 seconds
Error: while operating on slot `604': write open error, I/O error
Tue 16:52:00 write open error, I/O error

nsrjb: Jukebox error:  Tue 16:52:00 write open error, I/O error

Had anybody seen this problem before? I currently have a case open with
legato and they had not able to point out any problem. My environment
consists of:

Sun 280R with 2 CPUs, 1Gb RAM,
Solaris 8
Networker 7.1.3
Quantum P4000
SuperDLT 600
Sun JNI HBA FCX2-6562
SUN Storage Foundation 4.4.3

Thanks,

Manually load the volume into the drive using the library interface.
Run scanner -v /dev/rmt/6cbn

What output do you see ?

Lee Stecklov

--

Note: To sign off this list, send a "signoff networker" command via email

to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=