Veritas-bu

[Veritas-bu] Eight new SDLT320 drives not being recognzied by NBU Datacenter 3 .4

2004-05-04 16:23:38
Subject: [Veritas-bu] Eight new SDLT320 drives not being recognzied by NBU Datacenter 3 .4
From: jon.hardt AT cendant DOT com (Hardt, Jon)
Date: Tue, 4 May 2004 14:23:38 -0600
Recently, we upgraded our tape library (StorageTek L180) from 4 Quantum 8000
tape drives to 8 Quantum SDLT320 drives. However, only 2 drives are usable
at this point. The other 6 drives are reported as DOWN-TLD in NBU 3.4

Our setup:

StorageTek L180
8 SDLT320 drives connected through 2 StorageTek 3300 routers
Local host is Sun 220R running Solaris 8 -- fiber connectivity HBAs JNI 6410

The odd thing is that a sgscan produces the correct mappings:

bash-2.03# ./sgscan all
/dev/sg/c0t6l0: Cdrom: "TOSHIBA DVD-ROM SD-M1401"
/dev/sg/c0tw21000004cf20f03el0: Disk (???): "SEAGATE ST336605FSUN36G"
/dev/sg/c0tw21000004cf20f8d3l0: Disk (???): "SEAGATE ST336605FSUN36G"
/dev/sg/c2t4l0: Tape (/dev/rmt/0): "QUANTUM SDLT320"
/dev/sg/c2t4l1: Tape (/dev/rmt/1): "QUANTUM SDLT320"
/dev/sg/c2t4l2: Tape (/dev/rmt/2): "QUANTUM SDLT320"
/dev/sg/c2t4l3: Tape (/dev/rmt/3): "QUANTUM SDLT320"
/dev/sg/c2t5l0: Tape (/dev/rmt/4): "QUANTUM SDLT320"
/dev/sg/c2t5l1: Tape (/dev/rmt/5): "QUANTUM SDLT320"
/dev/sg/c2t5l2: Tape (/dev/rmt/6): "QUANTUM SDLT320"
/dev/sg/c2t5l3: Tape (/dev/rmt/7): "QUANTUM SDLT320"
/dev/sg/c2t9l0: Changer: "STK     L180"

However, when trying to use drives 2, 3, 4, 6, 7 and 8 we will get some sort
of varation of this messages in /var/adm/messages:

May  4 11:01:12 med01den avrd[5623]: [ID 914860 daemon.error] Fatal open
error on Drive6 (device 5, /dev/rmt/5cbn),
 errno = 6 (No such device or address), DOWN'ing it
May  4 11:01:12 med01den avrd[5623]: [ID 529501 daemon.error] Fatal open
error on Drive7 (device 6, /dev/rmt/6cbn),
 errno = 6 (No such device or address), DOWN'ing it

Same type of this when trying to stat the drive:

bash-2.03# mt -f /dev/rmt/0cbn stat      # This is one of the working drives
/dev/rmt/0cbn: Device busy
bash-2.03# mt -f /dev/rmt/1cbn stat
/dev/rmt/1cbn: No such device or address
bash-2.03# mt -f /dev/rmt/2cbn stat
/dev/rmt/2cbn: No such device or address
bash-2.03# mt -f /dev/rmt/3cbn stat
/dev/rmt/3cbn: No such device or address
bash-2.03# mt -f /dev/rmt/4cbn stat        # This is the other working drive
/dev/rmt/4cbn: Device busy
bash-2.03# mt -f /dev/rmt/5cbn stat
/dev/rmt/5cbn: No such device or address
bash-2.03# mt -f /dev/rmt/6cbn stat
/dev/rmt/6cbn: No such device or address
bash-2.03# mt -f /dev/rmt/7cbn stat
/dev/rmt/7cbn: No such device or address
bash-2.03#

All the expect links exist in /dev/rmt ... notations for drives 0 through
7... I can't figure out why NBU doesn't think there is anything attached at
the /dev/rmt/Xcbn addresses. I've done what seems like hundreds of
reconfiguration reboots, but no love. I am totally confused; if anyone has
any idea on why I can't use the other 6 drives or could give me some good
troubleshooting steps, please fire away.

---
Jon Hardt
Unix Systems Administrator
Cheap Tickets, part of Trip Networks
6560 Greenwood Plaza Blvd., Englewood, CO 80111
Office: (303) 357-3314
Mobile: (303) 522-7537
jon.hardt AT cendant DOT com
YIM: jonhardt
www.cheaptickets.com