Tru64 on GS60E: my DLT drives have disappeared!

[Vladimir Fabecic]

Now try:
# scu scan edt bus 3
# scu show edt
and post output

[Vladimir Fabecic]

You may have hardware problem with tape device on BUS 1 with SCSI ID 4.
So that may be the reason of your problem.
System can still see tape drive in BUS 3 SCSI ID 5.
What kernel parameters did you change (also do not think it had something to do with tape problem)?
First I would turn off the machine, and disconnect device on BUS 1 ID 4.
Then I would check if there were some software problems.

[Matt Hearn]

CAM Equipment Device Table (EDT) Information:

Bus/Target/Lun Device Type ANSI Vendor ID Product ID Revision N/W
-------------- ----------- ------ --------- ---------------- -------- ---
0 0 0 RAID SCSI-2 DEC HSG80CCL V87F W
0 0 1 Direct SCSI-2 DEC HSG80 V87F W
0 0 2 Direct SCSI-2 DEC HSG80 V87F W
0 0 3 Direct SCSI-2 DEC HSG80 V87F W
0 0 101 Direct SCSI-2 DEC HSG80 V87F W
0 0 102 Direct SCSI-2 DEC HSG80 V87F W
0 0 103 Direct SCSI-2 DEC HSG80 V87F W
0 0 104 Direct SCSI-2 DEC HSG80 V87F W
0 0 105 Direct SCSI-2 DEC HSG80 V87F W
0 0 106 Direct SCSI-2 DEC HSG80 V87F W
0 0 107 Direct SCSI-2 DEC HSG80 V87F W
0 0 108 Direct SCSI-2 DEC HSG80 V87F W
0 0 109 Direct SCSI-2 DEC HSG80 V87F W
0 1 0 RAID SCSI-2 DEC HSG80CCL V87F W
1 0 0 Changer SCSI-2 DEC TL800 (C) DEC 0525 W
2 0 0 RAID SCSI-2 DEC HSG80CCL V87F W
2 1 0 RAID SCSI-2 DEC HSG80CCL V87F W
2 1 1 Direct SCSI-2 DEC HSG80 V87F W
2 1 2 Direct SCSI-2 DEC HSG80 V87F W
2 1 3 Direct SCSI-2 DEC HSG80 V87F W
2 1 101 Direct SCSI-2 DEC HSG80 V87F W
2 1 102 Direct SCSI-2 DEC HSG80 V87F W
2 1 103 Direct SCSI-2 DEC HSG80 V87F W
2 1 104 Direct SCSI-2 DEC HSG80 V87F W
2 1 105 Direct SCSI-2 DEC HSG80 V87F W
2 1 106 Direct SCSI-2 DEC HSG80 V87F W
2 1 107 Direct SCSI-2 DEC HSG80 V87F W
2 1 108 Direct SCSI-2 DEC HSG80 V87F W
2 1 109 Direct SCSI-2 DEC HSG80 V87F W
3 5 0 Sequential SCSI-2 COMPAQ SuperDLT1 4B4B W
4 4 0 CD-ROM SCSI-2 DEC RRD47 (C) DEC 1206 N

I should note that the SuperDLT on bus 3 is a totally separate thing from from the TL891 library. The library drives are on the same bus as the changer, bus 1. I did run a scan on bus 1, but it hangs.

[Vladimir Fabecic]

So you did run a scan on bus 1, but it hangs.
This indicates there is some problem with something on that bus.
Turn off the library, change SCSI ID on other tape device and do:
# hwmgr -scan scsi

[DCBrown]

Sounds like a hardware issue. Disconnect ALL three devices from bus 1. Run a hwmgr scan component... it should now finish unless there is a scsi bus problem (i.e. termination or something else nasty like that).

If it finishes, attach changer, repeat.
If it finishes, remove changer, attach library box, repeat.

Are there any events in the binary.errlog? If so, what? Should be lots of cam errors and these should give a good indication of what's going on.

Bud

[Pieter 't Hart]

I would go to the library and from the front panel check it's settings.
If for some reason the changer and a drive have returned to a default SCSI-ID (both 0(!)) you get strange responses on this bus.
If so reconfigure the ID's from the front panel.
If you have multiple drives you might disconnect one of them from the bus.

As allready suggested check cables and terminator (replace if you have spare)
you mention a tlx drive so be sure the replacement terminator is a LVD-type.

[Rob Leadbeater]

Hi Matt,

If you look at the output of a "hwmgr show scsi -full" do you see any null devices on the relevant SCSI bus ?

If so, you may well have to get rid of these with "dsfmgr -R hwid ..." and "hwmgr delete component -id ..."

Hope this helps,

Regards,

Rob

[Matt Hearn]

Hey guys; turned out we had a number of hardware problems. The root cause was PROBABLY a scsi cable that's either too long or faulty, or both; the LVD SCSI controller is also bad, although that might be because I had to jiggle the bejebus out of it to replace the cable. :) The server had a spare SCSI card, so we're attached to that with a new, shorter cable, and the drives are there! Thanks for all your help!