From: Steven Timm (timm@fnal.gov)
Date: Thu Jan 23 2003 - 11:14:51 EST
I have an Alphaserver 2100A. Four processers, Tru64 4.0g.
Attached to SCSI bus 3 are two tape jukeboxes, each of
which use two SCSI ID's.
scu> set nexus bus 3 target 0 lun 0
Device: EXB-85058SQANXR1, Bus: 3, Target: 0, Lun: 0, Type: Sequential
Access
scu> set nexus bus 3 target 1 lun 0
Device: EXB-10e, Bus: 3, Target: 1, Lun: 0, Type: Medium Changer
scu> set nexus bus 3 target 2 lun 0
Device: EXB-85058SQANXR1, Bus: 3, Target: 2, Lun: 0, Type: Sequential
Access
scu> set nexus bus 3 target 3 lun 0
Device: EXB-10e, Bus: 3, Target: 3, Lun: 0, Type: Medium Changer
At one point in the past this setup did work fine.
The current symptom is that when I run the sys_check utility,
part of what it does is to execute
"scu scan edt 3"
And this process hangs
root 28685 0.0 0.0 2.32M 8K ?? U Jan 21 0:00.02
scu scan edt bus 3
No further tape action is possible until the machine is rebooted, on
this bus or any other.
If I do a
"set nexus bus 3 target 1 lun 0"
and then issue a "tur" command
it will wait for ever but it can be aborted with control-C.
The only change I know that happened to make it stop working is the
following: We changed from an alphaserver 2100 chassis
to an alphaserver 2100A chassis--so now there are two PCI buses
instead of one. But this scsi controller is still on
the pci0 bus so that shouldn't have changed anything.
If anyone has any ideas what might be wrong, please
let me know.
Steve Timm
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:04 EDT