[2834] in linux-scsi channel archive
Termination-probs?
daemon@ATHENA.MIT.EDU (Eike Jonas)
Wed Nov 19 12:25:44 1997
Date: Wed, 19 Nov 1997 18:04:56 +0100
From: Eike Jonas <mecca@rrz.uni-hamburg.de>
To: Linux SCSI list <linux-scsi@vger.rutgers.edu>
Hi
Since I added an new HD, my old one pulls off some strange behaviour..
AS I put my HD´s into an external case at the same time, I think that
the prob will be either the cable or the termination.
I got an Adaptec 2940 with an internal toshiba 4101 CDrom and external
Hds: NEC DSE2010S (this is the one that messes around), Quantum Atlas
XP32150.
At 1st the probs where real heavy and I got kernel panics more or less
often. Then I got a nu .9m SCSI-cable (before was 1.5m) and no more
kernel panics. Still the messages are the same:
like this (mostly before changing the cable):
Oct 25 23:02:32 mecca kernel: SCSI disk error : host 0 channel 0 id 0
lun 0 return code = 28000002
Oct 25 23:02:32 mecca kernel: Current error sd08:02: sense key Hardware
Error
Oct 25 23:02:32 mecca kernel: Additional sense indicates No seek
complete
Oct 25 23:02:32 mecca kernel: scsidisk I/O error: dev 08:02, sector
1741474, absolute sector 2994544
Oct 25 23:10:19 mecca kernel: (scsi0:0:0) Underflow - Wanted at least
4096, got 1536, residual SG count 1.
--
or this (not so often):
Oct 16 16:28:42 mecca kernel: ASC= 2 ASCQ= 0
Oct 16 16:28:42 mecca kernel: Raw sense data:0xf0 0x00 0x04 0x00 0x2d
0x6d 0x12 0x0e 0x00 0x00 0x00 0x00 0x02 0x00 0x00 0x80
Oct 16 16:28:42 mecca kernel: scsidisk I/O error: dev 08:02, sector
1723972
Oct 16 16:29:11 mecca kernel: (scsi0:0:0) Underflow - Wanted at least
8192, got 0, residual SG count 1.
or just this (mostly):
Nov 19 16:46:17 mecca kernel: SCSI disk error : host 0 channel 0 id 0
lun 0 return code = 28000002
Nov 19 16:46:17 mecca kernel: Current error sd08:02: sense key Hardware
Error
Nov 19 16:46:17 mecca kernel: Additional sense indicates No seek
complete
Nov 19 16:46:17 mecca kernel: scsidisk I/O error: dev 08:02, sector
1631398, absolute sector 2884468
I got an unpatched 2.0.31 running...
I wonder if this might be a termination-prob, but it´s just SCSI (nut
ultra not wide) and I already changed external termination to be made by
the quantum drive (nofx) and if anyybody knows something about this I´d
appreciate some info before spending 50 mark on active terminators.
--
CU
+---------------------------------------+
|Eike Jonas / mecca@rrz.uni-hamburg.de |
| > Gleiten statt schrubben < |
+---------------------------------------+
