Very Computer

Hi,

Recently we have a PC installed RedHat 6.2 with 2.2.14-5.0smp kernel
experiencing problems with SCSI. We have two SCSI hard disk in this
machine. The problem is sometimes the kernel falls into continuos loop
of messages like

May 22 07:34:49 asap kernel: scsi : aborting command due to timeout :
pid 14258, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 31 d8 e0 00 00
10 00
May 22 07:34:49 asap kernel: scsi : aborting command due to timeout :
pid 14259, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 31 d8 f8 00 00
70 00
May 22 07:34:49 asap kernel: scsi : aborting command due to timeout :
pid 14260, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 24 41 00 00 00
08 00
May 22 07:34:51 asap kernel: SCSI host 0 abort (pid 14259) timed out -
resetting
May 22 07:34:51 asap kernel: SCSI bus is being reset for host 0 channel
0.
May 22 07:34:54 asap kernel: (scsi0:0:1:0) Synchronous at 20.0
Mbyte/sec, offset 15.
May 22 07:35:25 asap kernel: scsi : aborting command due to timeout :
pid 14430, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 32 25 58 00 00
80 00
May 22 07:35:25 asap kernel: scsi : aborting command due to timeout :
pid 14431, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 32 25 d8 00 00
80 00
May 22 07:35:25 asap kernel: scsi : aborting command due to timeout :
pid 14432, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 28 29 e0 00 00
08 00
May 22 07:35:27 asap kernel: SCSI host 0 abort (pid 14431) timed out -
resetting
May 22 07:35:27 asap kernel: SCSI bus is being reset for host 0 channel
0.
May 22 07:35:30 asap kernel: (scsi0:0:1:0) Synchronous at 20.0
Mbyte/sec, offset 15.

and the machine hangs (it seems it can't do any IO operations anymore).

Is there anyone know the reason for this problem and how to fix it?

Thank you very much!

Xiaoqin

I would suspect hardware first and check the usual suspects: cabling,
termination, and ID conflicts. If all those look fine then perhaps the
drive itself is actually failing. I'd check cabling first though, given
the symptoms.

You say you recently installed RH 6.2, if you never experienced any
problems with this machine with other OS's or versions then perhaps some
software issue is to blame, but I would consider that less likely. Only If
I could do an A/B comparison with another version and consistantly see it
with one and not the other would I suspect software in a situation like
this.

--
-Duane
-DNAware SoftLabs

In a feverish moment of semi-lucidity, "Xiaoqin Qiu"

Quote:> Hi,
> Recently we have a PC installed RedHat 6.2 with 2.2.14-5.0smp kernel
> experiencing problems with SCSI. We have two SCSI hard disk in this
> machine. The problem is sometimes the kernel falls into continuos loop
> of messages like
> May 22 07:34:49 asap kernel: scsi : aborting command due to timeout :
> pid 14258, scsi0, channel 0, id 1, lun 0 Read (10) 00 00 31 d8 e0 00 00
> 10 00

[snip]

Bad/nonexistant termination on the cable.

Or unsupported controller.

Peter

I had a similar problem with an adaptec card but tried an advansys card
instead and the same thing happened. I put it down to incompatabilities
within the so called SCSI standard, as soon  as I removed my old SCSI 1
drives the problem went away. I'm pretty sure termination wasn't the problem
as these drives were in the middle of the chain and unterminated (I removed
the resistor packs to make sure).
I now use both SCSI controllers, one with my old drives and the other with
my newer CD writer and hard disks. This works flawlessly.
Hope this helps.