Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

SCSI problems with Qlogic ISP

1 view
Skip to first unread message

Marcus Williams

unread,
Jan 11, 2002, 10:21:50 AM1/11/02
to
Hi -

I've been seeing a lot of scsi errors recently on some of the drives
attached to an alpha we have. All of these appear to be sense/parity
errors of the form:

scsi0 channel 0 : resetting for second half of retries.
SCSI bus is being reset for host 0 channel 0.
SCSI disk error : host 0 channel 0 id 2 lun 0 return code = 18000002
[valid=0] Info fld=0x0, Current sd08:23: sense key Aborted Command
Additional sense indicates Scsi parity error
scsidisk I/O error: dev 08:23, sector 8913048

We're running 4 drives + one tape drive off a QLogic 1020 ISP adapter
(via a wide->narrow adapter). The total cable length exceeds 1.5m
which I've seen elsewhere on the web as being the max length for this
set up. The whole bus is terminated with an active terminator and the
QLogic adapter is set to auto-terminate. We're running Debian 2.2,
kernel 2.2.14 with the qlogicisp driver.

Are these sort of errors indicative of cabling/termination problems or
should I be looking elsewhere? If we swap around/replace the cables
the drive id causing the error changes so I cant believe its a drive
error. What worries me is the whole machine locks up solid after these
errors and I dont know whether these errors should cause that...

Any ideas?

Marcus

--
Marcus Williams - http://www.onq2.com
Quintic Ltd, 39 Newnham Rd, Cambridge, CB3 9EY


--
To UNSUBSCRIBE, email to debian-alp...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org

jaap.ho...@nl.abnamro.com

unread,
Jan 11, 2002, 10:44:38 AM1/11/02
to

>Hi -
>
>I've been seeing a lot of scsi errors recently on some of the drives
>attached to an alpha we have. All of these appear to be sense/parity
>errors of the form:
>
>scsi0 channel 0 : resetting for second half of retries.
>SCSI bus is being reset for host 0 channel 0.
>SCSI disk error : host 0 channel 0 id 2 lun 0 return code = 18000002
>[valid=0] Info fld=0x0, Current sd08:23: sense key Aborted Command
>Additional sense indicates Scsi parity error
>scsidisk I/O error: dev 08:23, sector 8913048
>
>We're running 4 drives + one tape drive off a QLogic 1020 ISP adapter
>(via a wide->narrow adapter). The total cable length exceeds 1.5m
>which I've seen elsewhere on the web as being the max length for this
>set up. The whole bus is terminated with an active terminator and the
>QLogic adapter is set to auto-terminate. We're running Debian 2.2,
>kernel 2.2.14 with the qlogicisp driver.
>
>Are these sort of errors indicative of cabling/termination problems or
>should I be looking elsewhere? If we swap around/replace the cables
>the drive id causing the error changes so I cant believe its a drive
>error. What worries me is the whole machine locks up solid after these
>errors and I dont know whether these errors should cause that...
>
>Any ideas?
>
>Marcus

Marcus,

This could be caused by bad cabling / termination.
You mention the use of a wide-->narrow convertor to
connect the tapedrive.
Does this adapter have "high byte termination" ?
(when you convert from wide(16 bit) to narrow(8 bit)
scsi , you need to terminate the upper 8 bits of the scsi bus.)
On the tapedrive (or the end of the lower 8 bits) you need to terminate
the lower 8 bits as well.

This setup could well confuse your auto termination too , maybe
you should try using a real terminator or setting jumpers on the
controller end too.

Jaap.

---------------------------------------------------------------------------
This message (including any attachments) is confidential and may be
privileged. If you have received it by mistake please notify the sender by
return e-mail and delete this message from your system. Any unauthorised
use or dissemination of this message in whole or in part is strictly
prohibited. Please note that e-mails are susceptible to change. ABN AMRO
Bank N.V. (including its group companies) shall not be liable for the
improper or incomplete transmission of the information contained in this
communication nor for any delay in its receipt or damage to your system.
ABN AMRO Bank N.V. (or its group companies) does not guarantee that the
integrity of this communication has been maintained nor that this
communication is free of viruses, interceptions or interference.
---------------------------------------------------------------------------

Paul Slootman

unread,
Jan 11, 2002, 10:44:26 AM1/11/02
to
On Fri 11 Jan 2002, Marcus Williams wrote:

> (via a wide->narrow adapter). The total cable length exceeds 1.5m

> QLogic adapter is set to auto-terminate. We're running Debian 2.2,

> Are these sort of errors indicative of cabling/termination problems or

I'd say the problem is most probably cable-related. Especially
wide->narrow adapters should be treated suspiciously...

Also, auto-terminate won't always work well.

Perhaps there's room to stick in a simple cheap symbios 860 card, which
is narrow? Otherwise, perhaps there are parameters for the module to
not do sync, or at least not so quickly, etc....


Paul Slootman

Marcus Williams

unread,
Jan 11, 2002, 11:02:14 AM1/11/02
to
> -----Original Message-----
> From: jaap.ho...@nl.abnamro.com
[mailto:jaap.ho...@nl.abnamro.com]
[snip]
> This could be caused by bad cabling / termination.
> You mention the use of a wide-->narrow convertor to
> connect the tapedrive.
> Does this adapter have "high byte termination" ?
> (when you convert from wide(16 bit) to narrow(8 bit)
> scsi , you need to terminate the upper 8 bits of the scsi bus.)
[snip]

Sorry, should have made it a bit clearer. I have:

Alpha (with QLogicISP (Wide SCSI)) -> Wide to narrow adapter -> disc
1 -> ... -> disc 4 -> tape drive

Not sure about the high byte termination - I've noticed that I can set
the Qlogic adapter to provide high byte termination though - would
that do the same?

Cheers

Marcus

--
Marcus Williams - http://www.onq2.com
Quintic Ltd, 39 Newnham Rd, Cambridge, CB3 9EY

--

Jay Estabrook

unread,
Jan 11, 2002, 11:07:44 AM1/11/02
to
On Fri, Jan 11, 2002 at 03:23:23PM -0000, Marcus Williams wrote:
>
> We're running 4 drives + one tape drive off a QLogic 1020 ISP adapter
> (via a wide->narrow adapter). The total cable length exceeds 1.5m
> which I've seen elsewhere on the web as being the max length for this
> set up. The whole bus is terminated with an active terminator and the
> QLogic adapter is set to auto-terminate. We're running Debian 2.2,
> kernel 2.2.14 with the qlogicisp driver.
>
> Are these sort of errors indicative of cabling/termination problems or
> should I be looking elsewhere? If we swap around/replace the cables
> the drive id causing the error changes so I cant believe its a drive
> error. What worries me is the whole machine locks up solid after these
> errors and I dont know whether these errors should cause that...
>
> Any ideas?

Yes; if at all possible, get the tape drive off the SCSI bus that has
the HDDs. Add another 1020, or possibly NCR810, to run only the tape,
or perhaps the tape plus CD, and keep only the HDDs on the original
SCSI bus.

The QLogic driver is notorious for having problems with tape drives,
and is very sensitive to termination. Identical setups to yours will
work flawlessly under Tru64, and have problems under Linux, because
of the drivers.

Good luck.

--Jay++

-----------------------------------------------------------------------------
Jay A Estabrook Alpha Engineering - LINUX Project
Compaq Computer Corp. - MRO1-2/K15 (508) 467-2080
200 Forest Street, Marlboro MA 01752 Jay.Es...@compaq.com
-----------------------------------------------------------------------------

0 new messages