Hello Managers:
Receieved 2 good leads within 1/2 hour of posting the message. Thanks to:
Mandell Degerness
David J. DeWolfe
for two good tips. The problem was a "panic (CPU 1) simple_lock: time
limit exceeded" ocurring on our server. Mandell noted that a new patch
may be available from DEC, and to contact Tech Support. David noted
(quite interestingly) that the problem seemed to be related to tape errors.
Sure enough, upon inspection of the UERF output, we had a tape error on our
TZ87 last night.
I've opened a problem log with DEC, and will post-summarize that result.
David gave me a mild hint (righly so) that the problem was previously
documented in the archives. I should have checked there first. Mea Culpa.
Scott
***************************************************************************
>From MDEGERNESS_at_galaxy.gov.bc.caWed Aug 14 11:55:00 1996
>Date: Wed, 14 Aug 1996 09:12:58 -0700 (PDT)
>From: Mandell Degerness of ITSD 389-3539 <MDEGERNESS_at_galaxy.gov.bc.ca>
>To: sbalneav_at_UWinnipeg.ca
>Subject: FWD: Kernel Panic - simple_lock
>
> Scott,
>
> It would appear from the message that you have a two CPU system.
> We have had this problem on our 2 CPU, 128 Mb, 2100 4/275 system
> since at least May. We have just received the latest patch kit
> from Digital and have not yet had a chance to try it (a production
> system - we are running with a single borrowed 5/300 processor
> until we can test the patch). Contact Digital Support,
> specifically, the contact we have been working with is George
> Stofko.
>
> Regards,
> Mandell Degerness
>
>From sxdjd_at_orca.alaska.eduWed Aug 14 11:55:05 1996
>Date: Wed, 14 Aug 1996 08:23:58 -0800
>From: "David J. DeWolfe" <sxdjd_at_orca.alaska.edu>
>To: "scott.balneaves" <scott.balneaves_at_UWINNIPEG.CA>
>Subject: Re: Kernel Panic - simple_lock
>
>Hi;
>
>Yes, we have seen this before. If you look through the list archives, you
>will find several detailed messages from one of my co-workers (Kurt Carlson,
>sxkac_at_alaska.edu) regarding our experiences with simple lock timeouts. In
>short, it resulted in kernel panics which resulted in adfvs corruptions and
>Oracle database corruptions. Kurt worked long an hard with Digital on this
>problem, and as I recall some problems were discovered and corrected in
>panic processing. Our worst case in regards to this was an oracle database
>that came up fine after the panic, yet 5 hours later died and would not come
>up. The recovery effort was approx 40 straight hours. We have received and
>applied patches which addressed several of the semi-related problems, and
>things seem to be pretty much OK now. However, our faith in Digital and
>Oracle was severly shaken by the course of events that all started with a
>"simple_lock: time limit exceeded".
>
>again, to reiterate, for complete details and all the actual facts please
>dig up the messages from the list archives (they would be within the last
>3.5 months).
>
>good luck...
>
>
>David J. DeWolfe
>Systems Programmer
>Statewide Office of Information Services
>University of Alaska
>907.474.7399
>sxdjd_at_orca.alaska.edu
>
>In a vicious struggle for survival intelligence emerges as the weapon of
> choice. - Nova, In Search of Human Origins
------------------------------------------------------------------------------
Scott Balneaves, U of W Networking Support | "There is hopeful symbolism in
Email: sbalneav_at_uwinnipeg.ca | the fact that flags do not wave
Homepage:
http://www.uwinnipeg.ca/~sbalneav/ | in a vacuum" - Arthur C. Clarke
Received on Wed Aug 14 1996 - 19:38:21 NZST