*BSD News Article 79949


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!nntp.coast.net!howland.erols.net!newsfeed.internetmci.com!feed1.news.erols.com!news
From: Ken Bigelow <kbigelow@www.play-hookey.com>
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Re: FreeBSD hangs
Date: Sat, 05 Oct 1996 07:28:04 +0000
Organization: Erol's Internet Services
Lines: 46
Message-ID: <32560E04.3163@www.play-hookey.com>
References: <FRODEF.96Oct1160443@dslab5.cs.uit.no>
Reply-To: kbigelow@www.play-hookey.com
NNTP-Posting-Host: kenjb05.play-hookey.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0 (Win16; U)

Frode Vatvedt Fjeld wrote:
> 
> I recently installed the latest FreeBSD snapshot on my PC. In most
> respects it works very well, except that sometimes (every few hours)
> everything freezes permanently. No message appears on the console. The
> reset-button appears to be the only way out. Everyone says FreeBSD is
> more stable than Linux (with which I'm more experienced, and this sort
> of thing rarely happened), so how come this happens all the time?
> 
> Does anyone have ideas where I should search for the problem? The
> snapshot in question is 2.2-960801-SNAP. I also tried briefly the
> 2.1.5-RELEASE, and the same thing happened there (I reinstalled
> _everything_ with the snapshot).
> 
> I discovered a file /var/crash/minfree once (it held the string
> "2048"), but I have no idea what that means.
> 
> Also, the top-3.3 package that came with the snapshot does not work,
> it allways coredumps when it tries to print the process-info.
> 
> I compiled my own kernel, and all devices appear to work well. It's an
> all SCSI-system (NCR 53C810 on motherboard, no IDE-drivers compiled
> in), otherwise it's a very standard PC configuration. I have 48M of
> swap, and I'm pretty sure it's not exhausted. The freeze has happened
> both under heavy load and no load at all.
> 
> Please help, I can't live with rebooting my computer every two hours!
> (I'd install Win95 if I could).
> 

Jump-at-conclusions guess, based on similar irregular behavior in a
totally different situation: Check for a possible flaky RAM somewhere.
Sometimes RAM can check out perfectly good, but then fail either
thermally or when one specific pattern is stored there.

We had a hell of a time with it until we swapped SIMMs one at a time to
locate the bad one. If you can, swap the whole set of SIMMs on the
machine you want to get working, and test the others on a separate
setup. In our situation it took several days, because the pattern
failure occurred only *after* the bad chip warmed up. Yecch!
-- 
Ken

Are you interested in   |
byte-sized education    |   http://www.play-hookey.com
over the Internet?      |