Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 11 Sep 1998 08:00:28 -0500
From:      Zach Heilig <zach@gaffaneys.com>
To:        ben@rosengart.com, current@FreeBSD.ORG
Subject:   Re: soft updates panic
Message-ID:  <19980911080028.A24185@znh.org>
In-Reply-To: <Pine.GSO.4.02.9809110658360.10301-100000@echonyc.com>; from Snob Art Genre on Fri, Sep 11, 1998 at 06:59:33AM -0400
References:  <Pine.GSO.4.02.9809102021450.17494-100000@echonyc.com> <Pine.GSO.4.02.9809110658360.10301-100000@echonyc.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Sep 11, 1998 at 06:59:33AM -0400, Snob Art Genre wrote:
> Another crash in make -j3 world last night, but no panic or core dump
> this time.

> It's disturbing to me to see this this close to release.

I have gotten a couple of crashes recently, but it refused to dump core (I
later figured out that crash dumps were set to go to /dev/sd0s4b, except
FreeBSD is on /dev/sd0s1...  and I missed the dumpon complaint)

two of the crashes I saw were related to soft-updates (perhaps not caused by
them, but definately related).  The other one seemed to be related to the
filesystem code, but not specifically softupdates.

I have noticed when I compile many things as myself (not root), like 3 kernels
(one for each of my machines), there are compiler sig-11's in one random
session:

make -j8 world (as root) as well as these kernels made under myself:

3 make -j4 -> one session ends with sig-11 or other strange
              (not signal based) compiler error.
2 make -j4, 1 make -j2 -> everything seems to work fine.

the 'make -j8 world' always works (unless it was one of the filesystem-related
crashes), one of the kernel compiles gets smacked.

swap usage hovers about 10-12% (~40-45 out of 384Meg), 64 Meg ram.  There is
some heavy paging at times (2-3 meg per interval in top).

If the limits in /etc/login.conf are 'per user' rather than 'per session',
I can see how one of the kernel compiles could be smacked.  It is not spelled
out very well which policy is in effect (the implication seems to be 'per
user').

-- 
Zach Heilig -- zach@gaffaneys.com
  A horse without a nose... never wins. (unknown)

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980911080028.A24185>