From owner-freebsd-bugs Fri Apr 3 21:40:24 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA00633 for freebsd-bugs-outgoing; Fri, 3 Apr 1998 21:40:24 -0800 (PST) (envelope-from owner-freebsd-bugs@FreeBSD.ORG) Received: from noao.edu (noao.edu [140.252.1.54]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id VAA00626 for ; Fri, 3 Apr 1998 21:40:17 -0800 (PST) (envelope-from grandi@noao.edu) Received: from mirfak.tuc.noao.edu (grandi@mirfak.tuc.noao.edu [140.252.1.9]) by noao.edu (8.8.8/8.8.7/SAG-02Dec97) with SMTP id WAA29207; Fri, 3 Apr 1998 22:40:11 -0700 (MST) (envelope-from grandi@noao.edu) Date: Fri, 3 Apr 1998 22:40:10 -0700 (MST) From: Steve Grandi X-Sender: grandi@mirfak.tuc.noao.edu Reply-To: Steve Grandi To: Vince Vielhaber cc: bugs@FreeBSD.ORG Subject: Re: Running out of places to turn... In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, 3 Apr 1998, Vince Vielhaber wrote: > > For a quick rundown, I installed 2.2.5 from CD on a P150 (Tyan MB) with > 256MB Ram, a 4GB and 6.4GB IDE and a pair of 9GB SCSI drives connected > to a 2940UW (that I have set to a max transfer rate of 10Mbps). This is > a news machine and I have the the ccd driver configured into the kernel > and I'm running the 2-9GB drives as one 18GB drive. The other kernel > changes are maxusers set to 256 and the mem size items set at 262100. > I was having similar problems.... How long does the expire take? My system would work fine as long as the expire would take a few minutes; but then something would happen and the expires would start taking 8 hours! And the system would soon crash; somtimes a hard lock up and sometimes with the same "page fault in Kernel mode" that you see. I finally figured it out...at least my system hasn't crashed for over a month. When expire is rebuilding the history file, it tries to fit the history.pag file in core. If it can't, expire resorts to using a disk file which is horribly inefficient and stresses the system something fierce...apparently enough to tickle a latent bug in FreeBSD that causes a subsequent crash. The solution, a common one for 2.2.5+ problems, is to edit /etc/login.conf; in particular the news entry. Here is what my entry now looks like: # # Settings used by news subsystem # news:\ :path=/usr/local/news/bin /bin /sbin /usr/bin /usr/sbin /usr/local/bin /usr/local/sbin:\ :cputime=infinity:\ :datasize=infinity:\ :datasize-cur=infinity:\ :stacksize=64M:\ :stacksize-cur=64M:\ :memorylocked=64M:\ :memoryuse=128M:\ :memoryuse-cur=128M:\ :filesize=infinity:\ :coredumpsize=infinity:\ :openfiles=512:\ :maxproc=128:\ :priority=0:\ :requirehome@:\ :umask=022:\ :tc=auth-root-defaults: Steve Grandi, National Optical Astronomy Observatories, Tucson, Arizona USA Internet: grandi@noao.edu Voice: +1 520 318-8228 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message