From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266304AbUHROYX (ORCPT ); Wed, 18 Aug 2004 10:24:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266427AbUHROYX (ORCPT ); Wed, 18 Aug 2004 10:24:23 -0400 Received: from penguin.cohaesio.net ([212.97.129.34]:47006 "EHLO mail.cohaesio.net") by vger.kernel.org with ESMTP id S266304AbUHROYU (ORCPT ); Wed, 18 Aug 2004 10:24:20 -0400 From: Anders Saaby Organization: Cohaesio A/S To: William Lee Irwin III Subject: Re: oom-killer 2.6.8.1 Date: Wed, 18 Aug 2004 16:24:24 +0200 User-Agent: KMail/1.7 Cc: linux-kernel@vger.kernel.org References: <200408181455.42279.as@cohaesio.com> <20040818140550.GY11200@holomorphy.com> In-Reply-To: <20040818140550.GY11200@holomorphy.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200408181624.25131.as@cohaesio.com> X-OriginalArrivalTime: 18 Aug 2004 14:24:20.0242 (UTC) FILETIME=[0D116F20:01C4852F] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wednesday 18 August 2004 16:05, William Lee Irwin III wrote: > On Wed, Aug 18, 2004 at 02:55:42PM +0200, Anders Saaby wrote: > > This is a high-volume NFS server running almost no user-space > > applications. It serves a handfull of web server NFS clients from a > > ~700G XFS filesystem. The machine has about 2.5 GB of RAM and 4G of > > swap (which is almost not in use - i may use 5-10 MB total tops). > > CONFIG_HIGHMEM and CONFIG_HIGHMEM4G enabled, SMP enabled, preempt > > disabled. Today the OOM killer kicked in - it seemed that swap was almost > > unused at the time (which is strange, as that should prevent the OOM > > killer from kicking in). > > Relevant part of the syslog follows (syslogd was killed too eventually): > > This seems to have been meant to resolve laptop_mode issues, but looks > like it didn't get applied. I'm not convinced it will help given that > you appear to have a vanilla ZONE_NORMAL slab OOM (858MB slab), but you > never know. Capturing /proc/slabinfo data may be more helpful. > > > Index: oom-2.6.8-rc1/mm/vmscan.c > =================================================================== > --- oom-2.6.8-rc1.orig/mm/vmscan.c 2004-07-14 06:17:13.876343912 -0700 > +++ oom-2.6.8-rc1/mm/vmscan.c 2004-07-14 06:22:15.986416200 -0700 > @@ -417,7 +417,8 @@ > goto keep_locked; > if (!may_enter_fs) > goto keep_locked; > - if (laptop_mode && !sc->may_writepage) > + if (laptop_mode && !sc->may_writepage && > + !PageSwapCache(page)) > goto keep_locked; > > /* Page is dirty, try to write it out here */ laptop_mode is not set on this server <- :-) - So I guess this is not relevant for my setup? /Saaby