From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Mon, 18 Feb 2002 23:58:00 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Mon, 18 Feb 2002 23:57:51 -0500 Received: from ebiederm.dsl.xmission.com ([166.70.28.69]:46928 "EHLO frodo.biederman.org") by vger.kernel.org with ESMTP id ; Mon, 18 Feb 2002 23:57:38 -0500 To: Andrew Morton Cc: john , linux-kernel@vger.kernel.org Subject: Re: kupdated using all CPU In-Reply-To: <20020218134041.A2586@doom.sfo.covalent.net> <3C717C72.72A994D3@zip.com.au> <3C717C72.72A994D3@zip.com.au> <20020218141920.D2586@doom.sfo.covalent.net> <3C71848A.3DA9BC42@zip.com.au> From: ebiederm@xmission.com (Eric W. Biederman) Date: 18 Feb 2002 21:53:08 -0700 In-Reply-To: <3C71848A.3DA9BC42@zip.com.au> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Andrew Morton writes: > john wrote: > > > > [Mon, Feb 18, 2002 at 02:13:06PM -0800] akpm@zip.com.au wrote: > > + john wrote: > > + > > > + > hi, > > + > ive searched all over and found many references to this problem, but > > + > never found an actual solution. the problem is that during heavy > > + > disk I/O, kupdated will periodically take up ALL the cpu. > > + > > + I've seen a couple of reports of this, nothing to indicate that it's > > + a common problem? > > + > > + In the other reports, it was related to extremely low disk throughput. > > + What does `hdparm -t /dev/hda' say? > > > > root@doom:~# hdparm -t /dev/hda > > > > /dev/hda: > > Timing buffered disk reads: 64 MB in 25.60 seconds = 2.50 MB/sec > > ugh. OK, I'll see if I cen reproduce this - there's no reason > why disk suckiness should cause high CPU load. But you need to > pay some attention to your IDE settings in kernel config, and > possibly tuning. Another possibility with the same symptoms is that there are simply very large I/O request starving everything else out. When some crucial data needs to be paged back in. john can you confirm you looked in top and saw kupdate taking 100% of the cpu? Eric