From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265531AbUBBAEm (ORCPT ); Sun, 1 Feb 2004 19:04:42 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265538AbUBBAEm (ORCPT ); Sun, 1 Feb 2004 19:04:42 -0500 Received: from websrv.werbeagentur-aufwind.de ([213.239.197.241]:39847 "EHLO mail.werbeagentur-aufwind.de") by vger.kernel.org with ESMTP id S265531AbUBBAEl (ORCPT ); Sun, 1 Feb 2004 19:04:41 -0500 Subject: load average calculation in linux 2.6 From: Christophe Saout To: linux-kernel@vger.kernel.org Content-Type: text/plain Message-Id: <1075680277.8818.38.camel@leto.cs.pocnet.net> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 Date: Mon, 02 Feb 2004 01:04:37 +0100 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I noticed my web server reboot while doing an rsync because the load was too high (and my watchdog decided that something was wrong...). The CPU was mostly idle but /proc/loadavg showed a load average of over 40. I assume that the iowait is also used for the load average calculation. Every process that's waiting for IO contributes a load average of one. Right? So rsync kept the disk busy and a lot of mail and apache processes were waiting for it. Is this the desired behavior or wrong?