From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261639AbULBOmS (ORCPT ); Thu, 2 Dec 2004 09:42:18 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261644AbULBOmS (ORCPT ); Thu, 2 Dec 2004 09:42:18 -0500 Received: from ns.virtualhost.dk ([195.184.98.160]:18401 "EHLO virtualhost.dk") by vger.kernel.org with ESMTP id S261639AbULBOmP (ORCPT ); Thu, 2 Dec 2004 09:42:15 -0500 Date: Thu, 2 Dec 2004 15:41:34 +0100 From: Jens Axboe To: Giuliano Pochini Cc: Linux Kernel Subject: Re: Time sliced CFQ io scheduler Message-ID: <20041202144129.GI10458@suse.de> References: <20041202130457.GC10458@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Dec 02 2004, Giuliano Pochini wrote: > > > On Thu, 2 Dec 2004, Jens Axboe wrote: > > > Case 4: write_files, random, bs=4k > > Just a thought... in this test the results don't look right. Why > aggregate bandwidth with 8 clients is higher than with 4 and 2 clients ? > In the cfq test with 8 clients aggregate bw is also higher than with > a single client. I don't know what happens with the 4 client case, but it's not that unlikely that aggregate bandwidth will be higher for more threads doing random writes, as request coalesching will help minimize seeks. But I did think it was strange with the 4 client case dip was strange, it was reproducable though (as are all the results, they have very little variance). -- Jens Axboe