From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 18 Jan 2001 20:41:14 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 18 Jan 2001 20:41:04 -0500 Received: from ns.virtualhost.dk ([195.184.98.160]:2319 "EHLO virtualhost.dk") by vger.kernel.org with ESMTP id ; Thu, 18 Jan 2001 20:40:49 -0500 Date: Fri, 19 Jan 2001 02:40:23 +0100 From: Jens Axboe To: Andrea Arcangeli Cc: Marcelo Tosatti , linux-kernel@vger.kernel.org Subject: Re: 2.4.1pre8 slowdown on dbench tests Message-ID: <20010119024023.B18209@suse.de> In-Reply-To: <20010119011629.C32087@athlon.random> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20010119011629.C32087@athlon.random>; from andrea@suse.de on Fri, Jan 19, 2001 at 01:16:29AM +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jan 19 2001, Andrea Arcangeli wrote: > On Thu, Jan 18, 2001 at 03:17:13PM -0200, Marcelo Tosatti wrote: > > Jens, can be the -blk patch the reason for the slowdown I'm seeing? > > This heuristic is way too aggressive: > > /* > * Try to keep 128MB max hysteris. If not possible, > * use half of RAM > */ > high_queued_sectors = (total_ram * 2) / 3; > low_queued_sectors = high_queued_sectors - MB(128); > if (low_queued_sectors < 0) > low_queued_sectors = total_ram / 2; > > /* > * for big RAM machines (>= 384MB), use more for I/O > */ > if (total_ram >= MB(384)) { > high_queued_sectors = (total_ram * 4) / 5; > low_queued_sectors = high_queued_sectors - MB(128); > } > > 2/3 of ram locked down in the I/O queue is way too much. 1/3 should be > ok. big RAM machines needs way less than 1/3 locked down. Yes I agree, that values should probably be tweaked a bit. I'll try and squeeze some testing in to generate the best possible values. -- * Jens Axboe * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/