From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: Playing with SATA NCQ Date: Mon, 30 May 2005 08:05:57 +0200 Message-ID: <20050530060556.GC7054@suse.de> References: <20050526140058.GR1419@suse.de> <429793C8.8090007@gmail.com> <42979C4F.8020007@pobox.com> <42979FA3.1010106@gmail.com> <20050528121258.GA17869@suse.de> <4299BD23.6010004@gmail.com> <20050529190259.GA29770@suse.de> <429A2238.8010604@gmail.com> <429A236D.8030307@pobox.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from brick.kernel.dk ([62.242.22.158]:54659 "EHLO nelson.home.kernel.dk") by vger.kernel.org with ESMTP id S261536AbVE3GE6 (ORCPT ); Mon, 30 May 2005 02:04:58 -0400 Content-Disposition: inline In-Reply-To: <429A236D.8030307@pobox.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Jeff Garzik Cc: Michael Thonke , linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org On Sun, May 29 2005, Jeff Garzik wrote: > Michael Thonke wrote: > >Well the subjective feeling is great through! What I noticed is a > >improvement on rsync (goes slighty faster drive to drive). > >The noise decrease now it only make noise on very heavy IO reads and > >writes. > > > It might be as simple as... NCQ means the drive is doing more work. > More work means more noise. You can argue both ways :-). Usually more noise means more seeks and thus less work done. -- Jens Axboe