From owner-freebsd-questions@FreeBSD.ORG Mon May 9 19:50:21 2005 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 94F2A16A4EA for ; Mon, 9 May 2005 19:50:21 +0000 (GMT) Received: from dan.emsphone.com (dan.emsphone.com [199.67.51.101]) by mx1.FreeBSD.org (Postfix) with ESMTP id 17DA843D2D for ; Mon, 9 May 2005 19:50:21 +0000 (GMT) (envelope-from dan@dan.emsphone.com) Received: (from dan@localhost) by dan.emsphone.com (8.13.1/8.13.3) id j49JoE0N057362; Mon, 9 May 2005 14:50:14 -0500 (CDT) (envelope-from dan) Date: Mon, 9 May 2005 14:50:14 -0500 From: Dan Nelson To: Amandeep Message-ID: <20050509195011.GH38839@dan.emsphone.com> References: <427FA802.90805@chamkila.org> <427FA8CD.8040405@centtech.com> <427FAB5F.6070508@chamkila.org> <427FAF73.7070702@centtech.com> <20050509190245.GF38839@dan.emsphone.com> <427FB663.3090309@chamkila.org> <427FBD12.2090101@chamkila.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <427FBD12.2090101@chamkila.org> X-OS: FreeBSD 5.4-STABLE X-message-flag: Outlook Error User-Agent: Mutt/1.5.8i cc: anderson@centtech.com cc: freebsd-questions@freebsd.org cc: john@essenz.com Subject: Re: Low HDD tranfer rate with FreeBSD 5.3-Release X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 09 May 2005 19:50:21 -0000 In the last episode (May 09), Amandeep said: > Amandeep wrote: > >>>>>>The transfer rate is about 15MB/s > >>>>>> > >>>>>>when I run > >>>>>>#dd if=/dev/zero of=/usr/junk bs=8192 > > > >OK with the above dd > > > >dd if=/dev/ad0 of=/dev/null bs=64k > > > >it gives me 56MB/s. from the start. > > So how does the block size makes the difference?? What is the true > transfer rate.? At the moment, 56MB/sec :) Those two speeds aren't directly comparable, since your first was testing writes through the filesystem with a small blocksize, and the second was testing reads to the raw device with a large blocksize. Usually reading, using raw devices, and using large blocksizes are faster then writing, using a filesystem, and using small blocksizes, so all three changes probably contributed a little to the speed difference. -- Dan Nelson dnelson@allantgroup.com