From mboxrd@z Thu Jan 1 00:00:00 1970 From: Emmanuel Florac Subject: Re: Multi threaded random IO Date: Fri, 17 Jul 2009 13:10:43 +0200 Message-ID: <20090717131043.19b8e7de@harpe.intellique.com> References: <4A60442E.7010206@gslab.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: xfs mailing list , linux-fsdevel@vger.kernel.org To: Milind Dumbare Return-path: Received: from smtp2-g21.free.fr ([212.27.42.2]:36872 "EHLO smtp2-g21.free.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934411AbZGQLKv convert rfc822-to-8bit (ORCPT ); Fri, 17 Jul 2009 07:10:51 -0400 In-Reply-To: <4A60442E.7010206@gslab.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Le Fri, 17 Jul 2009 02:28:14 -0700 Milind Dumbare =E9crivait: > I see multi threaded random Write faster than multi threaded random=20 > Read. Which is contradictory to standard "Reads faster that Writes" > law. >=20 > My test setup is. > 10MB record size, 2TB filesize, 8 threads.iozone >=20 > Any comments? Well it very much depends upon your underlying hardware. For instance 3Ware RAID cards with default settings read very poorly, they even get much slower in sequential read than sequential write. So basically you have to provide much more detailed hardware setup : storage hardware, attachment, number and type of drives, RAID level... --=20 ---------------------------------------- Emmanuel Florac | Intellique ---------------------------------------- -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel= " in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html