From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from 173-166-109-252-newengland.hfc.comcastbusiness.net ([173.166.109.252]:34241 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753270Ab3ERGxv (ORCPT ); Sat, 18 May 2013 02:53:51 -0400 Date: Sat, 18 May 2013 08:53:41 +0200 From: Jens Axboe Subject: Re: output - aggregate read bandwitdh Message-ID: <20130518065341.GJ19781@kernel.dk> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: fio-owner@vger.kernel.org List-Id: fio@vger.kernel.org To: Edoardo Comar Cc: fio@vger.kernel.org On Sat, May 18 2013, Edoardo Comar wrote: > I have a problem with fio built today from GIT > running the jobfile below, the aggregate bandwidth for 'read' output by > the latest fio is excessive > while it seems reasonable with an older fio version. > > fio-2.0.13 READ: io=51500KB, aggrb=857KB/s, > > fio-2.1-6-g4d01 READ: io=31057MB, aggrb=530017KB/s, > > > You should easily reproduce this with the data below. > I have reproduced it on my laptop (RHEL 6.4) as well as on other RHEL > systems. > on my laptop > READ: io=13028KB, aggrb=216KB/s, > or > READ: io=33204MB, aggrb=566519KB/s, > > running the same configuration > > I believe this to be a bug. I wonder if this is a regression from 2.0.15 to 2.1. The gfio stuff got merged in along the way, could definitely be an accounting issue introduced there. I'll take a look. -- Jens Axboe