From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757322Ab2FFQQN (ORCPT ); Wed, 6 Jun 2012 12:16:13 -0400 Received: from mx1.redhat.com ([209.132.183.28]:64979 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754155Ab2FFQQM (ORCPT ); Wed, 6 Jun 2012 12:16:12 -0400 Date: Wed, 6 Jun 2012 12:15:44 -0400 From: Vivek Goyal To: Linus Torvalds Cc: Fengguang Wu , LKML , "Myklebust, Trond" , linux-fsdevel@vger.kernel.org, Linux Memory Management List , Jens Axboe Subject: Re: write-behind on streaming writes Message-ID: <20120606161544.GA8133@redhat.com> References: <20120529155759.GA11326@localhost> <20120530032129.GA7479@localhost> <20120605172302.GB28556@redhat.com> <20120605174157.GC28556@redhat.com> <20120605184853.GD28556@redhat.com> <20120605201045.GE28556@redhat.com> <20120606025729.GA1197@redhat.com> <20120606121408.GB4934@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20120606121408.GB4934@redhat.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jun 06, 2012 at 08:14:08AM -0400, Vivek Goyal wrote: [..] > I think it is happening because sync_file_range() will send all > the writes as SYNC and it will compete with firefox IO. On the other > hand, flusher's IO will show up as ASYNC and CFQ will be penalize it > heavily and firefox's IO will be prioritized. And this effect should > just get worse as more processes do sync_file_range(). Ok, this time I tried the same test again but with 4 processes doing writes in parallel on 4 different files. And with sync_file_range() things turned ugly. Interactivity was very poor. firefox launch test took around 1m:45s with sync_file range() while it took only about 35seconds with regular flusher threads. So sending writeback IO synchronously wreaks havoc. Thanks Vivek