Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752470AbZIICK3 (ORCPT ); Tue, 8 Sep 2009 22:10:29 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751304AbZIICK2 (ORCPT ); Tue, 8 Sep 2009 22:10:28 -0400 Received: from bld-mail16.adl2.internode.on.net ([150.101.137.101]:51440 "EHLO mail.internode.on.net" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751192AbZIICK2 (ORCPT ); Tue, 8 Sep 2009 22:10:28 -0400 X-Greylist: delayed 916 seconds by postgrey-1.27 at vger.kernel.org; Tue, 08 Sep 2009 22:10:27 EDT Date: Wed, 9 Sep 2009 11:53:59 +1000 From: Dave Chinner To: Peter Zijlstra Cc: Chris Mason , Artem Bityutskiy , Jens Axboe , linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, hch@infradead.org, akpm@linux-foundation.org, jack@suse.cz, "Theodore Ts'o" , Wu Fengguang Subject: Re: [PATCH 8/8] vm: Add an tuning knob for vm.max_writeback_mb Message-ID: <20090909015359.GC7146@discord.disaster> References: <1252401791-22463-1-git-send-email-jens.axboe@oracle.com> <1252401791-22463-9-git-send-email-jens.axboe@oracle.com> <4AA633FD.3080006@gmail.com> <1252425983.7746.120.camel@twins> <20090908162936.GA2975@think> <1252428983.7746.140.camel@twins> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1252428983.7746.140.camel@twins> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1180 Lines: 28 On Tue, Sep 08, 2009 at 06:56:23PM +0200, Peter Zijlstra wrote: > On Tue, 2009-09-08 at 12:29 -0400, Chris Mason wrote: > > Either way, if pdflush or the bdi thread or whoever ends up switching to > > another file during a big streaming write, the end result is that we > > fragment. We may fragment the file (ext4) or we may fragment the > > writeback (xfs), but the end result isn't good. > > OK, so what we want is for a way to re-enter the whole > writeback_inodes() path onto the same file, right? No, that would take use back to the Bad Old Days where one large file write can starve out the other 10,000 small files that need to be written. The old writeback code used to end up in this way because it didn't rotate large files to the back of the dirty inode queue once wbc->nr_to_write was exhausted. This could cause files not to be written back for tens of minutes.... Cheers, Dave. -- Dave Chinner david@fromorbit.com -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/