From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kara Subject: Re: [RFC] ext3: per-process soft-syncing data=ordered mode Date: Mon, 28 Jan 2008 18:34:05 +0100 Message-ID: <20080128173405.GH14038@duck.suse.cz> References: <200801242336.00340.a1426z@gawab.com> <20080124225026.55fc6c20.diegocg@gmail.com> <200801260827.43284.a1426z@gawab.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Diego Calleja , linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, Jan Kara To: Al Boldi Return-path: Received: from styx.suse.cz ([82.119.242.94]:44244 "EHLO duck.suse.cz" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752120AbYA1ReG (ORCPT ); Mon, 28 Jan 2008 12:34:06 -0500 Content-Disposition: inline In-Reply-To: <200801260827.43284.a1426z@gawab.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Sat 26-01-08 08:27:43, Al Boldi wrote: > Diego Calleja wrote: > > El Thu, 24 Jan 2008 23:36:00 +0300, Al Boldi esc= ribi=F3: > > > Greetings! > > > > > > data=3Dordered mode has proven reliable over the years, and it do= es this > > > by ordering filedata flushes before metadata flushes. But this > > > sometimes causes contention in the order of a 10x slowdown for ce= rtain > > > apps, either due to the misuse of fsync or due to inherent behavi= our > > > like db's, as well as inherent starvation issues exposed by the > > > data=3Dordered mode. > > > > There's a related bug in bugzilla: > > http://bugzilla.kernel.org/show_bug.cgi?id=3D9546 > > > > The diagnostic from Jan Kara is different though, but I think it ma= y be > > the same problem... > > > > "One process does data-intensive load. Thus in the ordered mode the > > transaction is tiny but has tons of data buffers attached. If commi= t > > happens, it takes a long time to sync all the data before the commi= t > > can proceed... In the writeback mode, we don't wait for data buffer= s, in > > the journal mode amount of data to be written is really limited by = the > > maximum size of a transaction and so we write by much smaller chunk= s > > and better latency is thus ensured." > > > > > > I'm hitting this bug too...it's surprising that there's not many pe= ople > > reporting more bugs about this, because it's really annoying. > > > > > > There's a patch by Jan Kara (that I'm including here because bugzil= la > > didn't include it and took me a while to find it) which I don't kno= w if > > it's supposed to fix the problem , but it'd be interesting to try: >=20 > Thanks a lot, but it doesn't fix it. Hmm, if you're willing to test patches, then you could try a debug pa= tch: http://bugzilla.kernel.org/attachment.cgi?id=3D14574 and send me the output. What kind of load do you observe problems wit= h and which problems exactly? Honza --=20 Jan Kara SUSE Labs, CR - 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