From: Chris Lee Subject: Re: [Bug 14830] When other IO is running sync times go to 10 to 20 minutes Date: Tue, 19 Jan 2010 17:24:23 +0000 Message-ID: <4B55EAC7.3080004@cybericom.co.uk> References: <201001182358.o0INwBkc011745@demeter.kernel.org> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Cc: linux-ext4@vger.kernel.org To: bugzilla-daemon@bugzilla.kernel.org Return-path: Received: from mk-outboundfilter-4.mail.uk.tiscali.com ([212.74.114.32]:36206 "EHLO mk-outboundfilter-4.mail.uk.tiscali.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752881Ab0ASRYk (ORCPT ); Tue, 19 Jan 2010 12:24:40 -0500 In-Reply-To: <201001182358.o0INwBkc011745@demeter.kernel.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: bugzilla-daemon@bugzilla.kernel.org wrote: > http://bugzilla.kernel.org/show_bug.cgi?id=14830 > > > > > > --- Comment #3 from Michael Godfrey 2010-01-18 23:58:09 --- > >> This problem prevents production use of systems using this kernel. >> > > >> evokes a question: Do you have a kernel which behaved better for you? Which >> one? >> > > Yes. RHEL5.4 does not show this problem. It is the production > system that works in this environment. > > The response above is disappointing. Is sync response of 20 minutes, > including several task timeouts to be considered "normal?" > > >> If you think the time is inappropriately long, we can have a look at it >> but for that we'd need much more details like amount and nature of data writen >> (many small files vs a few large ones), time it takes sync to complete, speed >> of disks for sequential IO... >> > > I am sorry to have to tell you that in this environment we do not > deal in exclusively small or large files, we actually have quite a > few of both. When an rsync which transfers about 50GB of files of > various sizes is running, the hung condition is continuous until the rsync > completes. This is just a pretty typical load. You could try it > yourself. No special sizes of files are required. I think I > mentioned that the ext4 LVM is a RAID 50 3ware 9650SE-8LPML, > with 8 2T drives. Its throughput for reading and writing is good > when the system is not locked up. > > Is it possible that it is something allong the lines of what is described at this link: http://notemagnet.blogspot.com/2008/08/linux-write-cache-mystery.html If so a runtime adjustment might help you out. Chris.