From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758487Ab1LOBKR (ORCPT ); Wed, 14 Dec 2011 20:10:17 -0500 Received: from mga03.intel.com ([143.182.124.21]:12369 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754667Ab1LOBKO (ORCPT ); Wed, 14 Dec 2011 20:10:14 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.71,315,1320652800"; d="scan'208";a="85854318" Date: Thu, 15 Dec 2011 09:00:10 +0800 From: Wu Fengguang To: "Li, Shaohua" Cc: "Ted Ts'o" , "linux-ext4@vger.kernel.org" , Jan Kara , LKML , "linux-fsdevel@vger.kernel.org" , NeilBrown , linux-raid@vger.kernel.org, Jens Axboe Subject: Re: ext4 data=writeback performs worse than data=ordered now Message-ID: <20111215010010.GA14805@localhost> References: <20111214133400.GA18565@localhost> <20111214143014.GB18080@thunk.org> <1323910977.22361.423.camel@sli10-conroe> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1323910977.22361.423.camel@sli10-conroe> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > I found sometimes one disk hasn't any request inflight, but we can't > send request to the disk, because the scsi host's resource (the queue > depth) is used out, looks we send too many requests from other disks and > leave some disks starved. The resource imbalance in scsi isn't a new > problem, even 3.1 has such issue, so I'd think writeback introduces new > imbalance between the 12 disks. In fact, if I limit disk's queue depth > to 10, in this way the 12 disks will not impact each other in scsi > layer, the performance regression fully disappears for both writeback > and order mode. I observe similar issue in MD. The default q->nr_requests = BLKDEV_MAX_RQ; is too small for large arrays, and I end up doing echo 1280 > /sys/block/md0/queue/nr_requests in my tests. Thanks, Fengguang