Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1762336AbZATO0k (ORCPT ); Tue, 20 Jan 2009 09:26:40 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755294AbZATO0a (ORCPT ); Tue, 20 Jan 2009 09:26:30 -0500 Received: from brick.kernel.dk ([93.163.65.50]:21631 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754447AbZATO03 (ORCPT ); Tue, 20 Jan 2009 09:26:29 -0500 Date: Tue, 20 Jan 2009 15:24:57 +0100 From: Jens Axboe To: Mathieu Desnoyers Cc: akpm@linux-foundation.org, ltt-dev@lists.casi.polymtl.ca, Linus Torvalds , Ingo Molnar , linux-kernel@vger.kernel.org Subject: Re: [ltt-dev] [RFC PATCH] block: Fix bio merge induced high I/O latency Message-ID: <20090120142457.GO30821@kernel.dk> References: <20090117004439.GA11492@Krystal> <20090117162657.GA31965@Krystal> <20090117190437.GZ30821@kernel.dk> <20090118211234.GA4913@Krystal> <20090119182654.GT30821@kernel.dk> <20090120021055.GA6990@Krystal> <20090120073709.GC30821@kernel.dk> <20090120122855.GF30821@kernel.dk> <20090120142223.GD22421@Krystal> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090120142223.GD22421@Krystal> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 7957 Lines: 202 On Tue, Jan 20 2009, Mathieu Desnoyers wrote: > * Jens Axboe (jens.axboe@oracle.com) wrote: > > On Tue, Jan 20 2009, Jens Axboe wrote: > > > On Mon, Jan 19 2009, Mathieu Desnoyers wrote: > > > > * Jens Axboe (jens.axboe@oracle.com) wrote: > > > > > On Sun, Jan 18 2009, Mathieu Desnoyers wrote: > > > > > > I looked at the "ls" behavior (while doing a dd) within my LTTng trace > > > > > > to create a fio job file. The said behavior is appended below as "Part > > > > > > 1 - ls I/O behavior". Note that the original "ls" test case was done > > > > > > with the anticipatory I/O scheduler, which was active by default on my > > > > > > debian system with custom vanilla 2.6.28 kernel. Also note that I am > > > > > > running this on a raid-1, but have experienced the same problem on a > > > > > > standard partition I created on the same machine. > > > > > > > > > > > > I created the fio job file appended as "Part 2 - dd+ls fio job file". It > > > > > > consists of one dd-like job and many small jobs reading as many data as > > > > > > ls did. I used the small test script to batch run this ("Part 3 - batch > > > > > > test"). > > > > > > > > > > > > The results for the ls-like jobs are interesting : > > > > > > > > > > > > I/O scheduler runt-min (msec) runt-max (msec) > > > > > > noop 41 10563 > > > > > > anticipatory 63 8185 > > > > > > deadline 52 33387 > > > > > > cfq 43 1420 > > > > > > > > > > > > > Extra note : I have a HZ=250 on my system. Changing to 100 or 1000 did > > > > not make much difference (also tried with NO_HZ enabled). > > > > > > > > > Do you have queuing enabled on your drives? You can check that in > > > > > /sys/block/sdX/device/queue_depth. Try setting those to 1 and retest all > > > > > schedulers, would be good for comparison. > > > > > > > > > > > > > Here are the tests with a queue_depth of 1 : > > > > > > > > I/O scheduler runt-min (msec) runt-max (msec) > > > > noop 43 38235 > > > > anticipatory 44 8728 > > > > deadline 51 19751 > > > > cfq 48 427 > > > > > > > > > > > > Overall, I wouldn't say it makes much difference. > > > > > > 0,5 seconds vs 1,5 seconds isn't much of a difference? > > > > > > > > raid personalities or dm complicates matters, since it introduces a > > > > > disconnect between 'ls' and the io scheduler at the bottom... > > > > > > > > > > > > > Yes, ideally I should re-run those directly on the disk partitions. > > > > > > At least for comparison. > > > > > > > I am also tempted to create a fio job file which acts like a ssh server > > > > receiving a connexion after it has been pruned from the cache while the > > > > system if doing heavy I/O. "ssh", in this case, seems to be doing much > > > > more I/O than a simple "ls", and I think we might want to see if cfq > > > > behaves correctly in such case. Most of this I/O is coming from page > > > > faults (identified as traps in the trace) probably because the ssh > > > > executable has been thrown out of the cache by > > > > > > > > echo 3 > /proc/sys/vm/drop_caches > > > > > > > > The behavior of an incoming ssh connexion after clearing the cache is > > > > appended below (Part 1 - LTTng trace for incoming ssh connexion). The > > > > job file created (Part 2) reads, for each job, a 2MB file with random > > > > reads each between 4k-44k. The results are very interesting for cfq : > > > > > > > > I/O scheduler runt-min (msec) runt-max (msec) > > > > noop 586 110242 > > > > anticipatory 531 26942 > > > > deadline 561 108772 > > > > cfq 523 28216 > > > > > > > > So, basically, ssh being out of the cache can take 28s to answer an > > > > incoming ssh connexion even with the cfq scheduler. This is not exactly > > > > what I would call an acceptable latency. > > > > > > At some point, you have to stop and consider what is acceptable > > > performance for a given IO pattern. Your ssh test case is purely random > > > IO, and neither CFQ nor AS would do any idling for that. We can make > > > this test case faster for sure, the hard part is making sure that we > > > don't regress on async throughput at the same time. > > > > > > Also remember that with your raid1, it's not entirely reasonable to > > > blaim all performance issues on the IO scheduler as per my previous > > > mail. It would be a lot more fair to view the disk numbers individually. > > > > > > Can you retry this job with 'quantum' set to 1 and 'slice_async_rq' set > > > to 1 as well? > > > > > > However, I think we should be doing somewhat better at this test case. > > > > Mathieu, does this improve anything for you? > > > > I got this message when running with your patch applied : > cfq: forced dispatching is broken (nr_sorted=4294967275), please report this > (message appeared 10 times in a job run) Woops, missed a sort inc. Updated version below, or just ignore the warning. > Here is the result : > > ssh test done on /dev/sda directly > > queue_depth=31 (default) > /sys/block/sda/queue/iosched/slice_async_rq = 2 (default) > /sys/block/sda/queue/iosched/quantum = 4 (default) > > I/O scheduler runt-min (msec) runt-max (msec) > cfq (default) 523 6637 > cfq (patched) 564 7195 > > Pretty much the same. Can you retry with depth=1 as well? There's not much to rip back out, if everything is immediately sent to the device. > > Here is the test done on raid1 : > queue_depth=31 (default) > /sys/block/sd{a,b}/queue/iosched/slice_async_rq = 2 (default) > /sys/block/sd{a,b}/queue/iosched/quantum = 4 (default) > > I/O scheduler runt-min (msec) runt-max (msec) > cfq (default, raid1) 523 28216 > cfq (patched, raid1) 540 16454 > > With nearly same order of magnitude worse-case. diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index e8525fa..30714de 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1765,6 +1765,36 @@ cfq_update_idle_window(struct cfq_data *cfqd, struct cfq_queue *cfqq, } /* + * Pull dispatched requests from 'cfqq' back into the scheduler + */ +static void cfq_pull_dispatched_requests(struct cfq_data *cfqd, + struct cfq_queue *cfqq) +{ + struct request_queue *q = cfqd->queue; + struct request *rq; + + list_for_each_entry_reverse(rq, &q->queue_head, queuelist) { + if (rq->cmd_flags & REQ_STARTED) + break; + + if (RQ_CFQQ(rq) != cfqq) + continue; + + /* + * Pull off the dispatch list and put it back into the cfqq + */ + list_del(&rq->queuelist); + cfqq->dispatched--; + if (cfq_cfqq_sync(cfqq)) + cfqd->sync_flight--; + + cfq_add_rq_rb(rq); + q->nr_sorted++; + list_add_tail(&rq->queuelist, &cfqq->fifo); + } +} + +/* * Check if new_cfqq should preempt the currently active queue. Return 0 for * no or if we aren't sure, a 1 will cause a preempt. */ @@ -1820,8 +1850,14 @@ cfq_should_preempt(struct cfq_data *cfqd, struct cfq_queue *new_cfqq, */ static void cfq_preempt_queue(struct cfq_data *cfqd, struct cfq_queue *cfqq) { + struct cfq_queue *old_cfqq = cfqd->active_queue; + cfq_log_cfqq(cfqd, cfqq, "preempt"); - cfq_slice_expired(cfqd, 1); + + if (old_cfqq) { + __cfq_slice_expired(cfqd, old_cfqq, 1); + cfq_pull_dispatched_requests(cfqd, old_cfqq); + } /* * Put the new queue at the front of the of the current list, -- Jens Axboe -- 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/