From: Bart Van Assche <bvanassche@acm.org>
To: Mike Snitzer <snitzer@redhat.com>
Cc: Jens Axboe <axboe@kernel.dk>,
Christoph Hellwig <hch@infradead.org>,
Keith Busch <keith.busch@intel.com>,
device-mapper development <dm-devel@redhat.com>,
Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Subject: Re: [PATCH v3 0/8] dm: add request-based blk-mq support
Date: Tue, 06 Jan 2015 10:31:44 +0100 [thread overview]
Message-ID: <54ABAB80.70006@acm.org> (raw)
In-Reply-To: <20150105213557.GA5030@redhat.com>
On 01/05/15 22:35, Mike Snitzer wrote:
> On Fri, Jan 02 2015 at 12:53pm -0500,
> Bart Van Assche <bvanassche@acm.org> wrote:
>> Thanks, my tests confirm that this patch indeed fixes the issue I had
>> reported. Unfortunately this doesn't mean that the blk-mq multipath code
>> is already working perfectly. Most of the time I/O requests are
>> processed within the expected time but sometimes I/O processing takes
>> much more time than what I expected:
>>
>> # /usr/bin/time -f %e mkfs.xfs -f /dev/dm-0 >/dev/null
>> 0.02
>> # /usr/bin/time -f %e mkfs.xfs -f /dev/dm-0 >/dev/null
>> 0.02
>> # /usr/bin/time -f %e mkfs.xfs -f /dev/dm-0 >/dev/null
>> 8.68
>>
>> However, if I run the same command on the underlying device it always
>> completes within the expected time.
>
> I don't have very large blk-mq devices, but I can work on that.
> How large is the blk-mq device in question?
>
> Also, how much memory does the system have? Is memory fragmented at
> all? With this change the requests are cloned using memory allocated
> from block core's blk_get_request (rather than a dedicated mempool in DM
> core).
>
> Any chance you could use 'perf record' to try to analyze where the
> kernel is spending its time?
Hello Mike,
The device used in this test was a tmpfs file with a size of 16 MB. That
file had been created as follows: dd if=/dev/zero of=/dev/vdisk bs=1M
count=16. The initiator and target systems did have enough memory to keep
this tmpfs file in RAM all the time (32 GB and 4 GB respectively).
For the runs that took much longer than expected the CPU load was low.
This probably means that the system was waiting for one or another I/O
timer to expire. The output triggered by "echo w > /proc/sysrq-trigger"
during a run that took longer than expected was as follows:
SysRq : Show Blocked State
task PC stack pid father
kdmwork-253:0 D ffff8807c1fd3b78 0 10396 2 0x00000000
ffff8807c1fd3b78 ffff88083b6b6cc0 0000000000012ec0 ffff8807c1fd3fd8
0000000000012ec0 ffff880824225aa0 ffff88083b6b6cc0 ffff88081b0cb2c0
ffff88085fc537c8 ffff8807c1fd3c98 ffff8807f7a99d70 ffffe8ffffc43bc0
Call Trace:
[<ffffffff814d5230>] io_schedule+0xa0/0x130
[<ffffffff8125a3f7>] bt_get+0x117/0x1b0
[<ffffffff81256580>] ? blk_mq_queue_enter+0x30/0x2a0
[<ffffffff81094cf0>] ? prepare_to_wait_event+0x110/0x110
[<ffffffff8125a76f>] blk_mq_get_tag+0x9f/0xd0
[<ffffffff8125591b>] __blk_mq_alloc_request+0x1b/0x210
[<ffffffff812571c9>] blk_mq_alloc_request+0x139/0x150
[<ffffffff8124c16e>] blk_get_request+0x2e/0xe0
[<ffffffff8109a60d>] ? trace_hardirqs_on+0xd/0x10
[<ffffffffa07f7d0f>] __multipath_map.isra.15+0x1cf/0x210 [dm_multipath]
[<ffffffffa07f7d6a>] multipath_clone_and_map+0x1a/0x20 [dm_multipath]
[<ffffffffa039dbb5>] map_tio_request+0x1d5/0x3a0 [dm_mod]
[<ffffffff8109a53d>] ? trace_hardirqs_on_caller+0xfd/0x1c0
[<ffffffff81075cbe>] kthread_worker_fn+0x7e/0x1b0
[<ffffffff81075c40>] ? __init_kthread_worker+0x60/0x60
[<ffffffff81075bc8>] kthread+0xf8/0x110
[<ffffffff81075ad0>] ? kthread_create_on_node+0x210/0x210
[<ffffffff814dacac>] ret_from_fork+0x7c/0xb0
[<ffffffff81075ad0>] ? kthread_create_on_node+0x210/0x210
dmraid D ffff8807f4cafc88 0 25099 25064 0x00000000
ffff8807f4cafc88 ffff8807c0b52440 0000000000012ec0 ffff8807f4caffd8
0000000000012ec0 ffffffff81a194e0 ffff8807c0b52440 ffff8807c09ec1c0
ffff88085fc137c8 ffff88085ff8ce38 ffff8807f4cafd30 0000000000000082
Call Trace:
[<ffffffff814d5990>] ? bit_wait+0x50/0x50
[<ffffffff814d5230>] io_schedule+0xa0/0x130
[<ffffffff814d59bc>] bit_wait_io+0x2c/0x50
[<ffffffff814d578b>] __wait_on_bit_lock+0x4b/0xb0
[<ffffffff8113b45a>] __lock_page_killable+0x9a/0xa0
[<ffffffff81094d30>] ? autoremove_wake_function+0x40/0x40
[<ffffffff8113da78>] generic_file_read_iter+0x408/0x640
[<ffffffff8109a60d>] ? trace_hardirqs_on+0xd/0x10
[<ffffffff811d5f57>] blkdev_read_iter+0x37/0x40
[<ffffffff8119866e>] new_sync_read+0x7e/0xb0
[<ffffffff81199858>] __vfs_read+0x18/0x50
[<ffffffff81199916>] vfs_read+0x86/0x140
[<ffffffff81199a19>] SyS_read+0x49/0xb0
[<ffffffff814dad52>] system_call_fastpath+0x12/0x17
Bart.
next prev parent reply other threads:[~2015-01-06 9:31 UTC|newest]
Thread overview: 94+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-12-17 3:59 [PATCH v3 0/8] dm: add request-based blk-mq support Mike Snitzer
2014-12-17 3:59 ` [PATCH v3 1/8] block: require blk_rq_prep_clone() be given an initialized clone request Mike Snitzer
2014-12-17 3:59 ` [PATCH v3 2/8] block: initialize bio member of blk-mq request to NULL Mike Snitzer
2014-12-17 3:59 ` [PATCH v3 3/8] block: add blk-mq support to blk_insert_cloned_request() Mike Snitzer
2014-12-17 4:00 ` [PATCH v3 4/8] block: mark blk-mq devices as stackable Mike Snitzer
2014-12-17 4:00 ` [PATCH v3 5/8] dm: remove exports for request-based interfaces without external callers Mike Snitzer
2014-12-17 4:00 ` [PATCH v3 6/8] dm: split request structure out from dm_rq_target_io structure Mike Snitzer
2014-12-17 4:00 ` [PATCH v3 7/8] dm: submit stacked requests in irq enabled context Mike Snitzer
2014-12-17 4:00 ` [PATCH v3 8/8] dm: allocate requests from target when stacking on blk-mq devices Mike Snitzer
2014-12-17 22:35 ` Mike Snitzer
2014-12-17 21:42 ` [PATCH v3 0/8] dm: add request-based blk-mq support Keith Busch
2014-12-17 21:43 ` Jens Axboe
2014-12-17 23:06 ` Mike Snitzer
2014-12-18 1:41 ` Keith Busch
2014-12-18 4:58 ` Mike Snitzer
2014-12-19 14:32 ` Bart Van Assche
2014-12-19 15:38 ` Mike Snitzer
2014-12-19 17:14 ` Mike Snitzer
2014-12-22 15:28 ` Bart Van Assche
2014-12-22 18:49 ` Mike Snitzer
2014-12-23 16:24 ` Bart Van Assche
2014-12-23 17:13 ` Mike Snitzer
2014-12-23 21:42 ` Mike Snitzer
2014-12-24 13:02 ` Bart Van Assche
2014-12-24 18:21 ` Mike Snitzer
2014-12-24 18:55 ` Mike Snitzer
2014-12-24 19:26 ` Mike Snitzer
2015-01-02 17:53 ` Bart Van Assche
2015-01-05 21:35 ` Mike Snitzer
2015-01-06 8:59 ` Christoph Hellwig
2015-01-06 9:31 ` Bart Van Assche [this message]
2015-01-06 16:05 ` blk-mq request allocation stalls [was: Re: [PATCH v3 0/8] dm: add request-based blk-mq support] Mike Snitzer
2015-01-06 16:15 ` Jens Axboe
2015-01-07 10:33 ` Bart Van Assche
2015-01-07 15:32 ` Jens Axboe
2015-01-07 16:15 ` Mike Snitzer
2015-01-07 16:18 ` Jens Axboe
2015-01-07 16:22 ` Mike Snitzer
2015-01-07 16:24 ` Jens Axboe
2015-01-07 17:18 ` Mike Snitzer
2015-01-07 17:35 ` Jens Axboe
2015-01-07 20:09 ` Mike Snitzer
2015-01-07 20:40 ` Keith Busch
2015-01-09 19:49 ` Mike Snitzer
2015-01-09 21:07 ` Jens Axboe
2015-01-09 21:11 ` Jens Axboe
2015-01-09 21:40 ` Mike Snitzer
2015-01-09 21:56 ` Jens Axboe
2015-01-09 22:25 ` Mike Snitzer
2015-01-10 0:27 ` Jens Axboe
2015-01-10 1:48 ` Mike Snitzer
2015-01-10 1:59 ` Jens Axboe
2015-01-10 3:10 ` Mike Snitzer
2015-01-12 14:46 ` blk-mq request allocation stalls Bart Van Assche
2015-01-12 15:42 ` Jens Axboe
2015-01-12 16:12 ` Bart Van Assche
2015-01-12 16:34 ` Jens Axboe
2015-01-12 16:58 ` Mike Snitzer
2015-01-12 16:59 ` Jens Axboe
2015-01-12 17:04 ` Bart Van Assche
2015-01-12 17:09 ` Jens Axboe
2015-01-12 17:53 ` Keith Busch
2015-01-12 18:12 ` Jens Axboe
2015-01-12 18:22 ` Keith Busch
2015-01-12 18:35 ` Keith Busch
2015-01-12 19:11 ` Mike Snitzer
2015-01-12 20:21 ` Mike Snitzer
2015-01-13 12:29 ` Bart Van Assche
2015-01-13 14:17 ` Mike Snitzer
2015-01-13 14:28 ` dm + blk-mq soft lockup complaint Bart Van Assche
2015-01-13 16:20 ` Mike Snitzer
2015-01-14 9:16 ` Bart Van Assche
2015-01-14 18:59 ` Mike Snitzer
2015-01-15 8:11 ` Bart Van Assche
2015-01-15 15:43 ` Mike Snitzer
2015-01-15 15:55 ` Bart Van Assche
2015-01-13 14:59 ` blk-mq request allocation stalls Jens Axboe
2015-01-13 15:11 ` Keith Busch
2015-01-13 15:27 ` Keith Busch
2015-01-13 15:41 ` Mike Snitzer
2015-01-13 15:14 ` Mike Snitzer
2015-01-27 18:42 ` blk-mq DM changes for 3.20 [was: Re: blk-mq request allocation stalls] Mike Snitzer
2015-01-28 16:42 ` Jens Axboe
2015-01-28 17:44 ` Mike Snitzer
2015-01-28 17:49 ` Jens Axboe
2015-01-28 18:10 ` Mike Snitzer
2015-01-29 22:43 ` blk-mq DM changes for 3.20 [was: Re: blk-mq request allocation stalls]X Keith Busch
2015-01-29 23:09 ` Mike Snitzer
2015-01-29 23:44 ` Keith Busch
2015-01-30 0:32 ` Mike Snitzer
2015-01-12 19:05 ` blk-mq request allocation stalls Jens Axboe
2015-01-12 19:07 ` Mike Snitzer
2015-01-12 18:19 ` Mike Snitzer
2014-12-17 22:51 ` [PATCH v3 0/8] dm: add request-based blk-mq support Mike Snitzer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=54ABAB80.70006@acm.org \
--to=bvanassche@acm.org \
--cc=axboe@kernel.dk \
--cc=dm-devel@redhat.com \
--cc=hch@infradead.org \
--cc=j-nomura@ce.jp.nec.com \
--cc=keith.busch@intel.com \
--cc=snitzer@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox