From: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
To: Jens Axboe <axboe@kernel.dk>,
Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-fsdevel <linux-fsdevel@vger.kernel.org>,
Linux-MM <linux-mm@kvack.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Michal Hocko <mhocko@suse.com>,
Dave Chinner <david@fromorbit.com>, Mel Gorman <mgorman@suse.de>,
Johannes Weiner <hannes@cmpxchg.org>, Tejun Heo <tj@kernel.org>
Subject: Re: [PATCH v2] mm: implement write-behind policy for sequential file writes
Date: Mon, 23 Sep 2019 19:05:55 +0300 [thread overview]
Message-ID: <a5e769db-823d-9a48-b5f9-a4b406710c6f@yandex-team.ru> (raw)
In-Reply-To: <1882a6da-a599-b820-6257-11bbac02b220@kernel.dk>
On 23/09/2019 18.36, Jens Axboe wrote:
> On 9/20/19 5:10 PM, Linus Torvalds wrote:
>> On Fri, Sep 20, 2019 at 4:05 PM Linus Torvalds
>> <torvalds@linux-foundation.org> wrote:
>>>
>>>
>>> Now, I hear you say "those are so small these days that it doesn't
>>> matter". And maybe you're right. But particularly for slow media,
>>> triggering good streaming write behavior has been a problem in the
>>> past.
>>
>> Which reminds me: the writebehind trigger should likely be tied to the
>> estimate of the bdi write speed.
>>
>> We _do_ have that avg_write_bandwidth thing in the bdi_writeback
>> structure, it sounds like a potentially good idea to try to use that
>> to estimate when to do writebehind.
>>
>> No?
>
> I really like the feature, and agree it should be tied to the bdi write
> speed. How about just making the tunable acceptable time of write behind
> dirty? Eg if write_behind_msec is 1000, allow 1s of pending dirty before
> starting writbeack.
>
I haven't digged into it yet.
But IIRR writeback speed estimation has some problems:
There is no "slow start" - initial speed is 100MiB/s.
This is especially bad for slow usb disks - right after plugging
we'll accumulate too much dirty cache before starting writeback.
And I've seen problems with cgroup-writeback:
each cgroup has own estimation, doesn't work well for short-living cgroups.
next prev parent reply other threads:[~2019-09-23 16:06 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-20 7:35 [PATCH v2] mm: implement write-behind policy for sequential file writes Konstantin Khlebnikov
2019-09-20 7:39 ` Konstantin Khlebnikov
2019-09-23 14:52 ` Tejun Heo
2019-09-23 15:06 ` Konstantin Khlebnikov
2019-09-23 15:19 ` Tejun Heo
2019-09-24 7:39 ` Dave Chinner
2019-09-24 9:00 ` Konstantin Khlebnikov
2019-09-25 7:18 ` Dave Chinner
2019-09-25 8:15 ` Konstantin Khlebnikov
2019-09-25 23:25 ` Dave Chinner
2019-09-25 12:54 ` Theodore Y. Ts'o
2019-09-24 19:08 ` Linus Torvalds
2019-09-25 8:00 ` Dave Chinner
2019-09-20 23:05 ` Linus Torvalds
2019-09-20 23:10 ` Linus Torvalds
2019-09-23 15:36 ` Jens Axboe
2019-09-23 16:05 ` Konstantin Khlebnikov [this message]
2019-09-24 9:29 ` Konstantin Khlebnikov
2019-09-22 7:47 ` kbuild test robot
2019-09-23 0:36 ` [mm] e0e7df8d5b: will-it-scale.per_process_ops -7.3% regression kernel test robot
2019-09-23 19:11 ` Konstantin Khlebnikov
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=a5e769db-823d-9a48-b5f9-a4b406710c6f@yandex-team.ru \
--to=khlebnikov@yandex-team.ru \
--cc=axboe@kernel.dk \
--cc=david@fromorbit.com \
--cc=hannes@cmpxchg.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=mhocko@suse.com \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
/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;
as well as URLs for NNTP newsgroup(s).