From: Matthias Ferdinand <bcache@mfedv.net>
To: Eric Wheeler <bcache@lists.ewheeler.net>
Cc: Adriano Silva <adriano_da_silva@yahoo.com.br>,
Coly Li <colyli@suse.de>,
Bcache Linux <linux-bcache@vger.kernel.org>
Subject: Re: Bcache in writes direct with fsync. Are IOPS limited?
Date: Sat, 28 May 2022 09:22:30 +0200 [thread overview]
Message-ID: <YpHNts38gQMJspip@xoff> (raw)
In-Reply-To: <5b164113-364b-76a8-5bcc-94c1cec868db@ewheeler.net>
On Fri, May 27, 2022 at 06:27:53PM -0700, Eric Wheeler wrote:
> > I can say that the performance of tests after the write back command for
> > all devices greatly worsens the performance of direct tests on NVME
> > hardware. Below you can see this.
>
> I wonder what is going on there! I tried the same thing on my system and
> 'write through' is faster for me, too, so it would be worth investigating.
In Ceph context, it seems not unusual to disable SSD write back cache
and see much improved performance (or the other way round: see
surprisingly low performance with write back cache enabled):
https://yourcmc.ru/wiki/Ceph_performance#Drive_cache_is_slowing_you_down
Disk controllers seem to interpret FLUSH CACHE / FUA differently.
If bcache would set FUA for cache device writes while running fio
directly on the nvme device would not, that might explain the timing
difference.
Regards
Matthias
next prev parent reply other threads:[~2022-05-28 7:22 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <958894243.922478.1652201375900.ref@mail.yahoo.com>
2022-05-10 16:49 ` Bcache in writes direct with fsync. Are IOPS limited? Adriano Silva
2022-05-11 6:20 ` Matthias Ferdinand
2022-05-11 12:58 ` Adriano Silva
2022-05-11 21:21 ` Matthias Ferdinand
2022-05-18 1:22 ` Eric Wheeler
2022-05-23 14:07 ` Coly Li
2022-05-26 19:15 ` Eric Wheeler
2022-05-27 17:28 ` colyli
2022-05-28 0:58 ` Eric Wheeler
2022-05-23 18:36 ` [RFC] Add sysctl option to drop disk flushes in bcache? (was: Bcache in writes direct with fsync) Eric Wheeler
2022-05-24 5:34 ` Christoph Hellwig
2022-05-24 20:14 ` Eric Wheeler
2022-05-24 20:34 ` Keith Busch
2022-05-24 21:34 ` Eric Wheeler
2022-05-25 5:20 ` Christoph Hellwig
2022-05-25 18:44 ` Eric Wheeler
2022-05-26 9:06 ` Christoph Hellwig
2022-05-28 1:52 ` Eric Wheeler
2022-05-28 3:57 ` Keith Busch
2022-05-28 4:59 ` Christoph Hellwig
2022-05-28 12:57 ` Adriano Silva
2022-05-29 3:18 ` Keith Busch
2022-05-31 19:42 ` Eric Wheeler
2022-05-31 20:22 ` Keith Busch
2022-05-31 23:04 ` Eric Wheeler
2022-06-01 0:36 ` Keith Busch
2022-06-01 18:48 ` Eric Wheeler
[not found] ` <2064546094.2440522.1653825057164@mail.yahoo.com>
[not found] ` <YpTKfHHWz27Qugi+@kbusch-mbp.dhcp.thefacebook.com>
2022-06-01 19:27 ` Adriano Silva
2022-06-01 21:11 ` Eric Wheeler
2022-06-02 5:26 ` Christoph Hellwig
2022-05-25 5:17 ` Christoph Hellwig
[not found] ` <681726005.1812841.1653564986700@mail.yahoo.com>
2022-05-26 20:20 ` Bcache in writes direct with fsync. Are IOPS limited? Adriano Silva
2022-05-26 20:28 ` Eric Wheeler
2022-05-27 4:07 ` Adriano Silva
2022-05-28 1:27 ` Eric Wheeler
2022-05-28 7:22 ` Matthias Ferdinand [this message]
2022-05-28 12:09 ` Adriano Silva
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=YpHNts38gQMJspip@xoff \
--to=bcache@mfedv.net \
--cc=adriano_da_silva@yahoo.com.br \
--cc=bcache@lists.ewheeler.net \
--cc=colyli@suse.de \
--cc=linux-bcache@vger.kernel.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