From: Alberto Garcia <berto@igalia.com>
To: Dave Chinner <david@fromorbit.com>
Cc: Kevin Wolf <kwolf@redhat.com>,
Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
qemu-block@nongnu.org, Brian Foster <bfoster@redhat.com>,
qemu-devel@nongnu.org, Max Reitz <mreitz@redhat.com>,
linux-xfs@vger.kernel.org
Subject: Re: [PATCH 0/1] qcow2: Skip copy-on-write when allocating a zero cluster
Date: Mon, 24 Aug 2020 22:14:07 +0200 [thread overview]
Message-ID: <w51wo1nu7hs.fsf@maestria.local.igalia.com> (raw)
In-Reply-To: <20200823215907.GH7941@dread.disaster.area>
On Sun 23 Aug 2020 11:59:07 PM CEST, Dave Chinner wrote:
>> >> Option 4 is described above as initial file preallocation whereas
>> >> option 1 is per 64k cluster prealloc. Prealloc mode mixup aside, Berto
>> >> is reporting that the initial file preallocation mode is slower than
>> >> the per cluster prealloc mode. Berto, am I following that right?
>>
>> After looking more closely at the data I can see that there is a peak of
>> ~30K IOPS during the first 5 or 6 seconds and then it suddenly drops to
>> ~7K for the rest of the test.
>
> How big is the filesystem, how big is the log? (xfs_info output,
> please!)
The size of the filesystem is 126GB and here's the output of xfs_info:
meta-data=/dev/vg/test isize=512 agcount=4, agsize=8248576 blks
= sectsz=512 attr=2, projid32bit=1
= crc=1 finobt=1, sparse=1, rmapbt=0
= reflink=0
data = bsize=4096 blocks=32994304, imaxpct=25
= sunit=0 swidth=0 blks
naming =version 2 bsize=4096 ascii-ci=0, ftype=1
log =internal log bsize=4096 blocks=16110, version=2
= sectsz=512 sunit=0 blks, lazy-count=1
realtime =none extsz=4096 blocks=0, rtextents=0
>> I was running fio with --ramp_time=5 which ignores the first 5 seconds
>> of data in order to let performance settle, but if I remove that I can
>> see the effect more clearly. I can observe it with raw files (in 'off'
>> and 'prealloc' modes) and qcow2 files in 'prealloc' mode. With qcow2 and
>> preallocation=off the performance is stable during the whole test.
>
> What does "preallocation=off" mean again? Is that using
> fallocate(ZERO_RANGE) prior to the data write rather than
> preallocating the metadata/entire file?
Exactly, it means that. One fallocate() call before each data write
(unless the area has been allocated by a previous write).
> If so, I would expect the limiting factor is the rate at which IO can
> be issued because of the fallocate() triggered pipeline bubbles. That
> leaves idle device time so you're not pushing the limits of the
> hardware and hence none of the behaviours above will be evident...
The thing is that with raw (i.e. non-qcow2) images the number of IOPS is
similar, but in that case there are no fallocate() calls, only the data
writes.
Berto
next prev parent reply other threads:[~2020-08-24 20:15 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-14 14:57 [PATCH 0/1] qcow2: Skip copy-on-write when allocating a zero cluster Alberto Garcia
2020-08-14 14:57 ` [PATCH 1/1] " Alberto Garcia
2020-08-14 18:07 ` Vladimir Sementsov-Ogievskiy
2020-08-14 18:06 ` [PATCH 0/1] " Vladimir Sementsov-Ogievskiy
2020-08-17 10:10 ` Kevin Wolf
2020-08-17 15:31 ` Alberto Garcia
2020-08-17 15:53 ` Kevin Wolf
2020-08-17 15:58 ` Alberto Garcia
2020-08-17 18:18 ` Alberto Garcia
2020-08-18 8:18 ` Kevin Wolf
2020-08-19 14:25 ` Alberto Garcia
2020-08-19 15:07 ` Kevin Wolf
2020-08-19 15:37 ` Alberto Garcia
2020-08-19 15:53 ` Alberto Garcia
2020-08-19 17:53 ` Brian Foster
2020-08-20 20:03 ` Alberto Garcia
2020-08-20 21:58 ` Dave Chinner
2020-08-21 11:05 ` Brian Foster
2020-08-21 11:42 ` Alberto Garcia
2020-08-21 12:12 ` Alberto Garcia
2020-08-21 17:02 ` Brian Foster
2020-08-25 12:24 ` Alberto Garcia
2020-08-25 16:54 ` Brian Foster
2020-08-25 17:18 ` Alberto Garcia
2020-08-25 19:47 ` Brian Foster
2020-08-26 18:34 ` Alberto Garcia
2020-08-27 16:47 ` Brian Foster
2020-08-23 21:59 ` Dave Chinner
2020-08-24 20:14 ` Alberto Garcia [this message]
2020-08-21 12:59 ` Brian Foster
2020-08-21 15:51 ` Alberto Garcia
2020-08-23 22:16 ` Dave Chinner
2020-08-21 16:09 ` Alberto Garcia
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=w51wo1nu7hs.fsf@maestria.local.igalia.com \
--to=berto@igalia.com \
--cc=bfoster@redhat.com \
--cc=david@fromorbit.com \
--cc=kwolf@redhat.com \
--cc=linux-xfs@vger.kernel.org \
--cc=mreitz@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=vsementsov@virtuozzo.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;
as well as URLs for NNTP newsgroup(s).