qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Stefan Hajnoczi <stefanha@redhat.com>
To: virtio-fs@redhat.com, qemu-devel@nongnu.org
Cc: Liu Bo <bo.liu@linux.alibaba.com>,
	"Dr. David Alan Gilbert" <dgilbert@redhat.com>
Subject: Re: [Qemu-devel] [PATCH 0/4] virtiofsd: multithreading preparation part 3
Date: Wed, 7 Aug 2019 19:03:55 +0100	[thread overview]
Message-ID: <20190807180355.GA22758@stefanha-x1.localdomain> (raw)
In-Reply-To: <20190801165409.20121-1-stefanha@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 1931 bytes --]

On Thu, Aug 01, 2019 at 05:54:05PM +0100, Stefan Hajnoczi wrote:
> Performance
> -----------
> Please try these patches out and share your results.

Here are the performance numbers:

  Threadpool | iodepth | iodepth
     size    |    1    |   64
  -----------+---------+--------
  None       |   4451  |  4876
  1          |   4360  |  4858
  64         |   4359  | 33,266

A graph is available here:
https://vmsplice.net/~stefan/virtiofsd-threadpool-performance.png

Summary:

 * iodepth=64 performance is increased by 6.8 times.
 * iodepth=1 performance degrades by 2%.
 * DAX is bottlenecked by QEMU's single-threaded
   VHOST_USER_SLAVE_FS_MAP/UNMAP handler.

Threadpool size "none" is virtiofsd commit 813a824b707 ("virtiofsd: use
fuse_lowlevel_is_virtio() in fuse_session_destroy()") without any of the
multithreading preparation patches.  I benchmarked this to check whether
the patches introduce a regression for iodepth=1.  They do, but it's
only around 2%.

I also ran with DAX but found there was not much difference between
iodepth=1 and iodepth=64.  This might be because the host mmap(2)
syscall becomes the bottleneck and a serialization point.  QEMU only
processes one VHOST_USER_SLAVE_FS_MAP/UNMAP at a time.  If we want to
accelerate DAX it may be necessary to parallelize mmap, assuming the
host kernel can do them in parallel on a single file.  This performance
optimization is future work and not directly related to this patch
series.

The following fio job was run with cache=none and no DAX:

  [global]
  runtime=60
  ramp_time=30
  filename=/var/tmp/fio.dat
  direct=1
  rw=randread
  bs=4k
  size=4G
  ioengine=libaio
  iodepth=1

  [read]

Guest configuration:
1 vCPU
4 GB RAM
Linux 5.1 (vivek-aug-06-2019)

Host configuration:
Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz (2 cores x 2 threads)
8 GB RAM
Linux 5.1.20-300.fc30.x86_64
XFS + dm-thin + dm-crypt
Toshiba THNSFJ256GDNU (256 GB SATA SSD)

Stefan

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  parent reply	other threads:[~2019-08-07 18:04 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-01 16:54 [Qemu-devel] [PATCH 0/4] virtiofsd: multithreading preparation part 3 Stefan Hajnoczi
2019-08-01 16:54 ` [Qemu-devel] [PATCH 1/4] virtiofsd: process requests in a thread pool Stefan Hajnoczi
2019-08-05 12:02   ` Dr. David Alan Gilbert
2019-08-07  9:35     ` Stefan Hajnoczi
2019-08-01 16:54 ` [Qemu-devel] [PATCH 2/4] virtiofsd: prevent FUSE_INIT/FUSE_DESTROY races Stefan Hajnoczi
2019-08-05 12:26   ` Dr. David Alan Gilbert
2019-08-01 16:54 ` [Qemu-devel] [PATCH 3/4] virtiofsd: fix lo_destroy() resource leaks Stefan Hajnoczi
2019-08-05 15:17   ` Dr. David Alan Gilbert
2019-08-05 18:57     ` Dr. David Alan Gilbert
2019-08-06 18:58       ` Dr. David Alan Gilbert
2019-08-07  9:41       ` Stefan Hajnoczi
2019-08-01 16:54 ` [Qemu-devel] [PATCH 4/4] virtiofsd: add --thread-pool-size=NUM option Stefan Hajnoczi
2019-08-05  2:52 ` [Qemu-devel] [Virtio-fs] [PATCH 0/4] virtiofsd: multithreading preparation part 3 piaojun
2019-08-05  8:01   ` Stefan Hajnoczi
2019-08-05  9:40     ` piaojun
2019-08-07 18:03 ` Stefan Hajnoczi [this message]
2019-08-07 20:57   ` Vivek Goyal
2019-08-08  9:02     ` Stefan Hajnoczi
2019-08-08  9:53       ` Dr. David Alan Gilbert
2019-08-08 12:53         ` Vivek Goyal
2019-08-09  8:23           ` Stefan Hajnoczi
2019-08-10 21:35           ` Liu Bo
2019-08-09  8:21         ` Stefan Hajnoczi
2019-08-10 21:34           ` Liu Bo
2019-08-11  2:26           ` piaojun
2019-08-12 10:05             ` Stefan Hajnoczi
2019-08-12 11:58               ` piaojun
2019-08-12 12:51                 ` Dr. David Alan Gilbert
2019-08-08  8:10   ` piaojun
2019-08-08  9:53     ` Stefan Hajnoczi

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=20190807180355.GA22758@stefanha-x1.localdomain \
    --to=stefanha@redhat.com \
    --cc=bo.liu@linux.alibaba.com \
    --cc=dgilbert@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=virtio-fs@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;
as well as URLs for NNTP newsgroup(s).