From: Anthony Liguori <anthony@codemonkey.ws>
To: Aurelien Jarno <aurelien@aurel32.net>
Cc: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>,
Jan Kiszka <jan.kiszka@siemens.com>,
Marcelo Tosatti <mtosatti@redhat.com>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
Anthony Liguori <aliguori@linux.vnet.ibm.com>,
Paul Brook <paul@codesourcery.com>,
"Edgar E. Iglesias" <edgar.iglesias@gmail.com>,
Paolo Bonzini <pbonzini@redhat.com>,
Arun Bharadwaj <arun@linux.vnet.ibm.com>
Subject: Re: [Qemu-devel] Re: [PATCH 2/7] Enable I/O thread and VNC threads by default
Date: Wed, 09 Feb 2011 21:07:52 +0100 [thread overview]
Message-ID: <4D52F418.7070000@codemonkey.ws> (raw)
In-Reply-To: <20110209173523.GJ16429@hall.aurel32.net>
On 02/09/2011 06:35 PM, Aurelien Jarno wrote:
> On Tue, Feb 08, 2011 at 04:08:28PM +0100, Aurelien Jarno wrote:
>
>> Aurelien Jarno a écrit :
>>
>>> Paolo Bonzini a écrit :
>>>
>>>> On 02/08/2011 12:15 PM, Aurelien Jarno wrote:
>>>>
>>>>> however
>>>>> it should not be done ignoring all the*current* drawbacks of the
>>>>> iothread mode. We know them (at least for some of them), so let's try to
>>>>> solve them.
>>>>>
>>>> Let's also enumerate them.
>>>>
>>>>
>>> From what I know:
>>> - performance regression in TCG mode
>>>
>> I setup an x86_64 guest on an x86_64 host (Intel Xeon E5345). Nothing
>> was running except the standard daemons and the CPU governor was set to
>> "performance" on all CPU. I then compared the network performance using
>> netperf in default mode, through a tap interface and a virtio nic. I got
>> the following results (quite reproducible, std below 0.5):
>> - without IO thread: 107.36 MB/s
>> - with IO thread: 89.93 MB/s
>>
>>
> And the same test on the code from september 2009:
> - without IO thread: 141.8 MB/s
>
virtio-net is super finicky regarding mitigation strategies and their
relationship to the I/O thread. Different benchmarks will behave
differently. virtio-blk is probably a better device to test as you'll
get much more consistent results across different type of I/O patterns.
Regards,
Anthony Liguori
next prev parent reply other threads:[~2011-02-09 20:08 UTC|newest]
Thread overview: 101+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-24 21:00 [Qemu-devel] [RFC 0/7] Introduce hard dependency on glib Anthony Liguori
2011-01-24 21:00 ` [Qemu-devel] [PATCH 1/7] io-thread: make sure to initialize qemu_work_cond and qemu_cpu_cond Anthony Liguori
2011-02-08 8:53 ` [Qemu-devel] " Jan Kiszka
2011-02-08 9:01 ` Anthony Liguori
2011-01-24 21:00 ` [Qemu-devel] [PATCH 2/7] Enable I/O thread and VNC threads by default Anthony Liguori
2011-01-24 22:28 ` [Qemu-devel] " Anthony Liguori
2011-01-25 9:17 ` Edgar E. Iglesias
2011-01-25 13:34 ` Marcelo Tosatti
2011-02-07 10:12 ` Marcelo Tosatti
2011-02-07 16:03 ` Marcelo Tosatti
2011-02-07 16:23 ` Paolo Bonzini
2011-02-07 17:10 ` Jan Kiszka
2011-02-07 21:02 ` Anthony Liguori
2011-02-07 21:45 ` Aurelien Jarno
2011-02-08 2:09 ` Anthony Liguori
2011-02-08 7:26 ` Aurelien Jarno
2011-02-08 8:08 ` Paolo Bonzini
2011-02-08 8:50 ` Jan Kiszka
2011-02-08 9:05 ` Aurelien Jarno
2011-02-08 9:12 ` Anthony Liguori
2011-02-08 9:49 ` Paolo Bonzini
2011-02-08 9:51 ` Jan Kiszka
2011-02-08 9:58 ` Aurelien Jarno
2011-02-08 10:03 ` Jan Kiszka
2011-02-08 10:06 ` Aurelien Jarno
2011-02-08 10:16 ` Alexander Graf
2011-02-08 10:17 ` Stefan Hajnoczi
2011-02-08 10:27 ` Aurelien Jarno
2011-02-08 10:31 ` Paolo Bonzini
2011-02-08 10:40 ` Jan Kiszka
2011-02-08 18:05 ` Anthony Liguori
2011-02-08 11:29 ` Aurelien Jarno
2011-02-08 12:38 ` Riku Voipio
2011-02-08 10:21 ` Jan Kiszka
2011-02-08 10:26 ` Aurelien Jarno
2011-02-08 10:30 ` Jan Kiszka
2011-02-08 17:58 ` Anthony Liguori
2011-02-08 11:07 ` Tristan Gingold
2011-02-08 11:46 ` Aurelien Jarno
2011-02-08 12:07 ` Paolo Bonzini
2011-02-08 19:21 ` Anthony Liguori
2011-02-08 11:15 ` Aurelien Jarno
2011-02-08 12:10 ` Paolo Bonzini
2011-02-08 13:31 ` Aurelien Jarno
2011-02-08 15:08 ` Aurelien Jarno
2011-02-09 17:35 ` Aurelien Jarno
2011-02-09 20:07 ` Anthony Liguori [this message]
2011-02-11 0:03 ` Marcelo Tosatti
2011-02-08 19:17 ` Anthony Liguori
2011-02-08 13:30 ` Aurelien Jarno
2011-02-08 20:54 ` Anthony Liguori
2011-02-08 15:09 ` Aurelien Jarno
2011-02-09 17:13 ` Blue Swirl
2011-02-09 22:16 ` [Qemu-devel] " Stefan Weil
2011-02-10 7:34 ` Paolo Bonzini
2011-02-10 9:54 ` Paolo Bonzini
2011-02-10 19:46 ` Stefan Weil
2011-02-08 10:06 ` [Qemu-devel] " Paolo Bonzini
2011-02-07 18:35 ` Edgar E. Iglesias
2011-02-07 20:44 ` Aurelien Jarno
2011-02-07 21:30 ` Scott Wood
2011-02-07 20:47 ` Edgar E. Iglesias
2011-01-25 8:33 ` [Qemu-devel] [PATCH 0/2] vnc: the lost parts Corentin Chary
2011-01-25 8:33 ` [Qemu-devel] [PATCH 1/2] vl.c: set NULL upon deleting handlers in qemu_set_fd_handler2() Corentin Chary
2011-01-25 10:03 ` Stefan Hajnoczi
2011-01-25 10:13 ` Corentin Chary
2011-01-25 10:26 ` Stefan Hajnoczi
2011-01-25 12:05 ` Yoshiaki Tamura
2011-01-25 8:33 ` [Qemu-devel] [PATCH 2/2] vnc: qemu can die if the client is disconnected while updating screen Corentin Chary
2011-01-24 21:00 ` [Qemu-devel] [PATCH 3/7] Add support for glib based threading and convert qemu thread to use it Anthony Liguori
2011-01-25 14:24 ` Aurelien Jarno
2011-01-25 15:34 ` Anthony Liguori
2011-02-02 17:32 ` [Qemu-devel] " Paolo Bonzini
2011-02-02 17:35 ` Anthony Liguori
2011-01-24 21:00 ` [Qemu-devel] [PATCH 4/7] Get rid of QemuMutex and teach its callers about GStaticMutex Anthony Liguori
2011-01-24 22:24 ` [Qemu-devel] " Jan Kiszka
2011-01-25 0:02 ` Anthony Liguori
2011-01-25 7:39 ` Jan Kiszka
2011-01-24 21:00 ` [Qemu-devel] [PATCH 5/7] threads: get rid of QemuCond and teach callers about GCond Anthony Liguori
2011-01-24 21:00 ` [Qemu-devel] [PATCH 6/7] Teach vnc server to use GThread directly Anthony Liguori
2011-01-26 10:39 ` Stefan Hajnoczi
2011-01-24 21:00 ` [Qemu-devel] [PATCH 7/7] Rename QemuThread to QemuSThread to indicate that it is not a generic thread Anthony Liguori
2011-01-24 21:28 ` [Qemu-devel] Re: [RFC 0/7] Introduce hard dependency on glib Paolo Bonzini
2011-01-24 22:01 ` Anthony Liguori
2011-01-25 10:41 ` Paolo Bonzini
2011-01-25 11:14 ` Daniel P. Berrange
2011-01-25 11:21 ` Paolo Bonzini
2011-01-25 0:24 ` [Qemu-devel] " Anthony Liguori
2011-01-25 6:51 ` Edgar E. Iglesias
2011-01-25 10:24 ` Stefan Hajnoczi
2011-01-25 11:51 ` Gerd Hoffmann
2011-01-25 12:04 ` Daniel P. Berrange
2011-01-25 14:48 ` Stefano Stabellini
2011-01-25 17:48 ` Anthony Liguori
2011-01-25 18:12 ` Stefano Stabellini
2011-01-25 14:23 ` Aurelien Jarno
2011-01-25 15:35 ` Anthony Liguori
[not found] ` <20110126044710.GU9566@redhat.com>
2011-01-26 15:53 ` Anthony Liguori
2011-01-26 21:23 ` Stefan Hajnoczi
2011-01-26 22:12 ` Anthony Liguori
2011-01-26 17:48 ` Johannes Stezenbach
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=4D52F418.7070000@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=aliguori@linux.vnet.ibm.com \
--cc=arun@linux.vnet.ibm.com \
--cc=aurelien@aurel32.net \
--cc=edgar.iglesias@gmail.com \
--cc=jan.kiszka@siemens.com \
--cc=mtosatti@redhat.com \
--cc=paul@codesourcery.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@linux.vnet.ibm.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.