From: Anthony Liguori <anthony@codemonkey.ws>
To: Avi Kivity <avi@redhat.com>
Cc: Stefan Hajnoczi <stefanha@gmail.com>,
Christian Brunner <chb@muc.de>, Blue Swirl <blauwirbel@gmail.com>,
kvm@vger.kernel.org, qemu-devel@nongnu.org,
ceph-devel@vger.kernel.org
Subject: Re: [Qemu-devel] [RFC PATCH 1/1] ceph/rbd block driver for qemu-kvm
Date: Mon, 24 May 2010 14:19:34 -0500 [thread overview]
Message-ID: <4BFAD146.9090708@codemonkey.ws> (raw)
In-Reply-To: <4BFA5D07.8030309@redhat.com>
On 05/24/2010 06:03 AM, Avi Kivity wrote:
> On 05/24/2010 11:27 AM, Stefan Hajnoczi wrote:
>> On Sun, May 23, 2010 at 1:01 PM, Avi Kivity<avi@redhat.com> wrote:
>>> On 05/21/2010 12:29 AM, Anthony Liguori wrote:
>>>> I'd be more interested in enabling people to build these types of
>>>> storage
>>>> systems without touching qemu.
>>>>
>>>> Both sheepdog and ceph ultimately transmit I/O over a socket to a
>>>> central
>>>> daemon, right?
>>> That incurs an extra copy.
>> Besides a shared memory approach, I wonder if the splice() family of
>> syscalls could be used to send/receive data through a storage daemon
>> without the daemon looking at or copying the data?
>
> Excellent idea.
splice() eventually requires a copy. You cannot splice() to linux-aio
so you'd have to splice() to a temporary buffer and then call into
linux-aio. With shared memory, you can avoid ever bringing the data
into memory via O_DIRECT and linux-aio.
Regards,
Anthony Liguori
WARNING: multiple messages have this Message-ID (diff)
From: Anthony Liguori <anthony@codemonkey.ws>
To: Avi Kivity <avi@redhat.com>
Cc: kvm@vger.kernel.org, Stefan Hajnoczi <stefanha@gmail.com>,
qemu-devel@nongnu.org, Blue Swirl <blauwirbel@gmail.com>,
ceph-devel@vger.kernel.org, Christian Brunner <chb@muc.de>
Subject: Re: [Qemu-devel] [RFC PATCH 1/1] ceph/rbd block driver for qemu-kvm
Date: Mon, 24 May 2010 14:19:34 -0500 [thread overview]
Message-ID: <4BFAD146.9090708@codemonkey.ws> (raw)
In-Reply-To: <4BFA5D07.8030309@redhat.com>
On 05/24/2010 06:03 AM, Avi Kivity wrote:
> On 05/24/2010 11:27 AM, Stefan Hajnoczi wrote:
>> On Sun, May 23, 2010 at 1:01 PM, Avi Kivity<avi@redhat.com> wrote:
>>> On 05/21/2010 12:29 AM, Anthony Liguori wrote:
>>>> I'd be more interested in enabling people to build these types of
>>>> storage
>>>> systems without touching qemu.
>>>>
>>>> Both sheepdog and ceph ultimately transmit I/O over a socket to a
>>>> central
>>>> daemon, right?
>>> That incurs an extra copy.
>> Besides a shared memory approach, I wonder if the splice() family of
>> syscalls could be used to send/receive data through a storage daemon
>> without the daemon looking at or copying the data?
>
> Excellent idea.
splice() eventually requires a copy. You cannot splice() to linux-aio
so you'd have to splice() to a temporary buffer and then call into
linux-aio. With shared memory, you can avoid ever bringing the data
into memory via O_DIRECT and linux-aio.
Regards,
Anthony Liguori
next prev parent reply other threads:[~2010-05-24 19:19 UTC|newest]
Thread overview: 129+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-19 19:22 [RFC PATCH 1/1] ceph/rbd block driver for qemu-kvm Christian Brunner
2010-05-19 19:22 ` [Qemu-devel] " Christian Brunner
2010-05-20 20:31 ` Blue Swirl
2010-05-20 20:31 ` Blue Swirl
2010-05-20 21:18 ` Christian Brunner
2010-05-20 21:18 ` Christian Brunner
2010-05-20 21:29 ` Anthony Liguori
2010-05-20 21:29 ` Anthony Liguori
2010-05-20 22:16 ` Christian Brunner
2010-05-20 22:16 ` Christian Brunner
2010-05-21 5:28 ` Stefan Hajnoczi
2010-05-21 5:28 ` Stefan Hajnoczi
2010-05-21 6:13 ` MORITA Kazutaka
2010-05-21 6:13 ` MORITA Kazutaka
2010-05-21 5:54 ` MORITA Kazutaka
2010-05-21 5:54 ` MORITA Kazutaka
2010-05-23 12:01 ` Avi Kivity
2010-05-23 12:01 ` Avi Kivity
2010-05-24 7:12 ` MORITA Kazutaka
2010-05-24 7:12 ` MORITA Kazutaka
2010-05-24 11:05 ` Avi Kivity
2010-05-24 11:05 ` Avi Kivity
2010-05-24 11:42 ` MORITA Kazutaka
2010-05-24 11:42 ` MORITA Kazutaka
2010-05-24 11:56 ` Avi Kivity
2010-05-24 11:56 ` Avi Kivity
2010-05-24 12:07 ` Cláudio Martins
2010-05-24 12:07 ` Cláudio Martins
2010-05-24 14:01 ` MORITA Kazutaka
2010-05-24 14:01 ` MORITA Kazutaka
2010-05-24 19:07 ` Christian Brunner
2010-05-24 19:07 ` Christian Brunner
2010-05-24 19:38 ` Anthony Liguori
2010-05-24 19:38 ` Anthony Liguori
2010-05-25 9:14 ` Avi Kivity
2010-05-25 9:14 ` Avi Kivity
2010-05-25 13:17 ` Anthony Liguori
2010-05-25 13:17 ` Anthony Liguori
2010-05-25 13:25 ` Avi Kivity
2010-05-25 13:25 ` Avi Kivity
2010-05-25 13:29 ` Anthony Liguori
2010-05-25 13:29 ` Anthony Liguori
2010-05-25 13:36 ` Avi Kivity
2010-05-25 13:36 ` Avi Kivity
2010-05-25 13:54 ` Anthony Liguori
2010-05-25 13:54 ` Anthony Liguori
2010-05-25 13:57 ` Avi Kivity
2010-05-25 13:57 ` Avi Kivity
2010-05-25 14:02 ` Anthony Liguori
2010-05-25 14:02 ` Anthony Liguori
2010-05-26 8:44 ` Avi Kivity
2010-05-26 8:44 ` Avi Kivity
2010-05-25 14:01 ` Kevin Wolf
2010-05-25 14:01 ` Kevin Wolf
2010-05-25 16:21 ` Avi Kivity
2010-05-25 16:21 ` Avi Kivity
2010-05-25 17:12 ` Sage Weil
2010-05-25 17:12 ` Sage Weil
2010-05-25 17:12 ` Sage Weil
2010-05-26 5:24 ` MORITA Kazutaka
2010-05-26 5:24 ` MORITA Kazutaka
2010-05-26 8:46 ` Avi Kivity
2010-05-26 8:46 ` Avi Kivity
2010-05-24 19:16 ` Anthony Liguori
2010-05-24 19:16 ` Anthony Liguori
2010-05-25 9:19 ` Avi Kivity
2010-05-25 9:19 ` Avi Kivity
2010-05-25 13:26 ` MORITA Kazutaka
2010-05-25 13:26 ` MORITA Kazutaka
2010-05-24 8:27 ` Stefan Hajnoczi
2010-05-24 8:27 ` Stefan Hajnoczi
2010-05-24 11:03 ` Avi Kivity
2010-05-24 11:03 ` Avi Kivity
2010-05-24 19:19 ` Anthony Liguori [this message]
2010-05-24 19:19 ` Anthony Liguori
2010-05-25 9:22 ` Avi Kivity
2010-05-25 9:22 ` Avi Kivity
2010-05-25 11:02 ` Kevin Wolf
2010-05-25 11:02 ` Kevin Wolf
2010-05-25 11:25 ` Avi Kivity
2010-05-25 11:25 ` Avi Kivity
2010-05-25 12:03 ` Christoph Hellwig
2010-05-25 12:03 ` Christoph Hellwig
2010-05-25 12:13 ` Avi Kivity
2010-05-25 12:13 ` Avi Kivity
2010-05-25 13:25 ` Anthony Liguori
2010-05-25 13:25 ` Anthony Liguori
2010-05-25 13:31 ` Avi Kivity
2010-05-25 13:31 ` Avi Kivity
2010-05-25 13:35 ` Anthony Liguori
2010-05-25 13:35 ` Anthony Liguori
2010-05-25 13:38 ` Avi Kivity
2010-05-25 13:38 ` Avi Kivity
2010-05-25 13:55 ` Anthony Liguori
2010-05-25 13:55 ` Anthony Liguori
2010-05-25 14:01 ` Avi Kivity
2010-05-25 14:01 ` Avi Kivity
2010-05-25 14:05 ` Anthony Liguori
2010-05-25 14:05 ` Anthony Liguori
2010-05-25 15:00 ` Avi Kivity
2010-05-25 15:00 ` Avi Kivity
2010-05-25 15:01 ` Anthony Liguori
2010-05-25 15:01 ` Anthony Liguori
2010-05-25 16:16 ` Avi Kivity
2010-05-25 16:16 ` Avi Kivity
2010-05-25 16:21 ` Anthony Liguori
2010-05-25 16:21 ` Anthony Liguori
2010-05-25 16:27 ` Avi Kivity
2010-05-25 16:27 ` Avi Kivity
2010-05-25 13:53 ` Kevin Wolf
2010-05-25 13:53 ` Kevin Wolf
2010-05-25 13:55 ` Avi Kivity
2010-05-25 13:55 ` Avi Kivity
2010-05-25 14:03 ` Anthony Liguori
2010-05-25 14:03 ` Anthony Liguori
2010-05-25 15:02 ` Avi Kivity
2010-05-25 15:02 ` Avi Kivity
2010-05-25 14:09 ` Kevin Wolf
2010-05-25 14:09 ` Kevin Wolf
2010-05-25 15:01 ` Avi Kivity
2010-05-25 15:01 ` Avi Kivity
2010-05-20 23:02 ` Yehuda Sadeh Weinraub
2010-05-20 23:02 ` Yehuda Sadeh Weinraub
2010-05-23 7:59 ` Blue Swirl
2010-05-23 7:59 ` Blue Swirl
2010-05-24 2:17 ` Yehuda Sadeh Weinraub
2010-05-24 2:17 ` Yehuda Sadeh Weinraub
2010-05-25 20:13 ` Blue Swirl
2010-05-25 20:13 ` [Qemu-devel] " Blue Swirl
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=4BFAD146.9090708@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=avi@redhat.com \
--cc=blauwirbel@gmail.com \
--cc=ceph-devel@vger.kernel.org \
--cc=chb@muc.de \
--cc=kvm@vger.kernel.org \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@gmail.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.