From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:55831)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <Chris.Friesen@windriver.com>) id 1X9F4a-0004wz-IN
	for qemu-devel@nongnu.org; Mon, 21 Jul 2014 11:00:25 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <Chris.Friesen@windriver.com>) id 1X9F4S-0007n9-7G
	for qemu-devel@nongnu.org; Mon, 21 Jul 2014 11:00:16 -0400
Received: from mail.windriver.com ([147.11.1.11]:65221)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <Chris.Friesen@windriver.com>) id 1X9F4R-0007ji-Vl
	for qemu-devel@nongnu.org; Mon, 21 Jul 2014 11:00:08 -0400
Message-ID: <53CD2AE1.6080803@windriver.com>
Date: Mon, 21 Jul 2014 08:59:45 -0600
From: Chris Friesen <chris.friesen@windriver.com>
MIME-Version: 1.0
References: <53C9362C.8040507@windriver.com> <53C93C34.7030403@redhat.com>
	<53C949BA.9040204@windriver.com> <53C97FEB.9060208@redhat.com>
	<53C9A440.7020306@windriver.com> <53CA06ED.1090102@redhat.com>
	<53CA0FC4.8080802@windriver.com> <53CA1D06.9090601@redhat.com>
	<20140719084537.GA3058@irqsave.net>
In-Reply-To: <20140719084537.GA3058@irqsave.net>
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
Content-Transfer-Encoding: quoted-printable
Subject: Re: [Qemu-devel] is there a limit on the number of in-flight I/O
	operations?
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: =?ISO-8859-1?Q?Beno=EEt_Canet?= <benoit.canet@irqsave.net>, Paolo Bonzini <pbonzini@redhat.com>
Cc: qemu-devel@nongnu.org

On 07/19/2014 02:45 AM, Beno=EEt Canet wrote:

> I think in the throttling case the number of in flight operation is lim=
ited by
> the emulated hardware queue. Else request would pile up and throttling =
would be
> inefective.
>
> So this number should be around: #define VIRTIO_PCI_QUEUE_MAX 64 or som=
ething like than that.

Okay, that makes sense.  Do you know how much data can be written as=20
part of a single operation?  We're using 2MB hugepages for the guest=20
memory, and we saw the qemu RSS numbers jump from 25-30MB during normal=20
operation up to 120-180MB when running dbench.  I'd like to know what=20
the worst-case would be.

Thanks,
Chris