From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ian Campbell <Ian.Campbell@citrix.com>
Subject: Re: Phoronix Xen vs KVM vs Virtualbox benchmark with
	Ubuntu 11.10
Date: Wed, 2 Nov 2011 08:00:37 +0900
Message-ID: <1320188437.3084.28.camel@cthulhu.hellion.org.uk>
References: <20111031214036.GG12984@reaktio.net>
	<4EAFD891.8000204@icg.tugraz.at>
	<1320148634.3084.12.camel@cthulhu.hellion.org.uk>
	<20111101132639.GH12984@reaktio.net>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8bit
Return-path: <xen-devel-bounces@lists.xensource.com>
In-Reply-To: <20111101132639.GH12984@reaktio.net>
List-Unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xensource.com>
List-Help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-Subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
Sender: xen-devel-bounces@lists.xensource.com
Errors-To: xen-devel-bounces@lists.xensource.com
To: Pasi =?ISO-8859-1?Q?K=E4rkk=E4inen?= <pasik@iki.fi>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>, "dokter@icg.tugraz.at" <dokter@icg.tugraz.at>
List-Id: xen-devel@lists.xenproject.org

On Tue, 2011-11-01 at 09:26 -0400, Pasi Kärkkäinen wrote:
> On Tue, Nov 01, 2011 at 08:57:14PM +0900, Ian Campbell wrote:
> > On Tue, 2011-11-01 at 07:31 -0400, Mark Dokter wrote:
> > > In that test, the KVM machine uses virtio disk and nic devices, while
> > > the Xen machine uses the qemu emulated devices. Could that be the cause
> > > for the poor performance?
> > 
> > Very much so. Without using the PVHVM drivers in the Xen case this is
> > very much an apples to oranges comparison.
> > 
> 
> Actually the first page of the article says:
> "The only Xen issue encountered when testing it with an Ubuntu 11.10
> guest and host was the need for manually loading the xen-blkfront
> driver for disk support."
> 
> So it sounds like they actually did use PVHVM drivers..

Perhaps, the table of configuration details doesn't make it particularly
obvious (and as Alex points out they may not have been running what they
thought). Also as Alex mentioned it's not obvious which backend they
used in each case, e.g. the qemu based disk backend is not known to be
all that great.

It'd definitely be worth someone having a go at repeating even one of
the benchmarks where we did badly, to rule out these sorts of slip ups.

Ian.
> 
> -- Pasi
>