Xen project Mailing List

Re: [Xen-users] Networking with xen

To: "Quezada, Pedro" <PedroQ@xxxxxxxxxx>

From: "Grant McWilliams" <grantmasterflash@xxxxxxxxx>

Date: Wed, 15 Oct 2008 13:25:28 -0700

Cc: Luke S Crawford <lsc@xxxxxxxxx>, xen-users@xxxxxxxxxxxxxxxxxxx

Delivery-date: Wed, 15 Oct 2008 13:26:12 -0700

Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:references; b=f+uhUSClqGpNRVvWCIWMHxvbJUwk/aq3w2oNRZt12vi4UklDjJl3IeN2OLgxUPJy3o 6GrS6D18bf/lEpOXEFC0w/E1iwh66REUFL08x1baMhItyY1dC6npCkHh7XGfhf22E1B/ baypgIRWrnzBvJXhLd+Kv2b5rG6SWWfHSr3bk=

List-id: Xen user discussion <xen-users.lists.xensource.com>

On Wed, Oct 15, 2008 at 12:13 PM, Quezada, Pedro <PedroQ@xxxxxxxxxx> wrote:

hey I found this

http://www.vmware.com/pdf/hypervisor_performance.pdf

And since it's from vmware it obviously not biased... Notice that they didn't test Commercial Xen against the very expensive ESX but the open source one. VMware is the only one that can legally release performance figures of their product. Try to figure out why that is. I'm not knocking ESX because it's probably Xen's only competition but VMWare doesn't want anyone else to release numbers because someone might actually test an area that Xen comes out on top.

They show Xen as only having 3-6% of the speed of a native implementation on netperf when I'm seeing 10x that (and I'm still complaining). Also they only tested the Windows HVM drivers which are nothing more the QEMU emultated hardware. Again had they compared commercial software to commercial software things would have been different.

It's been known for quite a while that Xen has IO problems. I'm not sure how much of this has been fixed in 3.3 as apposed to 3.03.
http://media.brianmadden.com/briforumplayer/bfplayerdynamic.asp?id=296&sparky=1

Grant McWilliams

_______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-users