WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-users

RE: [Xen-users] poor harddisk performance HVM domain (Windows 2003guest)

To: "Matthijs ter Woord" <matthijsterwoord@xxxxxxxxx>, "Joost van den Broek" <joost@xxxxxxxxxxxxx>, xen-users@xxxxxxxxxxxxxxxxxxx
Subject: RE: [Xen-users] poor harddisk performance HVM domain (Windows 2003guest)
From: "Petersson, Mats" <Mats.Petersson@xxxxxxx>
Date: Mon, 8 May 2006 13:15:22 +0200
Delivery-date: Mon, 08 May 2006 04:15:39 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-users-request@lists.xensource.com?subject=help>
List-id: Xen user discussion <xen-users.lists.xensource.com>
List-post: <mailto:xen-users@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
Thread-index: AcZyj8h+2KEieNBURnqaZyE22TOEoQAAAyzQ
Thread-topic: [Xen-users] poor harddisk performance HVM domain (Windows 2003guest)
> -----Original Message-----
> From: Matthijs ter Woord [mailto:matthijsterwoord@xxxxxxxxx] 
> Sent: 08 May 2006 12:08
> To: Petersson, Mats; Joost van den Broek; 
> xen-users@xxxxxxxxxxxxxxxxxxx
> Subject: Re: [Xen-users] poor harddisk performance HVM domain 
> (Windows 2003guest)
> 
> Is VMWare + HVM then slow too?

I don't know - but the slowness isn't so much HVM as the way that
hardware is emulated in Xen and most importantly that there's no current
support for para-virtual drivers, or any drivers, in Xen. 

The same scenario without HVM wouldn't be faster - just much harder to
do the virtualization. It's the "save everything, go off do something
else and return" that takes the time. In a non-HVM solution, you get
"save everything, figure out what is going on [what caused the exception
or whatever it is], go off do something else and return". So it's
actually faster to do it this way, it's just that either way is a lot
slower than runing on real hardware... A single IO access to a 33MHz PCI
device takes around 300 clock cycles on a modern (2GHz) machine, so it's
not going to be that immediate even on real hardware... 

--
Mats
> 
> 
> 
> ----- Original Message -----
> From: "Petersson, Mats" <Mats.Petersson@xxxxxxx>
> To: "Matthijs ter Woord" <matthijsterwoord@xxxxxxxxx>; "Joost 
> van den Broek"
> <joost@xxxxxxxxxxxxx>; <xen-users@xxxxxxxxxxxxxxxxxxx>
> Sent: Monday, May 08, 2006 12:03 PM
> Subject: RE: [Xen-users] poor harddisk performance HVM domain (Windows
> 2003guest)
> 
> 
> 
> 
> > -----Original Message-----
> > From: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
> > [mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of 
> Matthijs 
> > ter Woord
> > Sent: 08 May 2006 10:30
> > To: Joost van den Broek; xen-users@xxxxxxxxxxxxxxxxxxx
> > Subject: Re: [Xen-users] poor harddisk performance HVM 
> domain (Windows 
> > 2003guest)
> >
> > <wishful thinking>
> >
> > Would a AoE driver for windows do the job too in combination with a 
> > AoE server xen machine?
> >
> > </wishful thinking>
> >
> Unfortunately the answer is no, because the network 
> performance from HVM is equally hampered by the intercepts 
> and context switches as the IDE performance.
> 
> --
> Mats
> 
> 
> 


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

<Prev in Thread] Current Thread [Next in Thread>