|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-users] Reply for the windows virtualization (Ko0nz) Xen-users Digest, Vol 27, Issue 110
Hi koonz
I noticed that you are using the cdrom parameter in the config file I
believe that's no longer used in recent versions of Xen. I am running
debian etch here atm and am trying to virtualize windows guests aswell
I am able to boot up windows with the following configs although they do
tend to hang at some point during the installation.
----- Config windows xp -----
import os, re
arch = os.uname()[4]
if re.search('64', arch):
arch_libdir = 'lib64'
else:
arch_libdir = 'lib'
name = "xp-001"
acpi = 1
apic = 1
pae = 1
builder = 'hvm'
device_model = '/usr/lib/xen-3.0.3-1/bin/qemu-dm'
kernel = "/usr/lib/xen-3.0.3-1/boot/hvmloader"
vcpus = 1
memory = 1024
#boot (a) floppy (c) hard disk (d) cd-rom
boot = 'dc'
vif = ['type=ioemu, bridge=xenbr0']
#LVM Based Disk
disk =
['phy:/dev/ovspc170/winxp,ioemu:hda,w','file:/home/winxp/winxp_sp2.iso,h
dc:cdrom,r']
on_poweroff = 'destroy'
on_reboot = 'destroy'
on_crash = 'destroy'
vnc = 1
vncviewer = 0
sdl = 0
stdvga = 0
Kind regards,
Ldj
-----Original Message-----
From: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
[mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of
xen-users-request@xxxxxxxxxxxxxxxxxxx
Sent: woensdag 23 mei 2007 2:55
To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: Xen-users Digest, Vol 27, Issue 110
Send Xen-users mailing list submissions to
xen-users@xxxxxxxxxxxxxxxxxxx
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users
or, via email, send a message with subject or body 'help' to
xen-users-request@xxxxxxxxxxxxxxxxxxx
You can reach the person managing the list at
xen-users-owner@xxxxxxxxxxxxxxxxxxx
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Xen-users digest..."
Today's Topics:
1. i can't "xm create winxp.hvm" (ko0nz)
2. Re: Networking problems (Alessandro R.)
3. problem upgrading to 3.1.0 (Dylan Martin)
4. Re: Firewire, PCI TV Tuner Card, PCI Wireless LAN Card, and
USB Device Support Under Windows XP Xen Guest (Teo En Ming)
----------------------------------------------------------------------
Message: 1
Date: Tue, 22 May 2007 23:19:29 +0200
From: ko0nz <yesiko0nz@xxxxxxxxx>
Subject: [Xen-users] i can't "xm create winxp.hvm"
To: xen-users@xxxxxxxxxxxxxxxxxxx
Message-ID:
<1f7e0c110705221419u417f353fo46e81c83440fd9d@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
hi!
i have a debian (lenny), 2.6.18-4-xen-vserver-686 kernel, i successed
running with the debian etch guest.
But trying the windows Xp with a partition, i can't:
here's the output of my "xm create winxp.hvmy"
Using config file "winxp.hvm".
/usr/lib/xen-3.0.3-1/bin/../lib/python/xen/xm/opts.py:520:
DeprecationWarning: Non-ASCII character '\xe2' in file winxp.hvm on
line 3, but no encoding declared; see
http://www.python.org/peps/pep-0263.html for details
execfile(defconfig, globs, locs)
VNC= 1
Error: Device 2048 (vbd) could not be connected. Hotplug scripts not
working.
here's my config:
# Kernel image file
kernel = "/usr/lib/xen-3.0.3-1/boot/hvmloader"
# The domain build function. HVM domain uses ?hvm?
builder = 'hvm'
# Initial memory allocation (in megabytes) for the new domain
memory = 512
# A name for your domain. All domains must have different names
name = "winxp"
# The number of cpus guest platform has, default=1
vcpus = 1
# Enable/disable HVM guest PAE, default=1 (enabled)
pae=0
#
# # Enable/disable HVM guest ACPI, default=1 (enabled)
acpi=1
#
# # Enable/disable HVM APIC mode, default=1 (enabled)
# # Note that this option is ignored if vcpus > 1
apic=1
## vif = [ ?type=ioemu, mac=aa:00:b0:00:00:11, bridge=xenbr0? ]
vif = [ 'type=ioemu, bridge=xenbr0' ]
#disk = [ 'file:/home/xen/domains/WinXP.img,ioemu:sda,w' ]
disk = [ 'phy:/dev/sda6/winxp,ioemu:sda,w' ]
# disk = [ 'phy:/dev/sda6,ioemu:sda,w' ]
device_model = '/usr/lib/xen-3.0.3-1/bin/qemu-dm'
#cdrom='/dev/scd0'
cdrom="/home/iso/winxp_pro.iso"
# enable SDL library for graphics, default = 0
sdl=1
# enable VNC library for graphics, default = 1
#vnc=0
vnc=1
# enable spawning vncviewer(only valid when vnc=1), default = 1
vncviewer=1
serial='pty'
# enable ne2000, default = 0(use pcnet
ne2000=0
# boot on floppy (a), hard disk (c) or CD-ROM (d)
boot='d'
thank you in advance for your help.
------------------------------
Message: 2
Date: Tue, 22 May 2007 23:33:35 +0200
From: "Alessandro R." <lord2y@xxxxxxxxx>
Subject: Re: [Xen-users] Networking problems
To: "Michel Drescher" <Michel.Drescher@xxxxxxxxxxxxxx>
Cc: xen-users@xxxxxxxxxxxxxxxxxxx
Message-ID:
<79a0772d0705221433k3274290ele3d4c7313d18e3f1@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=UTF-8; format=flowed
2007/5/22, Michel Drescher <Michel.Drescher@xxxxxxxxxxxxxx>:
[cut]
can you post an lspci of your machine?
--
Alessandro R.
------------------------------
Message: 3
Date: Tue, 22 May 2007 15:56:44 -0700
From: Dylan Martin <dmartin@xxxxxxxxxxxx>
Subject: [Xen-users] problem upgrading to 3.1.0
To: xen-users@xxxxxxxxxxxxxxxxxxx
Message-ID: <20070522225644.GC7229@xxxxxxxxxxxx>
Content-Type: text/plain; charset=us-ascii
Does the xm config file format change between 3.0.3 and 3.1.0?
I'm running fedora 6 and I just built and installed the xen HV and tools
from source but left the kernel in place. That means I'm running a
3.0.3 kernel on Dom0 with 3.1.0 HV and tools.
Anyway, I go to start a DomU that was working fine before upgrading
the HV and tools and I get
Using config file "/etc/xen/tintin2".
Error: (22, 'Invalid argument')
Not a very helpful error message. I poked around in create.py and found
a
place to print a traceback, and it looks like this:
Traceback (most recent call last):
File "/usr/lib64/python2.4/xen/xm/create.py", line 1051, in
make_domain
dominfo = server.xend.domain.create(config)
File "/usr/lib64/python2.4/xmlrpclib.py", line 1096, in __call__
return self.__send(self.__name, args)
File "//usr/lib64/python/xen/util/xmlrpcclient.py", line 118, in
__request
File "/usr/lib64/python2.4/xmlrpclib.py", line 1383, in __request
verbose=self.__verbose
File "//usr/lib64/python/xen/util/xmlrpcclient.py", line 55, in
request
File "/usr/lib64/python2.4/xmlrpclib.py", line 1147, in request
return self._parse_response(h.getfile(), sock)
File "/usr/lib64/python2.4/xmlrpclib.py", line 1286, in
_parse_response
return u.close()
File "/usr/lib64/python2.4/xmlrpclib.py", line 744, in close
raise Fault(**self._stack[0])
Since the error is coming from something that seems to be for parsing
XML and my config file is the same old non-xml config file I was using
with 3.0.3, I'm wondering if my config file is supposed to be XML now.
If so, is there a tool to convert a 3.0.3 config file to the XML
format?
Thanks
-Dylan
------------------------------
Message: 4
Date: Wed, 23 May 2007 08:53:23 +0800
From: "Teo En Ming" <space.time.universe@xxxxxxxxx>
Subject: Re: [Xen-users] Firewire, PCI TV Tuner Card, PCI Wireless LAN
Card, and USB Device Support Under Windows XP Xen Guest
To: "Petersson, Mats" <Mats.Petersson@xxxxxxx>
Cc: xen-users@xxxxxxxxxxxxxxxxxxx
Message-ID:
<f712b9cf0705221753g19a5563ft23c1aa451be3a3d8@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="iso-8859-1"
Thank you Petersson. Looks like I may want for the arrival of IOMMU
hardware.
On 5/22/07, Petersson, Mats <Mats.Petersson@xxxxxxx> wrote:
>
>
>
> > -----Original Message-----
> > From: Teo En Ming [mailto:space.time.universe@xxxxxxxxx]
> > Sent: 22 May 2007 16:39
> > To: Petersson, Mats
> > Cc: xen-users@xxxxxxxxxxxxxxxxxxx
> > Subject: Re: [Xen-users] Firewire, PCI TV Tuner Card, PCI
> > Wireless LAN Card, and USB Device Support Under Windows XP Xen Guest
> >
> > Another thing to worry about is harddisk access speed in
> > Windows guests. Video editing requires fast harddisk access
> > speeds. I could give the virtual machine lots of ram if I
> > have lots of physical memory to spare, so memory requirements
> > is not much of an issue in a windows guest. Unless memory
> > access is slower than native in a windows guest.
>
> The hard-disk speed should be reasonably equal to real hardware if you
> use para-virtual drivers (such as using XenExpress or any other
> "commercial" product that is supplying the same fucntionality).
> Para-virtual drivers stop the drive from being "emulated hardware",
but
> instead feed the disk-IO directly to Dom0 in one simple packet, which
is
> much better than about 5-6 transitions between Dom0 and the guest
before
> a single disk-IO is emulated in the basic system.
>
> Memory should be very close to the native speed. There is a difference
> in handling the page-table, but I would expect a video editing
software
> to attempt to avoid page-table operation in native mode (as they are
> somewhat slow in native mode too, even if they are x times faster than
> the virtualized version).
>
> >
> >
> > On 5/22/07, Teo En Ming <space.time.universe@xxxxxxxxx> wrote:
> >
> > Oh dear, I've let the video card requirement in Windows
> > guests slipped my mind. If I remember correctly, the virtual
> > video card in Windows guest is somewhat backward/obsolete,
> > and may not work with video editing software. Even if video
> > editing software can be successfully installed in a windows
> > guest, it may refuse to run/start due to an obsolete virtual
> > video card.
> >
> > Based on the same principle as AGP and PCI, I won't be
> > able to use PCI Express x16 video cards in Windows guests too.
>
> Yes, PCI-e is also a PCI architecture from a software and most
hardware
> standpoints - only the actual signalling between one point of hardware
> and another is (very) different from other PCI architectures [and, for
> completeness, from a software standpoint, there is some support for
> extended registers - but that's not really important here].
>
> --
> Mats
> >
> > Sigh...
> >
> >
> >
> > On 5/22/07, Petersson, Mats <Mats.Petersson@xxxxxxx> wrote:
> >
> >
> >
> > > -----Original Message-----
> > > From: Teo En Ming
> > [mailto:space.time.universe@xxxxxxxxx ]
> > > Sent: 22 May 2007 15:55
> > > To: Petersson, Mats
> > > Cc: xen-users@xxxxxxxxxxxxxxxxxxx
> > > Subject: Re: [Xen-users] Firewire, PCI TV
> > Tuner Card, PCI
> > > Wireless LAN Card, and USB Device Support
> > Under Windows XP Xen Guest
> > >
> > > Hi
> > >
> > > Thank you for your reply.
> > >
> > > May I know when will IOMMU hardware be
> > arriving? Any specific
> > > roadmap/dates?
> >
> > I don't work for the right part of AMD to know
> > the planned (or actual)
> > release-dates of new products, and I don't
> > quite know which product(s)
> > the IOMMU will go into. It's not going to
> > happen in the next few weeks,
> > I can assure you of that, but as I said, I
> > don't really know much about
> > which parts will come out when - I usually know
> > that some new product
> > has been released when it's announced by e-mail
> > to all AMDers.
> > >
> > > I think I will still be going for current
> > virtualization
> > > processors. I will still be able to install
> > video editing
> > > software inside Windows XP guests and do all
> > my video editing
> > > there, while I will move all other computing
> > activities to my
> > > linux host operating system.
> >
> > Yes, as far as I can determine, there's nothing
> > in Video editing that
> > would be hardware specific, so it should work
> > just fine in a virtual
> > Windows system. [Although if the graphics
> > requirements are high for the
> > video editing software, you may still need to
> > use a dedicated machine
> > for that, rather than a virtual machine, simply
> > to get the graphics
> > performance].
> > >
> > > Will I be able to play Windows-based PC games
> > inside Windows guests?
> >
> > Short answer: No.
> > Long answer: Yes, as long as they don't require
> > high-end 3D graphics.
> > You can't use 3D graphics cards for the same
> > reason as any other PCI
> > device (AGP8x is PCI from software and most
> > hardware standpoints, it's
> > just a different connector and somewhat
> > different clock and signaling).
> >
> > --
> > Mats
> > >
> > >
> > >
> > > On 5/22/07, Petersson, Mats < Mats.Petersson@xxxxxxx
> > > <mailto:Mats.Petersson@xxxxxxx > > wrote:
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From:
> > xen-users-bounces@xxxxxxxxxxxxxxxxxxx
> > <mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx>
> > > > [mailto:
> > xen-users-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of
> > > > Teo En Ming
> > > > Sent: 22 May 2007 14:44
> > > > To: xen-users@xxxxxxxxxxxxxxxxxxx
> > <mailto:xen-users@xxxxxxxxxxxxxxxxxxx>
> > > > Subject: [Xen-users] Firewire, PCI TV
> > Tuner Card, PCI
> > > > Wireless LAN Card, and USB Device
> > Support Under
> > > Windows XP Xen Guest
> > > >
> > > > Dear All,
> > > >
> > > > Assuming that I buy a HVM compatible
> > processor and
> > > > motherboard, and having installed a
> > linux host operating
> > > > system with a Xen kernel, I proceed
> > to install a Windows XP
> > > > guest virtual machine. The question is:
> > > >
> > > > Will I be able to use the firewire
> > ports, USB ports, TV Tuner
> > > > program and wireless LAN card inside
> > Windows XP guest VM?
> > >
> > > Nope, none of these devices (aside from
> > limited USB
> > > support, possibly),
> > > will work under Xen, since (at present)
> > there is no support to
> > > hide/assign PCI devices to the HVM
> > domain. This in turn
> > > is because of
> > > the fact that PCI devices access memory
> > directly, which
> > > isn't going to
> > > work when Xen has told "lies" [1] to
> > the Windows guest
> > > about where the
> > > memory is. So when the guest OS tells
> > the PCI device
> > > where in memory
> > > something is, it will not know that
> > this is not the
> > > ACTUAL physical
> > > address. And there's no easy way to
> > solve this in software only.
> > >
> > > In future generations of
> > processors/chipsets, there
> > > will be IOMMU
> > > hardware that allows us to redirect the
> > memory requests from a
> > > particular PCI device, so that we can
> > continue to hide
> > > the ACTUAL
> > > physical address and still use the PCI
> > devices within a
> > > guest. But
> > > that's a little way out at this time.
> > >
> > >
> > > [1] All operating systems want memory
> > to start at
> > > address zero. Since
> > > only one CAN have this address, guests
> > in HVM-mode will
> > > get a fake
> > > memory map that starts at zero and goes
> > to whatever
> > > size it's configured
> > > to. The fact that the ACTUAL physical
> > address of the
> > > guest's memory is
> > > somewhere else is completely hidden
> > from the guest by
> > > using either
> > > shadow-paging or hardware assisted
> > paging (AMD Nested
> > > paging or Intel's
> > > corresponding technology) [once this
> > technology reaches
> > > customers,
> > > sometime later this year or so].
> > >
> > >
> > > > Will I be able to do video editing
> > inside Windows XP guest
> > > > VM? Or is networking the one and only
> > feature that is
> > > > supported under Windows XP guest
> > operating system? And I
> > > > won't be able to use anything else
> > inside Windows XP guest?
> > >
> > > You should be able to edit video in the
> > guest, as long
> > > as you don't rely
> > > on hardware features in PCI devices to do
this.
> > >
> > > Likewise, I don't see why you need to
> > use Windows to
> > > connect to the
> > > Wireless network, you can just as well
> > hide the fact
> > > that it's wireless
> > > from Windows, and just use virtual
> > network device, and
> > > use the Linux
> > > bridge setting to connect it to the
> > physical Wireless device.
> > >
> > > But you are correct, that the current
> > technology only
> > > allows a limited
> > > set of hardware features within the
> > guest. This is a hardware
> > > restriction, and it's nothing to do
> > with Xen in itsels,
> > > but with the
> > > current state of hardware. Future generations
of
> > > hardware will remove
> > > some or all of these restrictions (but
> > leaving one remaining
> > > restriction: each guest will HAVE to
> > have it's own
> > > hardware to access -
> > > no sharing of a single device without
> > interfacing
> > > through a virtual
> > > device - this is because all OS's
> > requires that the
> > > hardware they
> > > control is their own. There are
> > hardware devices (such
> > > as network cards)
> > > that support "multi-access" by
> > providing multiple
> > > device-instances.
> > > These of course can be shared, as they
> > are from a
> > > software standpoint
> > > multiple devices, and each device will
> > thus have it's
> > > sole owner).
> > >
> > > --
> > > Mats
> > > >
> > > > Thank you.
> > > >
> > > >
> > >
> > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
> >
> >
> >
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xensource.com/archives/html/xen-users/attachments/20070523/
47618fdc/attachment.htm
------------------------------
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
End of Xen-users Digest, Vol 27, Issue 110
******************************************
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |