[Gllug] Xen and bridging

Adrian McMenamin adrian at mcmen.demon.co.uk
Tue May 16 13:33:58 UTC 2006


On Tue, May 16, 2006 1:14 pm, Daniel P. Berrange said:
>>
> Basically Xen creates one or more virtual interfaces in Dom0 for each
> guest OS that runs. These are referred to as vif0, vif1, etc in the
> host OS, or just plain eth0 in the guest OS. In a pure bridging mode,
> the vifX devices will be connected straight to the host's real eth0
> device, so packets to/from the guest go to/arrive from the physical
> network without the dom0's network stack really getting involved. So
> if you have DHCP on your LAN, the guests will auto-assign themselves
> IP addr in the normal manner.
>

Well, I read that somewhere but didn't quite believe it - I assumed some
automagical step had been left out :) The wonders of Linux!

So do these virtual devices also have virtual NIC hardware addresses - so
I can assign what my router calls static DHCP?

I am starting to see the power of all this now.


> There is a possible alternative setup where you don't bridge the vifX
> devices onto eth0, and instead setup a traditional NAT forwarding using
> iptables. This gives you an isolated subnet for your guests, which may
> or may not be desirable depending on your intended use cases for the
> guests.
>
> Dan.
> --
> |=-            GPG key: http://www.berrange.com/~dan/gpgkey.txt       -=|
> |=-       Perl modules: http://search.cpan.org/~danberr/              -=|
> |=-           Projects: http://freshmeat.net/~danielpb/               -=|
> |=-   berrange at redhat.com  -  Daniel Berrange  -  dan at berrange.com    -=|
> --
> Gllug mailing list  -  Gllug at gllug.org.uk
> http://lists.gllug.org.uk/mailman/listinfo/gllug
>


-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug




More information about the GLLUG mailing list