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/
Home Products Support Community News


Re: [Xen-users] Installing two versions of xen on the same machine

On Tue, Sep 8, 2009 at 5:40 PM, shankha<shankhabanerjee@xxxxxxxxx> wrote:
> Hi,
> Should i need to build the initrd image  or i can use the old one(the
> one which was build when i installed a previous version of xen
> hypervisor). I am am using the old kernel.

If you're using the old kernel you can continue to use the old initrd.
There's one possible change though: You might need to add "pci=nomsi"
to kernel parameter.


> On Tue, Sep 8, 2009 at 6:04 AM, shankha<shankhabanerjee@xxxxxxxxx> wrote:
>> Hi,
>> One more query. When i had build the source code of the xen
>> hypervisor (make world) it had build the linux kernel also(downloaded
>> the bits using mercurial).
>> Would it be possible to just build the hypervisor ?

Should be. I forgot what the command is (is it make tools?) try
looking at README (if available) or makefile.

Anyway, even if it builds and install a new kernel, you can simply
choose not to use the new (2.6.18) kernel and use your old suse kernel

>> Isn't the hypervisor and linux kernel linked ? Pardon my ignorance.
>> Like is the linux kernel  which is used is specially build to work
>> with the hyper visor or i can i can build any version of the kernel
>> and it should work with the hypervisor.

In general, yes, the hypervisor and kernel is connected. However newer
xen kernel has a compatibility option, by default selected to 3.0.2.
This means that as long as your hypervisor and userland has the same
version, and it's >= 3.0.2, it should work (minus some
version-specific feature)

I've been using Redhat's kernel-xen (built against Redhat's Xen
3.1.2+) with Xen 3.3 and 3.4, and it works great.


Xen-users mailing list