Re: [vNUMA v2][PATCH 2/8] public interface

[Andre Przywara <andre.przywara@amd.com>]

Keir Fraser wrote:
> On 03/08/2010 16:43, "Dulloor" <dulloor@gmail.com> wrote:
> 
>>> I would expect guest would see nodes 0 to nr_vnodes-1, and the mnode_id
>>> could go away.
>> mnode_id maps the vnode to a particular physical node. This will be
>> used by balloon driver in
>> the VMs when the structure is passed as NUMA enlightenment to PVs and
>> PV on HVMs.
>> I have a patch ready for that (once we are done with this series).
> 
> So what happens when the guest is migrated to another system with different
> physical node ids? Is that never to be supported? I'm not sure why you
> wouldn't hide the vnode-to-mnode translation in the hypervisor.
And what about if the node assignment changes at guest's runtime to 
satisfy load-balancing? I think we should have the opportunity to change 
the assignment, although this could be costly when it involves copying 
guest memory to another physical location. A major virtualization 
product ;-) solves this by a "hot pages first, the rest in background" 
algorithm. I see that this is definitely a future extension, but we 
shouldn't block this way already in this early stage.

Andre.

-- 
Andre Przywara
AMD-Operating System Research Center (OSRC), Dresden, Germany
Tel: +49 351 448-3567-12