Re: [dc] draft-khasnabish-vmmi-problems-00.txt

[Bhumip Khasnabish <[e-mail]>]

Tom,

Thanks.

Yes, seamless migration of VM and VNE can be problematic in both intra- and
inter-data-center environments, especially in multi-hypervisor case.

It may be very helpful to bring one or more of these
proprietary VM migration approaches to IETF for consideration
for standardization, if that is appropriate.
Sure, we'll update the draft to articulate these requirements.

Best.

Bhumip


On Tue, Jan 17, 2012 at 10:40 AM, Thomas Narten <[e-mail]> wrote:

> Bhumip,
>
> I skimmed this document and am having trouble figuring out what it is
> intended to do.
>
> The draft name itself has "problem" in it, but there is no single (or
> small set of) succinct problems listed. It's all very high level and
> hand wavy. I need help making the connection to an IETF action that
> could come out of this document.
>
> For example, it talks about VM migration.
>
> Is VM Migration a "problem" today? There are properietary approaches
> that the market seems to like OK.
>
> What is wrong with the current approaches? What is "broken" that needs
> fixing? Why should the IETF get involved in this space? What value
> would the IETF bring?
>
> Do you want to be able to do VM migration from one vendor's hypervisor
> to another vendor's?  If so, please just say so. Then we can see
> whether others here think that is an area the IETF (or some other SDO)
> should get involved in.
>
> Thomas
>
>