Re: [clouds] Announcing Clouds bar BoF during IETF-77 (March, 2010, Anaheim, CA)

[Sam Johnston <[e-mail]>]

On Tue, Feb 23, 2010 at 1:35 AM, Paul Hoffman <[e-mail]> wrote:

> At 1:08 PM -0800 2/22/10, Dave CROCKER wrote:
> >For defining 'cloud', one group I'm participating in decided it was happy
> with the NIST language:
> >
> >   <http://csrc.nist.gov/groups/SNS/cloud-computing/cloud-def-v15.doc>
>
> [ Disclaimer: I am working for NIST on a related document. ]
>
> The NIST definition is good for some environments, but it does not lend
> itself well to an IETF effort that will discuss provisioning, particularly
> the SaaS and PaaS parts. If we do not scope down more narrowly than
> "everything that can be virtualized", there is probably no reason to start
> work on how to provision that everything.
>

Agreed, the NIST document is not without its imperfections. It also hasn't
evolved a great deal, which I think is less to do with having got it right
from the start and more to do with having got it "right enough" for many. I
would also suggest that virtualisation is a red herring in that there is a
lot of focus here because that's where the legacy vendors are at (plus it's
the low hanging fruit - basically compute, storage and network resources and
not much else).

Things start to get interesting when you look at platforms (structured
storage, queues, runtimes and weird and wonderful contraptions like MTurk),
and as for the application layer, that stretches as far and wide as you
like.

Sam