A goal of grid computing is having to compile-in or statically configure as little as possible by out-of-band means. I've been calling this a posteriori computing (the opposite of having a priori knowledge). This is addressed by virtualization in all of the areas mentioned here. The real toughest nut will be true virtualization of service provisioning. This is certainly a fundamental research topic. --Craig At 04:34 AM 3/29/2006, Dennis Gannon wrote:

I agree with Bob. virtualization is critical to grid success. it has actually been there all along. the first grid standards virtualized access and identity. the next phase was virtualizing resources and now ogsa and other architectures are virtualizing services and applications. But this path has not always been clear and there are entire industries that are emerging, such as those virtualizing file systems/storage at the network level, or virtualizing computational resources.

however, i feel there is still a lot of hype to sort through. an interesting approach would be to do a pannel at ggf17 or ggf18 that looks at real virtualization use-cases and ask the question: how are grid standards helping to enable these? Bob, i still don't have a voice-of-the-community panel proposal from you. why don't you do it on this topic? dennis

On Tue, 28 Mar 2006, Robert Cohen wrote:

...

Hello all,

We did not get to virtualization today, but I am seeing many more vendors focusing on it. I think that one thing that would be very helpful in interacting with EGA, particularly with vendors who don't believe that GGF is responsive to their customers, would be to have a discussion on how to deal with the issue of virtualization with some GGF community and standards folks as well as EGA and perhaps a few interested sponsors.

One goal would be to begin to formulate a more encompassing and convergent GGF-EGA view of where virtualization is going and how it will impact the deployment of grids in research and commercial settings.

Another goal would be to identify a few initial steps that might let GGF and the GGF-EGA NEWORG begin to integrate virtualization into conferences and programs, plus discussions about the goals of the NEWORG.

A third goal would be to see how the emergence of GGF and EGA as leaders in bringing virtualization more visibility might enhance the position that the two organizations have in the press. If the press sees GGF-EGA taking on the importance of virtualization and insuring that 1) standards being developed for grids are on a path that is preparing grids for virtualization and 2) business models for virtualization are being taken seriously (we have already seen the impact in telcos), this might change the perception of GGF as just an SDO.

I would suggest an initial call under the GGF tent to discuss this, probably more among the community people first, but I am open to suggestions.

I do see the trend to virtualization as one that is beginning to subsume Grids under its definition but I think the two are really intertwined and that the issue of networks and grids is key in moving to virtualization.

Bob

Fogel, Robert wrote:

...

<<GGF VOC and VAF.ppt>>

Robert Fogel Grid & Service-Oriented IT WW Strategy and Business Development Intel Corp 408-765-6404

_____________________________________________ From: Fogel, Robert Sent: Monday, August 22, 2005 6:31 PM To: cads@ggf.org; Replogle, Joel (GGF); Steve Crumb Subject: GGF Community Council Bi-Weekly (REMINDER) When: Tuesday, March 28, 2006 10:00 AM-11:00 AM (GMT-06:00) Central Time (US & Canada). Where: 1-888-455-3044 and +1 210 839 8611, Passcode: 94172

(3/27) - Agenda below (3/14) - Added international call-in info. (3/13) - This is a placeholder for GGF Community Council Bi-Weekly meeting. Each week I will pre-send an agenda and any materials for our meeting.

Agenda:

o GGF17 update (Dennis Gannon) o Review of GGF17 content and unfilled sessions o Review roadmap/plan for VOC and VAF at GGF17, GGF18 and beyond o GGF18 Community program