[ogsa-wg] Request for Comments: InterGrid - Peering Architecture for Internetworking Disparate Grids

[Donal K. Fellows]

Rajkumar Buyya wrote:
> It will be great to hear your comments on the proposed InterGrid
> research agenda. I would also love to hear from those who are interested 
> in pursuing this line of research with focus on a specific issue/area of 
> your interest.

Interesting.

One point of note is that inter-grid peering does not need to be
symmetric, though asymmetries could be expected to result in
compensatory payments. Another is that I'd be highly surprised if all
grids were to allow outside agents as much control as your resource
allocation model seems to presume (indeed, this is almost certainly not
the case for business grids). You should also be aware that the task of
working out where to execute a task and how to actually carry out that
execution remains non-trivial in general, as things can really vary a
lot more than you'd expect if you'd only previously been exposed to a
largely-homogeneous grid like EGEE (e.g. you can't assume everything is
either runtime-deployed or pre-existing on the PATH).

Overall, I think you've got an interesting project planned, but it is a
lot more ambitious than it appears to be.

Donal.