On Wed, Jul 28, 2004 at 12:24:07AM -0500, David Buttrick scratched on the wall:
> OK. I buy your comparisons. And it had gone past me that Xgrid is a
> grid, and not a cluster-in-the-Beowulf-sense. That's a really good
> distinction.
>
> There is one limitation though - Rendezvous. If you look at that, then
> an Xgrid has more a traditional cluster network topology.
> All the nodes are connected on a single subnet.
That's true, but you don't *have* to use Rendezvous. You can
manually configure systems that are not on your subnet.
Also, see information from session 504 ("Rendezvous Update") at WWDC.
> its about guaranteeing that a result comes from who it says it
> does - "formal membership" as you said below.
I guess that phrase could be used for a lot of things. As you said,
one way to see that is a set of "known" machines, even if they are
organized in a grid-like fashion. It can be taken much further than
that, however. Cluster machines tend to be dedicated. A machine
that is part of a cluster does absolutely nothing else. This is much
less true of Xgrid machines (or grid machines in general). Cluster
machines all tend to be managed by the same person (or group) and are
as "cookie cutter" as possible-- again, grid machines aren't.
In many ways I think it is fair to say that a cluster is just a
extremely organized and formalized grid-- or that a grid is a very
disorganized cluster-- but I'm sure the computer science types will
yell at me for such comments. I think you get what I mean, however.
-j
--
Jay A. Kreibich | Integration & Software Eng.
email@hidden | Campus IT & Edu. Svcs.
<http://www.uiuc.edu/~jak> | University of Illinois at U/C
_______________________________________________
xgrid-users mailing list | email@hidden
Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/xgrid-users
Do not post admin requests to the list. They will be ignored.
