Clusters (was CLUG meeting 23 May 2002)

[Ian McCulloch]

On Wed, 22 May 2002, Richard Cottrill wrote:

[...]

> 
> Can anyone give me a concise description of what 'grid computing' means?
> From what I can tell it's essentially a different name for a beowolf
> cluster... I think there's supposed to be something about the nodes not
> being dedicated cluster machines in there too.

The idea behind the grid is complex, but the practical stuff is really 
protocols for single-sign-on, data management etc.  The idea is that you 
submit your huge numerical simulation to the 'grid', and it then executes 
on whatever compatible hardware it can find, a 'collectivization' of 
all computing resources + data storage (ultimately, it would be literally 
s/internet/grid/g).  In part, its also a free and (hopefully) properly 
designed+implemented version of M$ Passport etc, aimed currently at users 
of high performance computers.  There is also a grid implementation of MPI 
(message passing interface) for running parallel programs on a network of 
non-homogeneous architecture.  http://www.globus.org/ is where much of 
this stuff is happening.  I don't know specifically what you refer to with 
the comparison to a beowulf, I wouldn't be at all surprised if there is a 
grid-ified version.  Do you have any links?

Cheers,
Ian