Uses for a beowulf cluster?

[Gerry Creager]

Shachar Tal wrote:
> Transparent load sharing is exactly what I need. But I want a bit more. I
> want process migration, I want lastlog to be unified, etc. I don't need
> *anything* to run in parallel (not that it wouldn't be nice to have
> this), but I prefer the following scenario: each user logs in to a random
> (for the argument's sake) node, and does his work there. No interaction is
> needed for most of the work (pine, gcc, mathematica, flex, yacc, etc.),
> maybe sharing memory would be fine, but I suspect swapping pages to a
> local hard drive is a bit less expensive than shared memory across a
> 100Mb/s network.

What you're really describing is a small (or maybe, large) network of
workstations... and I don't think you're referring to a NOW.  If all are
running LINUX, have NFS and NIS and such, they will accomplish what
you've outlines, with the added benefit that they *CAN* be used in a
scalable parallel configuration if needed.  As already indicated, the
"cluster" we usually consider a "Beowulf cluseter" appears as a single
parallel entity, although it may in reality be a distributed lab of
computers, all running a coherent operating system and residing (for
discussions' sake) on a single isolated network, or might be a bunch of
computers doing nothing but parallel computation stuch in a rack or
wiring closet somewhere.

Preplanning for multiple use is a good idea, and I applaud you on it.

Gerry Creager
Mapping Sciences Laboratory
Texas A&M University