[Beowulf] PVFS on 80 proc (40 node) cluster

[Reuti]

> Reuti:
>
> > What programs will your cluster run and how are the users submitting the
> jobs? 
> 
> Jeff:
>
> - the program is a large physics code that does I/O
> (200KB or less) every 10 to 60 sec.  Every 10min or
> so, a 100MB file is written.

It's completely different from our requirements. We share /home with the 
(small) input files, and each node needs a large local /scratch space (100GB 
and more).

> - I want a *single* filesystem, so no local scratch
> will be used.

A single file system for /home and /scratch (will your software need a common 
/scratch space)? Is there any error correction in PVFS in case that a disk or 
node fails? Another solution could be IBM's GPFS if you need a big and fast 
common file space.

Cheers - Reuti