[PVFS2-users] Low performance of PVFS2 on our small cluster

[Robert Latham]

On Wed, Oct 27, 2004 at 09:15:57PM +0200, Silvestre Zabala wrote:
> One datapoint more: 
> I have a super-simple MPI program that reads and writes large 
> not overlapping chunks into a file. Running on 10 nodes it gets
> a read bandwidth of 450 MB/s, which is very good, as that's exactly the
> aggregated bw of all disks, but a write bw of only 30 MB/s. I hadn't
> expected such a difference. 

Me either.  

We've got a similar benchmark (mpi-io-test), and on jazz
(separate IO and computer nodes) PVFS2 produces an aggregate write
bandwidth of ~180 MB/sec, or roughly 22 MB per server -- which is
close to what we would expect given the low-end hard drives on these
nodes.  

If you compare mpi-io-test to your benchmark, are they doing the same
things?  

==rob

-- 
Rob Latham
Mathematics and Computer Science Division    A215 0178 EA2D B059 8CDF
Argonne National Labs, IL USA                B29D F333 664A 4280 315B