[OpenAFS-devel] performance study

Alf Wachsmann alfw@slac.stanford.edu
Mon, 14 Feb 2005 15:08:54 -0800 (PST)


On Mon, 14 Feb 2005, John S. Bucy wrote:
> AFAIK, noone has done a detailed performance evaluation of AFS in some
> time.  I was thinking of doing such a thing for a class project.
> Here are a few things that I've thought about looking at:
>
> 1.  RX.  How does it do on the wire?  How does it compare to modern
> sunrpc+gssapi?  How does it compare to TCP for bulk-data throughput.

Have a look at Hartmut Reuter's presentation at the last German
AFS Workshop:
        http://www-dat.lrz.de/dat/afs/ws/afsws2004/pres/
He has done some intersting experiments and measurements.


> 2.  CM.  How much overhead does the cache manager add above and beyond
> the underlying filsystem?  People have mumbled that the cache has
> scalability problems.  What are they?

I know the company Sine Nomine has a DOE project with Fermi lab
to investigate where the CM bottlenecks are. I don't think they have
started yet.

-- Alf.

-----------------------------------------------------------------------
  Alf Wachsmann                       | e-mail: alfw@slac.stanford.edu
  SLAC Computing Service              | Phone:  +1-650-926-4802
  2575 Sand Hill Road, M/S 97         | FAX:    +1-650-926-3329
  Menlo Park, CA 94025, USA           | Office: Bldg. 50/323
-----------------------------------------------------------------------
                http://www.slac.stanford.edu/~alfw (PGP)
-----------------------------------------------------------------------