Could you go into a little more detail about how the simulator accesses
data? Is it mostly sequential, mostly random, mostly reading, reading
and writing? What are the file sizes? How many are there?
And perhaps you could answer a few of the questions asked earlier, in
particular,
what cache manager switches do you use?
"Glew, Andy" wrote:
> I'm just trying to figure out why the trend is so much against
> AFS at the sites that I have returned to. Why various
> groups that I have to work with are religiously against AFS.
> Why people insist on setting up yet another CVS server (with lousy
> security, etc.) instead of sharing source code via AFS.
> Why the simulator jockeys manually replicate data via rsync
> instead of sharing via AFS.
