Eugene,
you did not take into account the
dispersion/dephasing between different processes. As cluster size and the
number of instances of parallel process increase,
the dispersion increases as well, making
different instances
to be a kind out of sync - not really out of sync, but just because
of different speed of execution on different nodes, delays, etc...
If you account for this, you get the result I mentioned.
Alex
----- Original Message -----
Sent: Thursday, September 09, 2010 11:32
PM
Subject: Re: [OMPI users] MPI_Reduce
performance
Alex A. Granovsky wrote:
Isn't in evident from
the theory of random processes and probability theory
that in the limit of infinitely
large cluster and parallel process,
the probability of deadlocks with current implementation is unfortunately
quite a finite quantity and in limit approaches
to unity regardless on any particular details of the
program.
No, not at all. Consider
simulating a physical volume. Each process is assigned to some small
subvolume. It updates conditions locally, but on the surface of its
simulation subvolume it needs information from "nearby" processes. It
cannot proceed along the surface until it has that neighboring
information. Its neighbors, in turn, cannot proceed until their
neighbors have reached some point. Two distant processes can be quite
out of step with one another, but only by some bounded amount. At some
point, a leading process has to wait for information from a laggard to
propagate to it. All processes proceed together, in some loose lock-step
fashion. Many applications behave in this fashion. Actually, in
many applications, the synchronization is tightened in that "physics" is made
to propagate faster than neighbor-by-neighbor.
As the number of
processes increases, the laggard might seem relatively slower in comparison,
but that isn't deadlock.
As the size of the cluster increases, the
chances of a system component failure increase, but that also is a different
matter.
_______________________________________________
users mailing
list
users@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users