itaps-parallel Micro-migration update

[Mark Miller]

Carl Ollivier-Gooch wrote:

> 2.  We want to avoid communication blocks as much as possible, and to
> facilitate communication latency hiding as much as possible.  This means
> I'm working in a paradigm where messages fly back and forth, and each
> process has to poll for newly-received messages from time to time.  This
> implies some mechanism for the implementation to inform the application
> when some request has been completed, or when some entity isn't
> available for local modification (i.e., locked) because communication is
> ongoing.

I recall early on in the march to Terascale in another project, a number of
parallel platforms at the time did not support the kinds of communication
paradigms we needed to be fully flexible. For example, I recall believing
we needed multi-threading on each processor in order to have one thread
handling MPI messages our library was shipping around and another thread
handling the interface to the client. I just want to mention this to make sure
we are not presently aware of any computing platforms that would invalidate
your assumptions. Do your assumptions require, for example, MPI-2. I
don't think they do but lets be sure.

Also, get some sleep. You are up late. But, so am I.

Mark



--
Mark C. Miller, Lawrence Livermore National Laboratory
email: mailto:miller86 at llnl.gov
(M/T/W) (925)-423-5901 (!!LLNL BUSINESS ONLY!!)
(Th/F)  (530)-753-8511 (!!LLNL BUSINESS ONLY!!)