Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!spool.mu.edu!sol.ctr.columbia.edu!emory!hubcap!fpst
From: mcmillan@cis.ohio-state.edu (Hal McMillan)
Newsgroups: comp.parallel
Subject: Functional (logic) program execution trees
Summary: Looking for followup to work of F.W. Burton
Keywords: dynamic load balancing
Message-ID: <1991Jun13.152613.12431@hubcap.clemson.edu>
Date: 13 Jun 91 14:27:25 GMT
Article-I.D.: hubcap.1991Jun13.152613.12431
Sender: news@cis.ohio-state.edu
Distribution: usa
Organization: The Ohio State University, Department of Computer and Information Science
Lines: 20
Approved: parallel@hubcap.clemson.edu



I have come across a paper, "Executing Functional Programs on a Virtual Tree
of Processors" (F. W. Burton, 1981), which has been used as a
reference by people doing dynamic load balancing work.  In particular,
I am interested in programs that execute in the form of a dynamic
tree.  The Burton paper has been referenced as a justification for
choosing which processes to migrate to other nodes, and which to
execute when a node is idle.  However, the Burton paper only presents
these ideas as heuristics, with a very casual justification, in the
section of the paper entitled "Execution of Applicative Programs on
the Network".

I would be interested in seeing analytical, simulation, and/or
experimental results justifying the choices that Burton recommends.
Can anyone point me to such references?

	Hal
--
Hal McMillan                | "It's an ideal principle, which can be verified
mcmillan@cis.ohio-state.edu |  only under ideal conditions.  Which means never.
72627.642@CompuServe.com    |  But it's still true."