Workshop on Parallel Scheduling
July 14-18, 1997, Schloß Dagstuhl
During the seventies and the early eighties, scheduling problems were the
subject of intensive research, with applications mostly in the area of
operations research and adjacent fields. After a period of somewhat reduced
activities, renewed and very vigorous interest in scheduling problems has
manifested itself during the last four or five years. This time, mainly the
areas of parallel and networked computing gave rise to new research
directions and projects. The increased availability of large parallel
computers as well as much improved network technologies have significantly
increased the need for efficient resource management. The technological
developments in the area of communication networks, like broad-band,
fiber-optics, or ATM networks, have made these areas particularly interesting
and important during these last few years.
As one consequence, a large number
of heuristics have been developed by industry to cope with these challenges,
as presented at recent workshops like the one on "Job Scheduling Strategies
for Parallel Processing" at the International Parallel Processing Symposium.
In our workshop on "Parallel Scheduling", we intend to concentrate on the
following questions and problems:
Which results of classical scheduling theory can be applied to more
modern problems like on-line scheduling, parallel, and distributed
Is it possible to analyze and improve the proposed heuristics, using a
more formal approach and foundations?
How can side conditions dictated by novel applications and technologies
be combined with existing theoretical approaches?
Are there fundamentally new problems and solutions for the new types of
scheduling problems? How can they be analyzed and evaluated?
The workshop is organized by
E.W. Mayr, and
Participation in this workshop is by invitation only.
The workshop takes place at the
Hans Stadtherr, 1996-10-11
Tobias Knopff, 1997-04-11