Resource Management for Distributed Memory Multicomputers
Baxter, Jeffrey John
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/71975
Description
Title
Resource Management for Distributed Memory Multicomputers
Author(s)
Baxter, Jeffrey John
Issue Date
1992
Doctoral Committee Chair(s)
Patel, Janak H.
Department of Study
Electrical Engineering
Discipline
Electrical Engineering
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
Engineering, Electronics and Electrical
Computer Science
Abstract
In this thesis we explore the problem of resource management for multicomputer systems. A variety of algorithms were developed for different task graph models. We first present a suite of static resource management algorithms. For acyclic graphs, the LAST algorithm provides fast processor allocation and intelligent processor usage. For nondeterministic task graphs, the remap algorithm uses profiling data to iteratively improve the allocation decisions. The remap algorithm provides an efficient, distributed implementation with each processor analyzing the tasks assigned to it. Finally, the template strategy provided a static allocation to dynamic tree-based flow graphs.
Next we developed the concept of hybrid resource management, combining static and dynamic strategies. Profiling-based migration uses a migration algorithm to address runtime load imbalances for nondeterministic task costs. The algorithm uses profiled data to make migration decisions with data strictly local to each processor. For dynamic flow graphs, we developed two hybrid resource management techniques, t$\sb-$hybrid, and hybrid2. The t$\sb-$hybrid strategy is a decoupled hybrid strategy, where the static and dynamic portions of the management operate independently from one another. In the hybrid2 strategy, a coupled hybrid strategy, decisions in one strategy affect decisions taken in the other strategy.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
