Date of Publication

1-2-1995

Document Type

Master's Thesis

Degree Name

Master of Science in Computer Science

Subject Categories

Computer Sciences

College

College of Computer Studies

Department/Unit

Computer Science

Thesis Adviser

Philip Chan

Defense Panel Chair

Peter K. Fernandez

Defense Panel Member

Harry Joson
Raymund Sison

Abstract/Summary

A distributed operating system loses its essence if it cannot manage its resources efficiently. But it is not simply a full load for its processors but an intelligent and fair one. The system must assign a task to a processor considering the processor's availability as well as its ability to satisfy the resource and computation requirements of that task. Many load distribution schemes allocate tasks based on the load of processors, but most of them deal only with an environment of processors with identical resource sets and tasks that have identical resource requirements. In this type of environment, all processors have the same ability to execute any task.

This research aims to design a dynamic, group-based load balancing algorithm for a distributed system with consideration for the diverse resource requirements of tasks and the various resource surplus of processors. This strategy represents the system as a collection of processor groups of common or similar resource sets. It also considers that a processor's resource surplus changes as these resources get allocated and released. A set of primitives will be defined to support this layer, and a load balancing algorithm will be written based on this abstract layer.

Abstract Format

html

Language

English

Format

Print

Accession Number

TG02319

Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Physical Description

1 volume (various foliations); 28 cm.

Keywords

Algorithms; Programming (Mathematics); Dynamic programming; Mathematical optimization

Upload Full Text

wf_yes

Download

Share

COinS