Date of Award

10-1-1996

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science

First Advisor

Dr. Heshem El-Rewini

Abstract

Parallel database systems are gaining popularity as a solution that provides scalability in large and growing databases. A parallel database system is a DBS which exploits multiprocessing systems to improve performance. Parallel database computers can be classified into three categories: shared memory, shared disk, and shared nothing. In shared memory, all resources, including main memory and disk units, are shared among several processors. In shared disk, a group of processors share a common pool of disks, but each processor has its own private main memory. In the shared-nothing system, every processor has its own memory and disk unit, that is, except for communication links, no resources are shared among the processors. In this work, we· compare the performance of the three architecture classes. Simulation models for the various architectures are introduced. Using these models, a number of experiments were conducted to compare the system performance of these architectures under different workloads and transaction models. The aim of this work is to provide a tool for evaluating the different architectures and their appropriateness for a specific database application.

Comments

A Thesis Presented to the Department of Computer Science and the Faculty of the Graduate College University of Nebraska In Partial Fulfillment of the Requirements for the Degree Master of Science University of Nebraska at Omaha. Copyright 1996 Emad Eldin Mohamed Mohamed

Download

COinS