Back to search

University of British Columbia

The ship model : a new computational model for distriputed systems

Abstract

dc:description

One of the fundamental goals of a distributed operating is to make a collection of computers connected by a network appear as a unified whole to the users of the system. The system relies heavily on the network to help maintain this illusion. If the network is not fast enough system performance will be noticeably affected and the system will fail to achieve this primary goal. Since the network is used to allow two entities on different machines to communicate with each other, it is possible to reduce the use of the network by allowing entities to migrate between machines. Two entities on the same machine will not require the network for their communication. A distributed system which supports such a mobility mechanism has two primary benefits. First, it can increase the performance and usability of the system since decreased network communication can generally increase performance. The performance gain is most notable over low speed networks where decreased use of the network is vital for the system to perform adequately. Second, it also makes the design of system interfaces simpler by removing features from the design that are necessary only for good remote interaction. This thesis investigates a mechanism that allows distributed programs to reduce their network usage by moving code segments between computers. The general idea of moving code is developed into an abstract model of distributed computation call the Ship Model. The Ship Model has a basic entity called a ship which contains code and data. It uses mobility both as a way of moving code and data between machines and as an inter-ship communication mechanism. In order to test the viability of the model a prototype implementation is constructed. The prototype is written in C and runs under UNIX with the two parts of the system connected either by a high speed ethernet network or a slow speed dialup line. A few example applications are implemented and tested under the prototype. Various measurements show that for many applications the Ship Model can provide increased performance in a high speed network setting as well as over a slow speed line. Various issues regarding the implementation of the Ship Model are also discussed.

Degree

thesis:*
Name thesis:degree_name
Master of Science - MSc
Level thesis:degree_level
master's
Discipline thesis:degree_discipline
Computer Science
Grantor dc:publisher
University of British Columbia
Year dc:date
1992

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Phillips, George William

Rights

dc:rights
Statement dc:rights
  • For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.
Language dc:language
eng

Identifiers

dc:identifier.*
Handle dc:identifier
http://hdl.handle.net/2429/4601

Chain of custody

source
Harvested from
University of British Columbia
Base URL
circle.library.ubc.ca/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Phillips, George William. The ship model : a new computational model for distriputed systems. master's thesis, University of British Columbia, 1992. http://hdl.handle.net/2429/4601