LE3. A278 2005

2005

Giles, Rick

Acadia University

Master of Science

Masters

Computer Science

Computer Science

As the key technology for network traffic distribution, Network Load Balancing (NLB) is widely applied to help balance the network load in distributed computing, video streaming and web dispatching. With a proper NLB algorithm, a Network Load Balancing System (NLBS) can establish a scalable and stable network to serve numerous clients without any service interruption. As a sub system of NLBS, a Web Server Load Balancing System (WSLBS) is presently the most popular system. This thesis expands the study of WSLBS and WSLB algorithms by proposing an improved WSLB algorithm, namely, an Optimized-Hybrid Algorithm (OHA). Combining both static and dynamic WSLB algorithms, the OHA provides a mechanism with which to distribute the workload. A simulator is designed to evaluate the OHA and compare it with other WSLB algorithms. Performance is compared in terms of mean response time, mean response rate, and errors. The simulator system implements in Java in order to alleviate several disadvantages of current load balancing systems.

The author retains copyright in this thesis. Any substantial copying or any other actions that exceed fair dealing or other exceptions in the Copyright Act require the permission of the author.

https://scholar.acadiau.ca/islandora/object/theses:111