Abstract

Cloud computing is becoming more essential day by day. The allure of the cloud is the significant value and benefits that people gain from it, such as reduced costs, increased storage, flexibility, and more mobility. Flexibility is one of the major benefits that cloud computing can provide in terms of scaling up and down the infrastructure of a network. Once traffic has increased on one server within the network, a load balancer instance will route incoming requests to a healthy instance, which is less busy and less burdened. When the full complement of instances cannot handle any more requests, past research has been done by Chieu et. al. that presented a scaling algorithm to address a dynamic scalability of web applications on a virtualized cloud computing environment based on relevant indicators that can increase or decrease servers, as needed. In this project, I implemented the proposed algorithm, but based on CPU Utilization threshold. In addition, two tests were run exploring the capabilities of different metrics when faced with ideal or challenging conditions. The results did find a superior metric that was able to perform successfully under both tests.

Library of Congress Subject Headings

Web applications; Cloud computing; Computer algorithms

Publication Date

2013

Document Type

Thesis

Advisor

Mason, Sharon

Advisor/Committee Member

Hill, Larry

Advisor/Committee Member

Leone, Jim

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works in February 2014. Physical copy available through RIT's The Wallace Library at: QA76.585 .A54 2013

Campus

RIT – Main Campus

Plan Codes

NETSYS-MS

Download

Share

COinS