In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
will defend her dissertation proposal
Cost Delay Aware Web Request Routing
The ever-increasing dependency on internet in our day to day life causes vast amount of applications to be deployed as web services. The web services are normally deployed on clusters of redundant servers replicated across different geographical locations to make them scalable, reliable and fault-tolerant. The proliferation of web services and the need for redundancy causes large number of servers to be run from Data centers. These servers draw enormous electricity and become a concern for Data center owners with increased electricity bills. Normally the requests are served based on the proximity of servers to the clients. But the nearest server does not always serve responses with optimal latency. The factors affecting the response time are: load on the servers, intensity of the incoming requests, the traffic on the links connecting servers to clients, the number of hops between servers and clients, etc. The United Sates electricity market possess spatio-temporal variations in electricity prices. If we try to optimize energy cost by serving a request from a location where the electricity cost is cheaper, the response will be served with increased delay and causing customer dissatisfaction. So, reducing the energy cost caused by web servers and serving web requests with reduced latency is a challenging task. We propose a solution to reduce the electricity bill for Data center owners and to serve the requests with reduced latency. Our proposed solution exploits Learning Automata to select servers using energy cost and delay awareness.
Date: Wednesday, February 28, 2018
Time: 3:00 PM
Place: PGH 501D
Advisors: Dr. Jaspal Subhlok, Dr. Omprakash Gnawali, Dr. Edgar Gabriel, Dr. Ricardo Lent
Faculty, students, and the general public are invited.