Date of Award
2009
Degree Type
Thesis
Degree Name
Master of Science
Program
Computer Science
Supervisor
Dr. Hanan Lutfiyya
Second Advisor
Dr. Mark Perry
Abstract
Grid computing is a form of distributed computing which is used by an organiza tion to handle its long-running computational tasks. Volunteer computing (desktop grid) is a type of grid computing that uses idle CPU cycles donated voluntarily by users, to run its tasks. In a desktop grid model, the resources are not dedicated. The job (computational task) is submitted for execution in the resource only when the resource is idle. There is no guarantee that the job which has started to execute in a resource will complete its execution without any disruption from user activity (such as keyboard click or mouse move). This problem becomes more challenging in a Peer-to-Peer (P2P) model of desktop grids where there is no central server which takes the decision on whether to allocate a job to a resource.
In this thesis we propose and implement a P2P desktop grid framework which does resource availability prediction. We try to improve the predictability of the system, by submitting the jobs on machines which have a higher probability of being available at a given time. We benchmark our framework and provide an analysis of our results.
Recommended Citation
Ramachandran, Karthick, "Decentralized Resource Availability Prediction in Peer-to-Peer Desktop Grids" (2009). Digitized Theses. 3893.
https://ir.lib.uwo.ca/digitizedtheses/3893