University of
Houston
Department of Computer Science
In partial
fulfillment of the Requirements for the Degree of
Master of Science
Richa Rajpal
will defend her thesis
State-of-the-art in Grid Scheduling
Abstract
Computational grids may combine many resources that are possibly geographically distributed to solve a large problem that cannot be solved on any one resource. The use of distributed resources is possible only when there is specialized middleware available that can mitigate the complexity of the grid environment to provide users with a seamless access to these resources.
In this thesis we have studied the features of resource management systems (RMSs) that are currently available within the context of grid scheduling. We consider the kinds of information needed by a grid scheduler to make intelligent scheduling decisions while dealing efficiently with problems inherent in the Grid. Grid scheduling requirements go beyond the capabilities of RMSs and we consider some ways to gather the required additional information.
Finally we evaluate two state-of-the-art schedulers that were developed specifically to work in a Grid environment. They follow different paradigms, while incorporating all the relevant Grid-related information, to provide efficient scheduling mechanisms for Grid applications.
Date:
Time:
Place: 550 PGH
Faculty, students and the general public are all invited
Thesis advisor: Dr. Barbara M. Chapman