Proposing and Evaluating Allocation Algorithms in a Grid Environment

Abstract

Distributed computing addresses workload challenges by aggregating and allocating computing resources to provide for unlimited processing power. In the last ten years, it has grown from a concept that would enable organizations to simply balance workloads across heterogeneous computer resources, to an ubiquitous solution that has been embraced by some of the world's leading organizations across multiple industry sectors. And while distributed computing harnesses the full potential of existing computer resources by effectively matching the supply of processing cycles with the demand created by applications, even more importantly it has paved the way for grid computing -a more powerful, yet global approach to resource sharing. The aim of this paper is to propose new allocation algorithms for workload management in a computing grid environment. A simulation tool is used to validate and estimate performance of these algorithms. The paper is organised as follows. After a brief introduction about grid environment and on the DataGrid Project in Section 1, Section 2 presents the proposal for novel allocation policies. Finally, Section 3 provides for a performance evaluation of the algorithms proposed, comparing their performances it with those offered by the actual solution adopted in the DataGrid Project.