Performance of file replication policies for real-time file access in data grids

Abstract

A variety of file replication and replacement algorithms were proposed for the Data Grids in the literature. On the other hand, real-time Grid applications are emerging in many disciplines of science and engineering. Thus, it is natural to ask about the real-time performance of these file replication and replacement algorithms. Based on this motivation, in this study, the performance of three replication algorithms, namely fast spread, cascading, and caching, are evaluated together with three different file replacement policies, including least recently used, least frequently used, and least urgent first, under various Grid settings. For this evaluation, a process oriented and discrete-event driven simulator is developed. A detailed set of simulation studies are conducted using the simulator and the results obtained are presented to elaborate on the real-time performance of the aforementioned replication algorithms.