Data Management Policies and Scheduling in Grid Computing

Grid computing is emerging as a new paradigm for solving large-scale problems and is becoming an established technology for providing transparent access to large-scale distributed computational resources. Resource allocation and application scheduling are two of the most important aspects of Grid computing. In general, a grid application also requires datasets that may not be available at the local computing site where the application has to be executed, and hence in this case the required data has to be fetched before running the application. In this paper, we tackle with the local scheduling problem by means of a rectangle packing model combined with different policies for dataset scheduling, with the aim of maximizing the system efficiency.