Benchmarking Grid Information Systems

Grid information systems play a central role in today's production Grid infrastructures, enabling the discovery of a range of information about the Grid services that exist in an infrastructure. As the number of services within these infrastructures continues to grow, it must be understood whether the current implementations are able to scale to meet the future requirements. Existing approaches for evaluating Grid information systems mainly focus on performance metrics and do not consider the quality of the information itself. This paper proposes a comprehensive benchmarking methodology for the evaluation of Grid information systems which includes a metric to assess the quality of the information returned. Using this methodology, two commonly used Grid information system implementations, Metadata Directory Service (MDS) and the Berkeley Database Information Index (BDII), are evaluated using data obtained from the Enabling Grids for E-SciencE (EGEE) production Grid.