DOE Science Accelerator (DSA) Scalability Prototype

Scaling issues arise in federated search environments when the number of sources searched becomes large. These problems are not well understood, as few large-scale federated search environments exist. In particular, these problems have not been studied in distributed grid-based environments. This project will design and develop a test framework to study the problems of scalability. The testbed will profile the performance of each of a number of grid configurations in terms of a number of characteristics. The partitioning of sources and the distribution of services across grid nodes will be modified to achieve optimal performance and demonstrate the feasibility of large-scale distributed federated searches. Commercial Applications and other Benefits as described by the awardee: A large-scale search grid would support the DOE Science Accelerator Initiative to promote global discovery. Other government and commercial organizations would benefit from the ability to effectively aggregate large numbers of content sources.