The Relational Grid Resource project explores data models for storage of grid information system data. Grid information includes information and attributes about resources in a computational or data grid, including compute servers, clusters, storage resources, people, groups of people, etc.
The needs of the computations that run on a grid impose a unique set of demands on a resource repository. Specifically, user-driven needs include complex queries over the resource repository and update rates that guarantee a level of freshness.
We developed a database benchmark/workload for purposes of better understanding relative strengths of different platforms for grid information servers. A grid information server is an essential component of the distributed grid middleware as it is a repository of resource information about entities on the Grid.
The benchmark/workload is a set of queries and 'scenarios', or short-duration workloads, developed from a common representation for resource data that is neutral to a particular data model.
DownloadRGRbench V1.0. (Latest update 05/17/06)
IU RGRbench Workload/Benchmark User Guide to Database Creation and Population
- Beth Plale, Resource Information Management in Grid Middleware: Evaluation of Multiple Platforms with a Benchmark/Workload, manuscript submitted to journal, 2006.
- Beth Plale, Craig Jacobs, Scott Jenson, Ying Liu, Charlie Moad, Rupali Parab, and Prajakta Vaidya, Understanding Grid Resource Information Management through a Synthetic Database Benchmark/Workload, 4th IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGrid2004), April 2004.
- Deepti Kodeboyina and Beth Plale Experiences with OGSA-DAI: Portlet Access and Benchmark Global Grid Forum Workshop on Designing and Building Grid Services, Chicago, Illinois, September 2003.
- Beth Plale, Craig Jacobs, Ying Liu, Charlie Moad, Rupali Parab, and Prajakta Vaidya, Benchmark Details of Synthetic Database Benchmark/Workload for Grid Resource Information Technical Report TR-583, August 2003.
- Beth Plale, Peter Dinda, and Gregor von Laszewski, Key Concepts and Services of a Grid Information Service, ISCA 15th International Parallel and Distributed Computing Systems (PDCS), September 2002.