Using Performance Prediction to Allocate Grid Resources
Seung-Hye Jang, Xingfu Wu, Valerie Taylor, Gaurang Mehta, Karan Vahi, Ewa Deelman · 2004
Large-scale applications often require computational grids to obta in the needed compute power for execution. Generally, users are given access to a collec tion of resources that can be used for execution. The collection of resources can be dynamic, and the use rs must decide which collection of heterogeneous and distributed resources to use. Howe ver, many users often do not have knowledge of all of the resource performance characterist ics on which to make informed decisions and therefore need automated tools to perform the mapping of jobs to the available resources. In this paper, we present a resource planner syste m that uses performance prediction, based upon historical data, to identify the appropriate resources. This system is used with a gravitational-wave physics experiment, LIGO, for which the initi al results indicate an average of 24% reduction in execution time using the performance prediction v ersus a random selection of resources.