作者: G. Obertelli , F. Berman , R. Wolski , H. Casanova
关键词:
摘要: The Computational Grid is a promising platform for the efficient execution of parameter sweep applications over large spaces. To achieve performance on Grid, such must be scheduled so that shared data files are strategically placed to maximize reuse, and application can adapt deliverable potential target heterogeneous, distributed resources. Parameter an important class would greatly benefit from development middleware embeds scheduler targets resources transparently. In this paper we describe user-level project, AppLeS Sweep Template (APST), uses application-level scheduling techniques [1] various technologies allow deployment Grid. We discuss several possible algorithms detail our software design. then current implementation APST using systems like Globus [2], NetSolve [3] Network Weather Service [4], present experimental results.