Modeling the effects of node heterogeneity on the performance of grid applications

被引：2
作者：
Cremonesi, Paolo [1 ]
Turrin, Roberto [1 ]
Alexandrov, Vassil N. [2 ]
机构：
[1] Politecnico di Milano, I-20133 Milan
[2] University of Reading, Berkshire
来源：
Journal of Networks | 2009年 / 4卷 / 09期
关键词：
Grid computing; Network contention; Order statistics; Synchronization overhead; Task scheduling;
D O I：
10.4304/jnw.4.9.837-854
中图分类号：
学科分类号：
摘要：
The performance benefit when using grid systems comes from different strategies, among which partitioning the applications into parallel tasks is the most important. However, in most cases the enhancement coming from partitioning is smoothed by the effects of synchronization overheads, mainly due to the high variability in the execution times of the different tasks, which, in turn, is accentuated by the large heterogeneity of grid nodes. In this paper we design hierarchical, queuing network performance models able to accurately analyze grid architectures and applications. Thanks to the model results, we introduce a new allocation policy based on a combination between task partitioning and task replication. The models are used to study two real applications and to evaluate the performance benefits obtained with allocation policies based on task replication. © 2009 Academy Publisher.
引用
收藏
页码：837 / 854
页数：17
相关论文
共 34 条
  • [1] Zhu H., Zou Y., Zha L., Vegabench: A benchmark tool for grid system software, Grid and Cooperative Computing Workshops, 2006, pp. 543-548, (2006)
  • [2] Snavely A., Chun G., Casanova H., Wijngaart R.F.V.D., Benchmarks for grid computing: A review of ongoing efforts and future directions, SIGMETRICS Perform. Eval. Rev, 30, (2003)
  • [3] Amdhal G.M., Validity of the single processor approach to achieving large scale computing capabilities, Proc. AFIPS 1967 Spring Joint Computer Conference, 30, pp. 483-485, (1967)
  • [4] Taylor V., Wu X., Geisler J., Li X., Lan Z., Stevens R., Hereld M., Judson I.R., Prophesy: An infrastructure for analyzing and modeling the performance of parallel and distributed applications, Proc. of the 9th IEEE International Symposium on High Performance Distributed Computing (HPDC), pp. 13-18, (2000)
  • [5] Muttoni L., Casale G., Granata F., Zanero S., Optimal number of nodes for computation in grid environments, Euromicro Conference on Parallel, Distributed and Network-Based Processing (PDP04), pp. 282-289, (2004)
  • [6] Lee B.-D., Weissman J.B., Adaptive resource selection for grid-enabled network services, Proc. of the 2nd IEEE International Symposium on Network Computing and Applications (NCA), pp. 75-82, (2003)
  • [7] Giersch A., Robert Y., Vivien F., Scheduling tasks sharing files on heterogeneous master-slave platforms, Elsevier/Euromicro Journal of Systems Architecture, (2005)
  • [8] Balsamo S., Donatiello L., Dijk N.M.V., Bound performance models of heterogeneous parallel processing systems, IEEE Transactions on Parallel and Distributed Systems (PDP), 9, 10, pp. 1041-1056, (1998)
  • [9] Ko S.-S., Cycle times in a serial fork-join network, Lecture Notes in Computer Science, pp. 758-766, (2007)
  • [10] Sun X.-H., Wu M., Grid harvest service: A system for long-term, application-level task scheduling, Proc. of the 17th International Parallel and Distributed Processing Symposium (IPDPS), pp. 25-31, (2003)
1234