Modeling user submission strategies on production grids

ABSTRACT

Production-grid users experience many system faults as well as high and variable latencies due to the scale, complexity and sharing of such infrastructures. To improve performance, they adopt different submission strategies, that are potentially aggressive for the infrastructure.

This work studies the impact of three different strategies. It is based on a probabilistic modeling of these strategies which are evaluated according to their performance, measured as the reduction of the latency expectation, and the infrastructure overhead, measured as the additional number of submitted jobs. A strategy cost criterion is then derived.

Experiments are performed using real workload traces collected from the EGEE production infrastructure. Under these conditions, a good balance between high performance and low overhead can be found.

REFERENCES

• 1: J. Andreeva, S. Campana, F. Fanzago, and J. Herrala. High-Energy Physics on the Grid: the ATLAS and CMS Experience. Journal of Grid Computing (JGC), 6(1):3--13, Mar. 2008.
• 2: G. Aparicio, I. Blanquer Espert, and V. Hernández Garcıa. A Highly Optimized Grid Deployment: the Metagenomic Analysis Example. In Global Healthgrid: e-Science Meets Biomedical INformatics (Healthgrid'08), pages 105--115, Chicago, USA, May 2008. Healthgrid, IOS Press.
• 3: H. Casanova. Benefits and Drawbacks of Redundant Batch Requests. Journal of Grid Computing (JGC), 2(5):888--903, 2007.
• 4: K. Christodoulopoulos, V. Gkamas, and E. A. Varvarigos. Statistical Analysis and Modeling of Jobs in a Grid Environment. Journal of Grid Computing (JGC), 6(1):77--101, Mar. 2008.
• 5: Dror G. Feitelson, Workload Modeling for Performance Evaluation, Performance Evaluation of Complex Systems: Techniques and Tools, Performance 2002, Tutorial Lectures, p.114-141, January 2002
• 6: C. Germain, C. Loomis, J. T. Moscicki, and R. Texier. Scheduling for Responsive Grids. Journal of Grid Computing (JGC), 6(1):15--27, Mar. 2008.
• 7: Tristan Glatard , Diane Lingrand , Johan Montagnat , Michel Riveill, Impact of the execution context on Grid job performances, Proceedings of the Seventh IEEE International Symposium on Cluster Computing and the Grid, p.713-718, May 14-17, 2007 [doi>10.1109/CCGRID.2007.62]
• 8: Tristan Glatard , Xavier, Optimizing jobs timeouts on clusters and production grids, Proceedings of the Seventh IEEE International Symposium on Cluster Computing and the Grid, p.100-107, May 14-17, 2007 [doi>10.1109/CCGRID.2007.78]
• 9: Nicolas Jacq , Vincent Breton , Hsin-Yen Chen , Li-Yung Ho , Martin Hofmann , Vinod Kasam , Hurng-Chun Lee , Yannick Legré , Simon C. Lin , Astrid Maaí , Emmanuel Medernach , Ivan Merelli , Luciano Milanesi , Giulio Rastelli , Matthieu Reichstadt , Jean Salzemann , Horst Schwichtenberg , Ying-Ta Wu , Marc Zimmermann, Virtual screening on large scale grids, Parallel Computing, v.33 n.4-5, p.289-301, May, 2007 [doi>10.1016/j.parco.2007.02.010]
• 10: H. Li, D. Groep, and L. Walters. Workload Characteristics of a Multi-cluster Supercomputer. In Job Scheduling Strategies for Parallel Processing, pages 176--193. Springer Verlag, 2004.
• 11: D. Lingrand, J. Montagnat, and T. Glatard. Estimation of latency on production grid over several weeks. In ICT4Health, Manila, Philippines, Feb. 2008.
• 12: M. Niinimaki, X. Zhou, A. Depeursinge, A. Geissbuhler, and H. Muller. Building a Community Grid for Medical Image Analysis inside a Hospital, a Case Study. In S. Olabarriaga, D. Lingrand, and J. Montagnat, editors, MICCAI-Grid Workshop, pages 3--12, New York, NY, USA, Sept. 2008.
• 13: Mikko Juhani Pitkanen , Xin Zhou , Antti E. J. Hyvarinen , Henning Muller, Using the Grid for Enhancing the Performance of a Medical Image Search Engine, Proceedings of the 2008 21st IEEE International Symposium on Computer-Based Medical Systems, p.367-372, June 17-19, 2008 [doi>10.1109/CBMS.2008.30]
• 14: G. Sabin, R. Kettimuthu, A. Rajan, and P. Sadayappan. Schedulilng of Parallel Jobs in a Heterogeneous Multi-Site Environment. In JSSPP'03, volume LNCS 2872, pages 87--104, 2003.
• 15: Jennifer M. Schopf , Francine Berman, Stochastic scheduling, Proceedings of the 1999 ACM/IEEE conference on Supercomputing (CDROM), p.48-es, November 14-19, 1999, Portland, Oregon, United States [doi>10.1145/331532.331580]
• 16: Vijay Subramani , Rajkumar Kettimuthu , Srividya Srinivasan , P. Sadayappan, Distributed Job Scheduling on Computational Grids Using Multiple Simultaneous Requests, Proceedings of the 11th IEEE International Symposium on High Performance Distributed Computing, p.359, July 24-26, 2002

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Modeling techniques

General Terms:
Performance

Keywords:
grid computing, modelisation, submission strategy