Journal of Grid Computing

, Volume 12, Issue 2, pp 371–398 | Cite as

GMATE: Dynamic Tuning of Parallel Applications in Grid Environment

  • Genaro Costa
  • Anna Sikora
  • Josep Jorba
  • Tomàs Margalef


Performance is a main issue in parallel application development. Dynamic tuning is a technique that changes certain applications’ parameters on-line to improve their performance adapting the execution to actual conditions. To perform that, it is necessary to collect measurements, analyze application behavior and carry out tuning actions during the application execution. Computational Grids present proclivity for dynamic changes in the environment during the application execution. Therefore, dynamic tuning tools are necessary to reach the expected performance indexes of applications on those environments. This paper addresses the dynamic tuning of parallel/distributed applications on Computational Grids. We analyze Grid environments to determine their characteristics and we present the development of dynamic tuning tool GMATE enabled for such environments. The performance analysis is based on performance models that indicate how to improve the application execution. A particular problem which provokes performance bottlenecks is the load imbalance in Master/Worker applications. A heuristic to dynamically tune granularity of work and number of workers is proposed. Finally, we describe the experimental validation of the performance model and its applicability on a set of real parallel applications.


Computational Grid Performance model Dynamic tuning 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Agarwala, S., Poellabauer, C., Kong, J., Schwan, K., Wolf, M.: System-level resource monitoring in high-performance computing environments. J. Grid Computing 1(3), 273–289 (2003)CrossRefMATHGoogle Scholar
  2. 2.
    Allcock, W.E., Bresnahan, J., Kettimuthu, R., Link, J.M.: The globus extensible input/output system (xio): A protocol independent io system for the Grid. In: 19th International Parallel and Distributed Processing Symposium (IPDPS 2005), CD-ROM / Abstracts Proceedings, 4–8 April 2005, Denver, CA, USA. IEEE Computer Society (2005). doi:  10.1109/IPDPS.2005.429
  3. 3.
    Argollo, E.: Performance Prediction and Tuning in a Multi-Cluster Environment. Ph.D. thesis, Departament d’Arquitectura de Computadors i Sistemes Operatius, Universitat Autonoma de Barcelona (2006)Google Scholar
  4. 4.
    Bayucan, A., Henderson, R., Jones, J., Lesiak, C., Mann, B., Nitzberg, B., Proett, T., Utley, J.: Portable Batch System Administrator Guide. Veridian Systems PBS Products Dept, 2672 Bayshore Parkway, Suite 810 Mountain View, CA 94043 (2000)Google Scholar
  5. 5.
    Berman, F., Casanova, H., Chien, A., Cooper, K., Dail, H., Dasgupta, A., Deng, W., Dongarra, J., Johnsson, L., Kennedy, K., et al.: New Grid scheduling and rescheduling methods in the grads project. Int. J. Parallel Prog. 33(2–3), 209–229 (2005)CrossRefGoogle Scholar
  6. 6.
    Berman, F., Wolski, R., Casanova, H., Cirne, W., Dail, H., Faerman, M., Figueira, S., Hayes, J., Obertelli, G., Schopf, J., Shao, G., Smallen, S., Spring, N., Su, A., Zagorodnov, D.: Adaptive computing on the Grid using apples. IEEE Trans. Parallel Distrib. Syst. 14(4), 369–382 (2003)CrossRefGoogle Scholar
  7. 7.
    Bharadwaj, V., Ghose, D., Robertazzi, T.G.: Divisible load theory: A new paradigm for load scheduling in distributed systems. Clust. Comput. 6(1), 7–17 (2003)CrossRefGoogle Scholar
  8. 8.
    Buck, B., Hollingsworth, J.K.: An api for runtime code patching. Int. J. High Perform. Comput. Appl. 14(4), 317–329 (2000). doi:  10.1177/109434200001400404 CrossRefGoogle Scholar
  9. 9.
    Caymes-Scutari, P., Morajko, A., Margalef, T., Luque, E.: Automatic generation of dynamic tuning techniques. In: Euro-Par, pp. 13–22 (2007)Google Scholar
  10. 10.
    César, E., Mesa, J.G., Sorribes, J., Luque, E.: Modeling master-worker applications in poetries. In: HIPS, pp. 22–30 (2004)Google Scholar
  11. 11.
    Costa, G., Jorba, J., Morajko, A., Margalef, T., Luque, E.: Performance models for dynamic tuning of parallel applications on computational Grids. In: CLUSTER, pp. 376–385 (2008)Google Scholar
  12. 12.
    Costa, G., Morajko, A., Margalef, T., Luque, E.: Automatic tuning in computational Grids. In: Proceedings of the 8th international conference on Applied parallel computing: state of the art in scientific computing, PARA’06, pp. 381–389. Springer-Verlag, Berlin, Heidelberg (2007). URL
  13. 13.
    De Sarkar, A., Roy, S., Ghosh, D., Mukhopadhyay, R., Mukherjee, N.: An adaptive execution scheme for achieving guaranteed performance in computational Grids. J. Grid Computing 8(1), 109–131 (2010)CrossRefGoogle Scholar
  14. 14.
    Fitzgerald, S.: Grid information services for distributed resource sharing. In: Proceedings of the 10th IEEE International Symposium on High Performance Distributed Computing, HPDC ’01, p. 181. IEEE Computer Society, Washington, DC, USA (2001). URL
  15. 15.
    Foster, I., Kesselman, C.: The Grid 2: Blueprint for a New Computing Infrastructure, 2nd edn. Morgan Kaufmann Publishers Inc., San Francisco (2003)Google Scholar
  16. 16.
    Foster, I., Kesselman, C., Tuecke, S.: The anatomy of the Grid: enabling scalable virtual organizations. Int. J. High Perform. Comput. Appl. 15, 200–222 (2001)CrossRefGoogle Scholar
  17. 17.
    Fountain, T.: Parallel Computing: Principles and Practice. Cambridge University Press (2006). URL
  18. 18.
    Genaud, S., Grunberg, M., Mongenet, C.: Experiments in running a scientific mpi application on Grid’5000. In: 4th High Performance Grid Computing International Workshop, IPDPS Conference Proceedings. IEEE (2007). URL
  19. 19.
    Germain-Renaud, C., Loomis, C., Mościcki, J.T., Texier, R.: Scheduling for responsive Grids. J. Grid Computing 6(1), 15–27 (2008)CrossRefGoogle Scholar
  20. 20.
    Gerndt, M.: Performance Tools for the Grid: State of the Art and Future: Apart White Paper. Research Report Series Lehrstuhl Fur Rechnertechnik und Rechnerorganisation Technische Universitat Munchen Series. Shaker Verlag GmbH (2004). URL
  21. 21.
    Hablot, L., Gluck, O., Mignot, J.C., Genaud, S., Primet, P.VB.: Comparison and tuning of mpi implementations in a Grid context. In: Proceedings of the 2007 IEEE International Conference on Cluster Computing, CLUSTER ’07, pp. 458–463. IEEE Computer Society, Washington, DC (2007)Google Scholar
  22. 22.
    Hager, G., Wellein, G.: Introduction to High Performance Computing for Scientists and Engineers, 1st edn. CRC Press, Inc., Boca Raton (2010)CrossRefGoogle Scholar
  23. 23.
    Heymann, E., Senar, M.A., Luque, E., Livny, M.: Adaptive scheduling for master-worker applications on the computational Grid. In: Proceedings of the First IEEE/ACM International Workshop on Grid Computing, Grid ’00, pp. 214–227. Springer-Verlag (2000)Google Scholar
  24. 24.
    Hollingsworth, J.K., Keleher, P.J.: Prediction and adaptation in active harmony. Clust. Comput. 2(3), 195–205 (1999). doi:  10.1023/A:1019034926845 CrossRefGoogle Scholar
  25. 25.
    Javadi, B., Abawajy, J.H.: Performance analysis of heterogeneous multi-cluster systems. In: Proceedings of the 2005 International Conference on Parallel Processing Workshops, ICPPW ’05, pp. 493–500 (2005)Google Scholar
  26. 26.
    Karonis, N.T., Toonen, B., Foster, I.: Mpich-g2: A Grid-enabled implementation of the message passing interface. J. Parallel Distrib. Comput. 63(5), 551–563 (2003). doi:  10.1016/S0743-7315(03)00002-9 CrossRefMATHGoogle Scholar
  27. 27.
    Keller, J., Schiffmann, W.: Guiding performance tuning for Grid schedules. In: Proceedings of the 2009 IEEE International Symposium on Parallel & Distributed Processing, IPDPS ’09, pp 1–6. IEEE Computer Society, Washington, DC (2009). doi:  10.1109/IPDPS.2009.5161169 CrossRefGoogle Scholar
  28. 28.
    Kertész, A., Kacsuk, P.: Grid interoperability solutions in Grid resource management. IEEE Syst. J. 3(1), 131–141 (2009)CrossRefGoogle Scholar
  29. 29.
    Kondo, D., Chien, A.A., Casanova, H.: Scheduling task parallel applications for rapid turnaround on enterprise desktop Grids. J. Grid Comput 5(4), 379–405 (2007)CrossRefGoogle Scholar
  30. 30.
    MacDougall, M.H.: Simulating computer systems: techniques and tools. MIT Press, Cambridge (1987)Google Scholar
  31. 31.
    Maghraoui, K., Desell, T.J., Szymanski, B.K., Varela, C.A.: Dynamic malleability in iterative mpi applications. In: 7th International Symposium on Cluster Computing and the Grid, pp. 591–598 (2008)Google Scholar
  32. 32.
    Miller, B., Cortes, A., Senar, M.A., Livny, M.: The tool d& #230;mon protocol (tdp). In: Proceedings of the 2003 ACM/IEEE conference on Supercomputing, SC ’03, p. 19. ACM, New York (2003)Google Scholar
  33. 33.
    Miller, B.P., Callaghan, M.D., Cargille, J.M., Hollingsworth, J.K., Irvin, R.B., Karavanic, K.L., Kunchithapadam, K., Newhall, T.: The paradyn parallel performance measurement tool. Computer 28(11), 37–46 (1995). doi:  10.1109/2.471178 CrossRefGoogle Scholar
  34. 34.
    Morajko, A.: Dynamic Tuning of Parallel/Distributed Applications. Ph.D. thesis, Departament d’Arquitectura de Computadors i Sistemes Operatius, Universitat Autonoma de Barcelona (2003)Google Scholar
  35. 35.
    Morajko, A., Caymes-Scutari, P., Margalef, T., Luque, E.: Automatic tuning of data distribution using factoring in master/worker applications. In: International Conference on Computational Science (2), pp. 132–139 (2005)Google Scholar
  36. 36.
    Morajko, A., Margalef, T., Luque, E.: Design and implementation of a dynamic tuning environment. J. Parallel Distrib. Comput. 67(4), 474–490 (2007)CrossRefMATHGoogle Scholar
  37. 37.
    Morajko, A., Morajko, O., Margalef, T., Luque, E.: Mate: Dynamic performance tuning environment. In: Euro-Par, pp. 98–106 (2004)Google Scholar
  38. 38.
    Ribler, R., Vetter, J., Simitci, H., Simitci, H., Reed, D.A.: Autopilot: Adaptive control of distributed applications. In: Proceedings of the 7th IEEE Symposium on High-Performance Distributed Computing, pp. 172–179 (1998)Google Scholar
  39. 39.
    Ribler, R.L., Simitci, H., Reed, D.A.: The autopilot performance-directed adaptive control system. Future Gener. Comput. Syst. 18(1), 175–187 (2001). doi:  10.1016/S0167-739X(01)00051-6 CrossRefMATHGoogle Scholar
  40. 40.
    Shende, S.S., Malony, A.D.: The tau parallel performance system. Int. J. High Perform. Comput. Appl. 20(2), 287–311 (2006). doi:  10.1177/1094342006064482 CrossRefGoogle Scholar
  41. 41.
    Team, G.: The Dynamically-Updated Request Online Coallocator (2007)Google Scholar
  42. 42.
    Thain, D., Tannenbaum, T., Livny, M.: Distributed computing in practice: The condor experience. Concurr. Comput. Pract. Experience 17, 323–356 (2005)CrossRefGoogle Scholar
  43. 43.
    Truong, H.L., Fahringer, T.: Scalea-g: A unified monitoring and performance analysis system for the Grid. Sci. Program 12(4), 225–237 (2004) URL Google Scholar
  44. 44.
    Vadhiyar, S.S., Dongarra, J.J.: Self adaptivity in Grid computing. Concurr. Comput. Pract. Experience 17(2–4), 235–257 (2005)CrossRefGoogle Scholar
  45. 45.
    Wolski, R.: Dynamically forecasting network performance using the network weather service. Clust. Comput. 1(1), 119–132 (1998). doi:  10.1023/A:1019025230054 CrossRefGoogle Scholar
  46. 46.
    Zanikolas, S., Sakellariou, R.: A taxonomy of Grid monitoring systems. Future Gener. Comput. Syst. 21(1), 163–188 (2005). doi:  10.1016/j.future.2004.07.002 CrossRefGoogle Scholar
  47. 47.
    Zhang, J., Luo, J.: Scheduling mixed-parallel application onto multicluster Grid with background workloads. In: CSCWD, pp. 429–436 (2011)Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  • Genaro Costa
    • 1
  • Anna Sikora
    • 2
  • Josep Jorba
    • 3
  • Tomàs Margalef
    • 2
  1. 1.Instituto de Humanidades, Artes e CiênciasUniversidade Federal da BahiaSalvador, BahiaBrazil
  2. 2.Computer Architecture and Operating Systems DepartmentUniversitat Autònoma de BarcelonaBarcelonaSpain
  3. 3.Estudis d’Informàtica, Multimèdia i TelecomunicacióUniversitat Oberta de CatalunyaBarcelonaSpain

Personalised recommendations