The Effect of Correlating Quantum Allocation and Job Size for Gang Scheduling

  • Gaurav Ghare
  • Scott T. Leutenegger
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1659)

Abstract

Gang scheduling is an effective scheduling policy for multiprocessing workloads with significant interprocess synchronization and is in common use in real installations. In this paper we show that significant improvement in the job slowdown metric can be achieved simply by allocating a different number of quanta to different rows “control groups” depending on the number of processes belonging to jobs in a given row. Specifically, we show that allocating the number of quanta inversely proportionally to the number of processes per job in that row results in 20 – 50% smaller slowdowns without significantly affecting mean job response time. Incorporating these suggestions in to real schedulers would require the addition of only a few lines of simple code, hence this work should have an immediate practical impact.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    J.M. Barton and N. Bitar, A scalable multi-discipline, multiple-processor scheduling framework for irix. In Proc. of the IPPS’95 Workshop on Job Scheduling Strategies for Parallel Processing, pages 45–69. Springer LNCS:949, 1995.Google Scholar
  2. 2.
    S-H. Chiang and M. Vernon, Dynamic vs. static quantum-based parallel processor allocation. In Proc. of the IPPS’96 Workshop on Job Scheduling Strategies for Parallel Processing, pages 200–223. Springer LNCS:1162, 1996.CrossRefGoogle Scholar
  3. 3.
    Thinking Machines Corporation, Connection machine cm-5 techinical summary. Technical report, nov 1992.Google Scholar
  4. 4.
    Intel Supercomputer Systems Divsion, Paragon users guide. Technical Report Order number 312489-003, jun 1994.Google Scholar
  5. 5.
    D. Feitelson. Packing schemes for gang scheduling, In Proc. of the IPPS’96 Workshop on Job Scheduling Strategies for Parallel Processing, pages 65–88. Springer LNCS:1162, 1996.Google Scholar
  6. 6.
    D. Feitelson and M.A. Jette, Improved utilization and responsiveness with gang scheduling. In Proc. of the IPPS’97 Workshop on Job Scheduling Strategies for Parallel Processing, pages 238–261. Springer LNCS:1291, 1997.Google Scholar
  7. 7.
    D. Feitelson and L. Rudolph, Metrics and benchmarking for parallel job scheduling. In Proc. of the IPPS’98 Workshop on Job Scheduling Strategies for Parallel Processing, pages 1–24. Springer LNCS:1459, 1998.CrossRefGoogle Scholar
  8. 8.
    D. G. Feitelson and L. Rudolph, Gang scheduling performance benefits for finegrain synchronization. Journal of Parallel and Distributed Computing, 16 (4):306–318, December 1992.MATHCrossRefGoogle Scholar
  9. 9.
    D. G. Feitelson and L. Rudolph, Evaluation of design choices for gang scheduling using distributed hierarchical control. Journal of Parallel and Distributed Computing, 35 (1):18–34, May 1996.MATHCrossRefGoogle Scholar
  10. 10.
    D.G. Feitelson and B. Bitzberg, Job characteristics of a production parallel scientific workload on the nasa ames ipsc/860. In Proc. of the IPPS’95 Workshop on Job Scheduling Strategies for Parallel Processing, pages 337–360. Springer LNCS:949, apr 1995.Google Scholar
  11. 11.
    D.G. Feitelson and L. Rudolph, Distrubuted hierarchical control for parallel processing. IEEE Computer, 23 (5):65–77, 1990.Google Scholar
  12. 12.
    P. A. Fishwick, Simulation Model Design and Execution: Building Digital Worlds. Prentice Hall, 1995.Google Scholar
  13. 13.
    P.A. Fishwick, Simpack: Getting started with simulation programming in c and c++. In Proc. 1992 Winter Simulation Conference, pages 154–162, 1992.Google Scholar
  14. 14.
    A. Hori, H Tezuka, and Y Ishikawa, Overhead analysis of preemptive gang scheduling. In Proc. of the IPPS’98 Workshop on Job Scheduling Strategies for Parallel Processing, pages 217–230. Springer LNCS:1459, 1998.CrossRefGoogle Scholar
  15. 15.
    A. Hori, H Tezuka, Y Ishikawa, N. Soda, H Konaka, and M. Maeda, Implementation of gang-scheduling on workstation cluster. In Proc. of the IPPS’96 Workshop on Job Scheduling Strategies for Parallel Processing, pages 126–129. Springer LNCS:1162, 1996.CrossRefGoogle Scholar
  16. 16.
    S. Hotovy. Workload evolution on the cornell theory center ibm sp2. In Proc. Of the IPPS’96 Workshop on Job Scheduling Strategies for Parallel Processing, pages27–40. Springer LNCS:1162, 1996.CrossRefGoogle Scholar
  17. 17.
    N. Islam, A. Prodromidis, M.S. Squillante, A.S. Gopal, and L.L. Fong. Extensible resource scheduling for parallel scientific applications. In Proc. Eight SIAM Conf on Parallel Processing for Scientific Computation, mar 1997.Google Scholar
  18. 18.
    N. Islam, A. Prodromidis, M.S. Squillante, L.L. Fong, and A.S. Gopal. Extensible resource management for cluster computing. In Proc. 1997 Int. Conf. On Distributed Computing Systems (ICDCS-97), 1997.Google Scholar
  19. 19.
    M. Jette, Expanding symmetric multiprocessor capability through gang scheduling. In Proc. of the IPPS’98 Workshop on Job Scheduling Strategies for Parallel Processing, pages 199–216. Springer LNCS:1459, 1998.CrossRefGoogle Scholar
  20. 20.
    S.T. Leutenegger and M.K. Vernon, Performance of multiprogrammed multiprocessor scheduling policies. In Proc. 1990 ACM SIGMETRICS, May 1990.Google Scholar
  21. 21.
    R. K. Mansharamani and M.K. Vernon, Properties of the eqs paralle processor allocation policy. Technical Report 1192, Computer Sciences Department, University of Wisconsin-Madision, nov 1993.Google Scholar
  22. 22.
    J. Ousterhout, Scheduling techniques for concurrent systems. In Proc. of Distributed Computing Systesms Conference, pages 22–30, 1982.Google Scholar
  23. 23.
    P. Sabalvarro, S. Pakin, W. Weihl, and A.A. Chien. Dynamic coscheduling on workstation clusters. In Proc. of the IPPS’98 Workshop on Job Scheduling Strategies for Parallel Processing, pages 231–256. Springer LNCS:1459, 1998.CrossRefGoogle Scholar
  24. 24.
    U. Schwiegelshohn and R. Yahyopour, Improving first-come-frist-serve job scheduling by gang scheduling. In Proc. of the IPPS’98 Workshop on Job Scheduling Strategies for Parallel Processing, pages 180–198. Springer LNCS:1459, 1998.CrossRefGoogle Scholar
  25. 25.
    M.S. Squillante, F. Want, and M. Papaefthymiou, Stochastic analysis of gang scheduling in parallel and distributed systems. Performance Evaluation, 27:273–296.Google Scholar
  26. 26.
    F. Wang, H. Franke, M Papaefthymiou, P Pattnaik, L. Rudolph, and M. Squillante, A gang scheduing design for multiprogrammed parallel computing environments. In Proc. of the IPPS’96 Workshop on Job Scheduling Strategies for Parallel Processing, pages 111–125. Springer LNCS:1162, 1996.CrossRefGoogle Scholar
  27. 27.
    F. Wang, M Papaefthymiou, and M. Squillante, Performance evaluation of gang scheduling for parallel and distributed multiprogramming. In Proc. of the IPPS’97 Workshop on Job Scheduling Strategies for Parallel Processing, pages 277–298. Springer LNCS:1291, 1997.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1999

Authors and Affiliations

  • Gaurav Ghare
    • 1
  • Scott T. Leutenegger
    • 2
  1. 1.Mathematics and Computer Science DepartmentUniversity of DenverDenver
  2. 2.Mathematics and Computer Science DepartmentUniversity of DenverDenver

Personalised recommendations