On the Development of an Efficient Coscheduling System

  • B.B. Zhou
  • R.P. Brent
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2221)

Abstract

Applying gang scheduling can alleviate the blockade problem caused by exclusively space-sharing scheduling. To simply allow jobs to run simultaneously on the same processors as in the conventional gang scheduling, however, may introduce a large number of time slots in the system. In consequence the cost of context switches will be greatly increased, and each running job can only obtain a small portion of resources including memory space and processor utilisation and so no jobs can finish their computations quickly. In this paper we present some experimental results to show that to properly divide jobs into different classes and to apply different scheduling strategies to jobs of different classes can greatly reduce the average number of time slots in the system and significantly improve the performance in terms of average slowdown.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    A. Batat and D. G. Feitelson, Gang scheduling with memory considerations, Proceedings of 14th International Parallel and Distributed Processing Symposium, Cancun, May 2000, pp.109–114.Google Scholar
  2. 2.
    A. B. Downey, A parallel workload model and its implications for processor allocation, Proceedings of 6th International Symposium on High Performance Distributed Computing, Aug 1997.Google Scholar
  3. 3.
    D. G. Feitelson, Packing schemes for gang scheduling, In Job Scheduling Strategies for Parallel Processing, D. G. Feitelson and L. Rudolph (eds.), Lecture Notes Computer Science, Vol. 1162, Springer-Verlag, 1996, pp.89–110.Google Scholar
  4. 4.
    D. G. Feitelson and L. Rudolph, Gang scheduling performance benefits for finegrained synchronisation, Journal of Parallel and Distributed Computing, 16(4), Dec. 1992, pp.306–318.MATHCrossRefGoogle Scholar
  5. 5.
    D. Lifka, The ANL/IBM SP scheduling system, In Job Scheduling Strategies for Parallel Processing, D. G. Feitelson and L. Rudolph (Eds.), Lecture Notes Computer Science, Vol. 949, Springer-Verlag, 1995, pp.295–303.Google Scholar
  6. 6.
    J. K. Ousterhout, Scheduling techniques for concurrent systems, Proceedings of Third International Conference on Distributed Computing Systems, May 1982, pp.20–30.Google Scholar
  7. 7.
    J. Skovira, W. Chan, H. Zhou and D. Lifka, The EASY-LoadLeveler API project, In Job Scheduling Strategies for Parallel Processing, D. G. Feitelson and L. Rudolph (Eds.), Lecture Notes Computer Science, Vol. 1162, Springer-Verlag, 1996.Google Scholar
  8. 8.
    K. Suzaki, H. Tanuma, S. Hirano, Y. Ichisugi and M. Tukamoto, Time sharing systems that use a partitioning algorithm on mesh-connected parallel computers, Proceedings of the Ninth International Conference on Distributed Computing Systems, 1996, pp.268–275.Google Scholar
  9. 9.
    Y. Zhang, H. Franke, J. E. Moreira and A. Sivasubramaniam, Improving parallel job scheduling by combining gang scheduling and backfilling techniques, Proceedings of 14th International Parallel and Distributed Processing Symposium, Cancun, May 2000, pp.133–142.Google Scholar
  10. 10.
    B. B. Zhou, R. P. Brent, C. W. Johnson and D. Walsh, Job re-packing for enhancing the performance of gang scheduling, Proceedings of 5th Workshop on Job Scheduling Strategies for Parallel Processing, San Juan, April 1999, pp.129–143.Google Scholar
  11. 11.
    B. B. Zhou, D. Walsh and R. P. Brent, Resource allocation schemes for gang scheduling, Proceedings of 6th Workshop on Job Scheduling Strategies for Parallel Processing, Cancun, May 2000, pp.45–53.Google Scholar
  12. 12.
    B. B. Zhou and R. P. Brent, Gang scheduling with a queue for large jobs, accepted by 15th International Parallel and Distributed Processing Symposium, San Francisco, April 2001.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2001

Authors and Affiliations

  • B.B. Zhou
    • 1
  • R.P. Brent
    • 2
  1. 1.School of Computing and MathematicsDeakin UniversityGeelongAustralia
  2. 2.Oxford University Computing laboratoryOxfordUK

Personalised recommendations