Competitive Two-Level Adaptive Scheduling Using Resource Augmentation

  • Hongyang Sun
  • Yangjie Cao
  • Wen-Jing Hsu
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5798)


As multi-core processors proliferate, it has become more important than ever to ensure efficient execution of parallel jobs on multiprocessor systems. In this paper, we study the problem of scheduling parallel jobs with arbitrary release time on multiprocessors while minimizing the jobs’ mean response time. We focus on non-clairvoyant scheduling schemes that adaptively reallocate processors based on periodic feedbacks from the individual jobs. Since it is known that no deterministic non-clairvoyant algorithm is competitive for this problem, we focus on resource augmentation analysis, and show that two adaptive algorithms, Agdeq and Abgdeq, achieve competitive performance using O(1) times faster processors than the adversary. These results are obtained through a general framework for analyzing the mean response time of any two-level adaptive scheduler. Our simulation results verify the effectiveness of Agdeq and Abgdeq by evaluating their performances over a wide range of workloads consisting of synthetic parallel jobs with different parallelism characteristics.


Malleable jobs Mean response time Non-clairvoyant algorithm Online scheduling Resource augmentation analysis Two-level adaptive scheduling Simulations 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Agrawal, K., He, Y., Hsu, W.-J., Leiserson, C.E.: Adaptive scheduling with parallelism feedback. In: PPoPP, New York City, NY, USA, pp. 100–109 (2006)Google Scholar
  2. 2.
    Agrawal, K., He, Y., Leiserson, C.E.: An empirical evaluation of work stealing with parallelism feedback. In: ICDCS, Lisbon, Portugal, pp. 19–29 (2006)Google Scholar
  3. 3.
    Agrawal, K., He, Y., Leiserson, C.E.: Adaptive work stealing with parallelism feedback. In: PPoPP, San Jose, CA, USA, pp. 112–120 (2007)Google Scholar
  4. 4.
    Bansal, N., Chan, H.L., Lam, T.W., Lee, L.K.: Scheduling for speed bounded processors. In: ICALP, Reykjavik, Iceland, pp. 409–420 (2008)Google Scholar
  5. 5.
    Bansal, N., Pruhs, K., Stein, C.: Speed scaling for weighted flow time. In: SODA, New Orleans, LA, USA, pp. 805–813 (2007)Google Scholar
  6. 6.
    Becchetti, L., Leonardi, S.: Nonclairvoyant scheduling to minimize the total flow time on single and parallel machines. Journal of the ACM 51(4), 517–539 (2004)CrossRefMathSciNetGoogle Scholar
  7. 7.
    Berman, P., Coulston, C.: Speed is more powerful than clairvoyance. Nordic Journal of Computing 6(2), 181–193 (1999)zbMATHMathSciNetGoogle Scholar
  8. 8.
    Blumofe, R.D., Leiserson, C.E.: Scheduling multithreaded computations by work stealing. Journal of the ACM 46(5), 720–748 (1999)zbMATHCrossRefMathSciNetGoogle Scholar
  9. 9.
    Borodin, A., El-Yaniv, R.: Online computation and competitive analysis. Cambridge University Press, New York (1998)zbMATHGoogle Scholar
  10. 10.
    Brucker, P.: Scheduling Algorithms. Springer, New York (2001)zbMATHGoogle Scholar
  11. 11.
    Chan, H.-L., Edmonds, J., Lam, T.-W., Lee, L.-K., Marchetti-Spaccamela, A., Pruhs, K.: Nonclairvoyant speed scaling for flow and energy. In: STACS, Freiburg, Germany, pp. 409–420 (2009)Google Scholar
  12. 12.
    Deng, X., Gu, N., Brecht, T., Lu, K.: Preemptive scheduling of parallel jobs on multiprocessors. In: SODA, Philadelphia, PA, USA, pp. 159–167 (1996)Google Scholar
  13. 13.
    Edmonds, J.: Scheduling in the dark. In: STOC, Atlanta, GA, USA, pp. 179–188 (1999)Google Scholar
  14. 14.
    Edmonds, J., Chinn, D.D., Brecht, T., Deng, X.: Non-clairvoyant multiprocessor scheduling of jobs with changing execution characteristics. In: STOC, El Paso, TX, USA, pp. 120–129 (1997)Google Scholar
  15. 15.
    Edmonds, J., Datta, S., Dymond, P.: TCP is competitive against a limited adversary. In: SPAA, San Diego, CA, USA, pp. 174–183 (2003)Google Scholar
  16. 16.
    Edmonds, J., Pruhs, K.: Scalably scheduling processes with arbitrary speedup curves. In: SODA, New York, NY, USA, pp. 685–692 (2009)Google Scholar
  17. 17.
    Feitelson, D.G.: Job scheduling in multiprogrammed parallel systems (extended version). IBM Research Report RC19790 (87657) 2nd Revision (1997)Google Scholar
  18. 18.
    He, Y., Hsu, W.-J., Leiserson, C.E.: Provably efficient two-level adaptive scheduling. In: Frachtenberg, E., Schwiegelshohn, U. (eds.) JSSPP 2006. LNCS, vol. 4376, pp. 1–32. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  19. 19.
    He, Y., Hsu, W.-J., Leiserson, C.E.: Provably efficient online non-clairvoyant adaptive scheduling. In: IPDPS, Long Beach, CA, USA, pp. 1–10 (2007)Google Scholar
  20. 20.
    He, Y., Sun, H., Hsu, W.-J.: Adaptive scheduling of parallel jobs on functionally heterogeneous resources. In: ICPP, Xi’an, China, p. 43 (2007)Google Scholar
  21. 21.
    Kalyanasundaram, B., Pruhs, K.: Speed is as powerful as clairvoyance. In: FOCS, Milwaukee, WI, USA, pp. 214–221 (1995)Google Scholar
  22. 22.
    Kalyanasundaram, B., Pruhs, K.: Minimizing flow time nonclairvoyantly. In: FOCS, Miami Beach, FL, USA, p. 345 (1997)Google Scholar
  23. 23.
    Lam, T.W., Lee, L.-K., To, I.K.-K., Wong, P.W.H.: Speed scaling functions for flow time scheduling based on active job count. In: Halperin, D., Mehlhorn, K. (eds.) ESA 2008. LNCS, vol. 5193, pp. 647–659. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  24. 24.
    McCann, C., Vaswani, R., Zahorjan, J.: A dynamic processor allocation policy for multiprogrammed shared-memory multiprocessors. ACM Transactions on Computer Systems 11(2), 146–178 (1993)CrossRefGoogle Scholar
  25. 25.
    Motwani, R., Phillips, S., Torng, E.: Non-clairvoyant scheduling. In: SODA, Austin, TX, USA, pp. 422–431 (1993)Google Scholar
  26. 26.
    Phillips, C.A., Stein, C., Torng, E., Wein, J.: Optimal time-critical scheduling via resource augmentation (extended abstract). In: STOC, El Paso, TX, USA, pp. 140–149 (1997)Google Scholar
  27. 27.
    Pruhs, K.: Competitive online scheduling for server systems. ACM SIGMETRICS Performance Evaluation Review 34(4), 52–58 (2007)CrossRefGoogle Scholar
  28. 28.
    Pruhs, K., Torong, E., Sgall, J.: Online scheduling. In: Handbook of scheduling: Algorithms, models, and performance analysis, ch. 15, CRC Press, Boca Raton (2004)Google Scholar
  29. 29.
    Robert, J., Schabanel, N.: Non-clairvoyant batch set scheduling: Fairness is fair enough. In: Arge, L., Hoffmann, M., Welzl, E. (eds.) ESA 2007. LNCS, vol. 4698, pp. 741–753. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  30. 30.
    Robert, J., Schabanel, N.: Non-clairvoyant scheduling with precedence constraints. In: SODA, San Francisco, CA, USA, pp. 491–500 (2008)Google Scholar
  31. 31.
    Sen, S.: Dynamic processor allocation for adaptively parallel jobs. Master’s thesis, Massachusetts Institute of technology (2004)Google Scholar
  32. 32.
    Sun, H., Hsu, W.-J.: Adaptive B-Greedy (ABG): A simple yet efficient scheduling algorithm. In: SMTPS in conjunction with IPDPS, Miami, FL, USA, pp. 1–8 (2008)Google Scholar
  33. 33.
    Tucker, A., Gupta, A.: Process control and scheduling issues for multiprogrammed shared-memory multiprocessors. In: SOSP, New York, NY, USA, pp. 159–166 (1989)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Hongyang Sun
    • 1
  • Yangjie Cao
    • 2
  • Wen-Jing Hsu
    • 1
  1. 1.School of Computer EngineeringNanyang Technological UniversitySingapore
  2. 2.School of Electronic and Information EngineeringXi’an Jiaotong UniversityShanxiP.R. China

Personalised recommendations