Advertisement

Master-Slave Tasking on Asymmetric Networks

  • Cyril Banino-Rokkones
  • Olivier Beaumont
  • Lasse Natvig
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4128)

Abstract

This paper presents new techniques for master-slave tasking on tree-shaped networks with fully heterogeneous communication and processing resources. A large number of independent, equal-sized tasks are distributed from the master node to the slave nodes for processing and return of result files. The network links present bandwidth asymmetry, i.e. the send and receive bandwidths of a link may be different. The nodes can overlap computation with at most one send and one receive operation. A centralized algorithm that maximizes the platform throughput under static conditions is presented. Thereafter, we propose several distributed heuristics making scheduling decisions based on information estimated locally. Extensive simulations demonstrate that distributed heuristics are better suited to cope with dynamic environments, but also compete well with centralized heuristics in static environments.

Keywords

Linear Programming Problem Child Node Master Node Bandwidth Utilization Throughput Rate 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Hong, B., Prasanna, V.K.: Performance Optimization of a De-centralized Task Allocation protocol via bandwidth and buffer management. In: CLADE, p. 108 (2004)Google Scholar
  2. 2.
    Casanova, H.: SimGrid: A Toolkit for the Simulation of Application Scheduling. In: Proceedings of the 1st International Symposium on Cluster Computing and the Grid, p. 430. IEEE Computer Society, Los Alamitos (2001)CrossRefGoogle Scholar
  3. 3.
    Banino-Rokkones, C., Beaumont, O., Natvig, L.: Master-Slave Tasking on Asymmetric Tree-Shaped Networks. Technical Report 02/06, NTNU (2006), URL: http://www.idi.ntnu.no/~banino/research/research.html
  4. 4.
    Robertazzi, T.: Processor Equivalence for a Linear Daisy Chain of Load Sharing Processors. IEEE Trans. Aerospace and Electronic Systems 29, 1216–1221 (1993)CrossRefGoogle Scholar
  5. 5.
    Bharadwaj, V., Ghose, D., Mani, V., Robertazzi, T.: Scheduling Divisible Loads in Parallel and Distributed Systems. IEEE Computer Society Press, Los Alamitos (1996)Google Scholar
  6. 6.
    Drozdowski, M., Wolniewicz, P.: Experiments with scheduling divisible tasks in clusters of workstations. In: Bode, A., Ludwig, T., Karl, W.C., Wismüller, R. (eds.) Euro-Par 2000. LNCS, vol. 1900, pp. 311–319. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  7. 7.
    Barlas, G.D.: Collection-Aware Optimum Sequencing of Operations and Closed-Form Solutions for the Distribution of a Divisible Load on Arbitrary Processor Trees. IEEE Trans. Parallel Distrib. Syst. 9(5), 429–441 (1998)CrossRefGoogle Scholar
  8. 8.
    Blazewicz, J., Drozdowski, M., Guinand, F., Trystram, D.: Scheduling a Divisible Task in a Two-dimensional Toroidal Mesh. In: Proceedings of the third international conference on Graphs and optimization, pp. 35–50. Elsevier Science Publishers BV, Amsterdam, The Netherlands (1999)Google Scholar
  9. 9.
    Adler, M., Gong, Y., Rosenberg, A.L.: Optimal Sharing of Bags of Tasks in Heterogeneous Clusters. In: 15th ACM Symp. on Parallelism in Algorithms and Architectures (SPAA 2003), pp. 1–10. ACM Press, New York (2003)Google Scholar
  10. 10.
    Beaumont, O., Marchal, L., Robert, Y.: Scheduling Divisible Loads with Return Messages on Heterogeneous Master-Worker Platforms. In: Bader, D.A., Parashar, M., Sridhar, V., Prasanna, V.K. (eds.) HiPC 2005. LNCS, vol. 3769, pp. 123–132. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  11. 11.
    Kreaseck, B., Carter, L., Casanova, H., Ferrante, J.: Autonomous Protocols for Bandwidth-Centric Scheduling of Independent-Task Applications. In: IPDPS 2003: Proceedings of the 17th International Symposium on Parallel and Distributed Processing, Washington, pp. 26.1. IEEE Computer Society, Los Alamitos (2003)Google Scholar
  12. 12.
    Dutot, P.F.: Complexity of Master-slave Tasking on Heterogeneous Trees. European Journal on Operationnal Research 164(3), 690–695 (2005)CrossRefMathSciNetGoogle Scholar
  13. 13.
    Rosenberg, A.L.: Sharing Partitionable Workloads in Heterogeneous NOWs: Greedier is not Better. In: Cluster Computing 2001, pp. 124–131. IEEE Computer Society Press, Los Alamitos (2001)Google Scholar
  14. 14.
    Banino, C., Beaumont, O., Carter, L., Ferrante, J., Legrand, A., Robert, Y.: Scheduling Strategies for Master-Slave Tasking on Heterogeneous Processor Platforms. IEEE Transactions on Parallel and Distributed Systems 15(4), 319–330 (2004)CrossRefGoogle Scholar
  15. 15.
    Hong, B., Prasanna, V.K.: Distributed Adaptive Task Allocation in Heterogeneous Computing Environments to Maximize Throughput. In: International Parallel and Distributed Processing Symposium IPDPS 2004, p. 52b. IEEE Computer Society Press, Los Alamitos (2004)Google Scholar
  16. 16.
    Bertsimas, D., Gamarnik, D.: Asymptotically optimal algorithm for job shop scheduling and packet routing. Journal of Algorithms 33(2), 296–318 (1999)MATHCrossRefMathSciNetGoogle Scholar
  17. 17.
    Hong, B., Prasanna, V.K.: Bandwidth-Aware Resource Allocation for Heterogeneous Computing Systems to Maximize Throughput. In: ICPP, pp. 539–546 (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Cyril Banino-Rokkones
    • 1
  • Olivier Beaumont
    • 2
  • Lasse Natvig
    • 1
  1. 1.Norwegian University of Science and TechnologyTrondheimNorway
  2. 2.LaBRI, UMR CNRS 5800, Domaine UniversitaireTalence CedexFrance

Personalised recommendations