This Article 
 Bibliographic References 
 Add to: 
Allocating Task Interaction Graphs to Processors in Heterogeneous Networks
September 1997 (vol. 8 no. 9)
pp. 908-925

Abstract—The problem of allocating task interaction graphs (TIGs) to heterogeneous computing systems to minimize job completion time is investigated. The only restriction is that the interprocessor communication cost is the same for any pair of processors. This is suitable for local area network based systems, such as Ethernet, as well as fully interconnected multiprocessor systems. An optimal polynomial solution exists if sufficient homogeneous processors and communication capacity are available. This solution is generalized to obtain two faster heuristics, one for the case of homogeneous processors and the other for heterogeneous processors. The heuristics were tested extensively with 60,900 systematically generated random TIGs and shown to be stable independent of the size of the TIG. A performance model is also proposed to predict the performance of the heuristic algorithms, and it is successful in explaining the experimental results qualitatively.

[1] D. Bernstein, M. Rodeh, and I. Gertner, "On the Complexity of Scheduling Problems for Parallel/Pipelined Machines," IEEE Trans. Computers, vol. 38, no. 9, pp. 1,308-1,313, Sept. 1989.
[2] S.H. Bokhari, "Partitioning Problems in Parallel, Pipelined and Distributed Computing," IEEE Trans. Computers, vol. 37, no. 1, pp. 48-57, Jan. 1988.
[3] N.S. Bowen, C.N. Nikolaou, and A. Ghafoor, “On the Assignment Problem of Arbitrary Process Systems to Heterogeneous Distributed Computer Systems,” IEEE Trans. Computers, vol. 41, no. 3, Mar. 1992.
[4] T. Bultan and C. Aykanat, "A New Mapping Heuristic Based on Mean Field Annealing," J. Parallel and Distributed Computing, vol. 16, pp. 292-305, Dec. 1992.
[5] T.L. Casavant and J.G. Kuhl,“A taxonomy of scheduling in general-purpose distributed computing systems,” IEEE Trans. on Software Engineering, vol. 14, no. 2. Feb. 1988.
[6] V. Chaudhary and J.K. Aggarwal, "A Generalized Scheme for Mapping Parallel Algorithms," IEEE Trans. Parallel and Distributed Systems, Mar. 1993, pp. 328-346.
[7] H. Clark and B. McMillin,“DAWGS—A distributed compute Srever utilizing idle workstations,”J. Parallel Distrib. Comput., pp. 175–186, Feb. 1992.
[8] E.G. Coffman Jr., Computer and Job Shop Scheduling Theory.New York: Wiley, 1976.
[9] E.G. Coffman Jr., M.R. Garey, and D.S. Johnson, "A Application of Bin-Packing to Multiprocessor Scheduling," SIAM J. Computing, vol. 7, pp. 1-17, Feb. 1978.
[10] K. Efe, "Heuristic Models of Task Assignment Scheduling in Distributed Systems," Computer, vol. 15, pp. 50-56, June 1982.
[11] A.V. Goldberg and R.E. Tarjan, "A New Approach to the Maximum Flow Problem," Proc. 18th Ann. Symp. Theory of Computing, pp. 136-146, 1987.
[12] R.L. Graham, "Bounds on Multiprocessing Timing Anomalies," SIAM J. Applied Math., vol. 17, pp. 416-429, Mar. 1969.
[13] S. Hurley, "Taskgraph Mapping Using a Genetic Algorithm: A Comparison of Fitness Functions," Parallel Computing, vol. 19, pp. 1,313-1,317, Nov. 1993.
[14] B.W. Kernighan and S. Lin, "An Efficient Heuristic Procedure for Partitioning Graphs," Bell System Technical J., vol. 49, pp. 291-307, Feb. 1970.
[15] P. Krueger and R. Chawla, "The Stealth Distributed Schedular," Proc. 11th Int'l Conf. Distributed Computing Systems, pp. 336-343, May 1991.
[16] V.M. Lo, "Task Assignment to Minimize Completion Time," Proc. Fifth Int'l Conf. Distributed Computing Systems, pp. 329-336, May 1985.
[17] S. Manoharan, "Taxonomy for Assignment in Parallel Processor Systems," Proc. Advanced Computer Technology, Reliable Systems, and Applications, Fifth Ann. European Computer Conf. CompEuro '91, pp. 143-147, May 1991.
[18] M.W. Mutka, "A Comparison of Workload Models of the Capacity Available for Sharing Among Privately Owned Workstations," Proc. 24th Ann. Hawaii Int'l Conf. System Sciences, pp. 353-362, Jan. 1991.
[19] D.M. Nicol and D.R. O'Hallaron, "Improved Algorithms for Mapping Pipelined and Parallel Computations," IEEE Trans. Computers, vol. 40, no. 3, pp. 295-306, Mar. 1991.
[20] J. Ramanujam, F. Erçal, and P. Sadayappan, "Task Allocation by Simulated Annealing," Proc. Int'l Conf. Supercomputing, vol. III, pp. 475-497, May 1988.
[21] C.-C. Shen and W.-H. Tsai, "A Graph Matching Approach to Optimal Task Assignment in Distributed Computing Systems Using a Minimax Criterion," IEEE Trans. Computers, vol. 34, no. 3, pp. 197-203, Mar. 1985.
[22] C.A. Waldspurger, T. Hogg, B.A. Huberman, J.O. Kephart, and W.S. Stornetta, Spawn: A Distributed Computational Economy IEEE Trans. Software Eng., vol. 18, no. 2, pp. 103-117, Feb. 1992.
[23] J. Wang, S. Zhou, K. Ahmed, and W. Long, "Lsbatch: A Distributed Load Sharing Batch System," Technical Report CSTI-286, Computer Systems Research Inst., Univ. of Toronto, Apr. 1993.
[24] C.M. Woodside and G.G. Monforton, “Fast Allocation of Processes in Distributed and Parallel Systems,” IEEE Trans. Parallel and Distributed Systems, vol. 4, no. 2, pp. 164-174, Feb. 1993.

Index Terms:
Task allocation, task interaction graph, heterogeneous network, shared communication medium, parallel program, minimum elapsed time.
Chi-Chung Hui, Samuel T. Chanson, "Allocating Task Interaction Graphs to Processors in Heterogeneous Networks," IEEE Transactions on Parallel and Distributed Systems, vol. 8, no. 9, pp. 908-925, Sept. 1997, doi:10.1109/71.615437
Usage of this product signifies your acceptance of the Terms of Use.