Efficient trigger-broadcasting in heterogeneous clusters
Pages 628 - 642
Abstract
Broadcasts in parallel computing environments are often used to trigger ''personal'' computations at the processors (or, nodes) that comprise the system. (The qualifier ''personal'' means that the triggered computations may differ in type and complexity at each node.) We present an algorithm for trigger-broadcasting in a node-heterogeneous cluster of workstations, which comes predictably close to minimizing the time for completing both the broadcast and the computations it triggers. The algorithm orchestrates its broadcast taking account of: the speeds of the cluster's constituent workstations, the speed of the cluster's network, and the complexities of the computations that the broadcast triggers. The algorithm is within a constant factor of optimal when the speeds of the cluster's workstations and of its network are independent of the number of workstations. The algorithm is exactly optimal when the cluster is homogeneous-no matter how diverse the ''personal'' computations are.
References
[1]
{1} M. Adler, Y. Gong, A.L. Rosenberg, Optimal sharing of bags of tasks in heterogeneous clusters, 15th ACM Symposium on Parallelism in Algorithms and Architectures (SPAA'03), 2003, pp. 1-10.]]
[2]
{2} A. Alexandrov, M.I. Ionescu, K.E. Schauser, C. Scheiman, LogGP: incorporating long messages into the LogP model for parallel computation, J. Parallel Distrib. Comput. 44 (1997) 71-79.]]
[3]
{3} T.E. Anderson, D.E. Culler, D.A. Patterson, The Cluster Team, A case for NOW (networks of workstations), IEEE Micro 15 (1995) 54-64.]]
[4]
{5} G. Ausiello, P. Crecensi, G. Gambosi, V. Kann, A. Marchetti-Spaccamela, M. Protasi, Complexity and Approximation, Springer, Berlin, 1999.]]
[5]
{6} M. Banikazemi, V. Moorthy, D.K. Panda, Efficient collective communication on heterogeneous networks of workstations, International Conference on Parallel Processing, 1998, pp. 460-467.]]
[6]
{7} A. Bar-Noy, S. Guha, J. Naor, B. Schieber, Message multicasting in heterogeneous networks, SIAM J. Comput. 30 (2000) 347-358.]]
[7]
{8} A. Bar-Noy, S. Kipnis, Designing broadcasting algorithms in the postal model for message-passing systems, Math. Systems Theory 27 (1994) 431-452.]]
[8]
{9} D. Barth, P. Fraigniand, Approximation algorithms for structured communication problems, Ninth ACM Symposium on Parallel Algorithms and Architectures, 1997, pp. 180-188. (See also: Tech. Rpt. 1239, LRI, Univ. Paris-Sud. (http://www.lri.fr/~pierre)).]]
[9]
{10} P.B. Bhat, V.K. Prasanna, C.S. Raghavendra, Adaptive communication algorithms for distributed heterogeneous systems, J. Parallel Distrib. Comput. 59 (1999) 252-279.]]
[10]
{11} P.B. Bhat, V.K. Prasanna, C.S. Raghavendra, Efficient collective communication in distributed heterogeneous systems, 19th IEEE International Conference on Distributed Computing and Systems, 1999, pp. 15-24.]]
[11]
{12} F. Cappello, P. Fraigniaud, B. Mans, A.L. Rosenberg, An algorithmic model for heterogeneous clusters: rationale and experience. Intl. J. Foundations of Computer Science, to appear, see also: HiHCoHP--toward a realistic communication model for hierarchical hyperclusters of heterogeneous processors, International Parallel and Distributed Proceedings Symposium (IPDPS'01), 2001.]]
[12]
{13} J. Cohen, P. Fraigniaud, J.-C. Konig, A. Raspaud, Broadcasting and multicasting in cut-through routed networks, 11th IEEE International Parallel Processing Symposium, 1997, pp. 734-738.]]
[13]
{14} D. Culler, R.M. Karp, D. Patterson, A. Sahay, K.E. Schauser, E. Santos, R. Subramonian, T. Von Eicken, LogP: towards a realistic model of parallel computation, Comm. ACM 39 (1996) 78-85.]]
[14]
{15} X. Du, X. Zhang, Coordinating parallel processes on networks of workstations, J. Parallel Distrib. Comput. 46 (1997) 125-135.]]
[15]
{16} M. Elkin, G. Kortsarz, Combinatorial logarithmic approximation algorithm for directed telephone broadcast problem, 34th ACM Symposium on Theory of Computing, 2002.]]
[16]
{17} M. Elkin, G. Kortsarz, Sublogarithmic approximation for telephone multicast: path out of jungle, 14th ACM-SIAM Symposium on Discrete Algorithms, 2003, pp. 76-85.]]
[17]
{18} M. Elkin, G. Kortsarz, Approximation algorithm for directed telephone multicast problem, 30th International Colloquium on Automata, Languages and Programming, 2003, pp. 212-223.]]
[18]
{19} P. Fraigniaud, Approximation algorithms for minimum-time broadcast under the vertex-disjoint paths mode, in: 19th European Symposium on Algorithms (ESA '01), Lecture Notes in Computer Science, vol. 2161, Springer, Berlin, 2001, pp. 440-451.]]
[19]
{20} P. Fraigniaud, E. Lazard, Methods and problems of communication in usual networks, Discrete Appl. Math. 53 (1994) 79-133.]]
[20]
{21} P. Fraigniaud, S. Vial, Approximation algorithms for broadcasting and gossiping, J. Parallel Distrib. Comput. 43 (1997) 47-55.]]
[21]
{22} M.R. Garey, D.S. Johnson, Computers and Intractability, W.H. Freeman, San Francisco, 1979.]]
[22]
{23} N.G. Hall, W.-P Liu, J.B. Sidney, Scheduling in broadcast networks, Networks 32 (1998) 233-253.]]
[23]
{24} S.M. Hedetniemi, S.T. Hedetniemi, A. Liestman, A survey of gossiping and broadcasting in communication networks, Networks 18 (1986) 319-349.]]
[24]
{25} J. Hromkovič, R. Klasing, B. Monien, R. Peine, Dissemination of information in interconnection networks (broadcasting and gossiping), in: F. Hsu, D.-Z. Du (Eds.), Combinatorial Network Theory, Science Press, American Mathematical Society, Providence, RI, 1997, pp. 125-212.]]
[25]
{26} S.L. Johnsson, C.-T. Ho, Optimum broadcasting and personalized communication in hypercubes, IEEE Trans. Comput. 38 (1989) 1249 -1268.]]
[26]
{27} R.M. Karp, A. Sahay, E. Santos, K.E. Schauser, Optimal broadcast and summation in the LogP model, Fifth ACM Symposium on Parallel Algorithms and Architectures, 1993, pp. 142-153.]]
[27]
{28} R. Kesavan, K. Bondalapati, D.K. Panda, Multicast on irregular switch-based networks with wormhole routing, Third International Symposium on High-Performance Computer Architecture, 1997, pp. 48-57.]]
[28]
{29} J.-S. Kim, S. Ha, C.S. Jhon, Efficient barrier synchronization mechanism for the BSP model on message-passing architectures, 12th IEEE International Parallel Processing Symposium, 1998, pp. 255-259.]]
[29]
{30} G. Kortsarz, D. Peleg, Approximation algorithms for minimum time broadcast, SIAM J. Discrete Math. 8 (1995) 401-427.]]
[30]
{31} R. Libeskind-Hadas, J. Hartline, Efficient multicast in heterogeneous networks of workstations, International Conference on Parallel Proceedings, 2000, pp. 403-410.]]
[31]
{32} R. Libeskind-Hadas, J. Hartline, P. Boothe, G. Rae, J. Swisher, On multicast algorithms for heterogeneous networks of workstations, J. Parallel Distrib. Comput. 61 (2001) 1665-1679.]]
[32]
{33} R. Libeskind-Hadas, D. Mazzoni, R. Rajagopalan, Optimal contention-free unicast-based multicasting in switch-based networks of workstations, 12th IEEE International Parallel Processing Symposium, 1998, pp. 358-364.]]
[33]
{34} P. Liu, T.-H. Sheng, Broadcast scheduling optimization for heterogeneous clusters systems, 12th ACM Symposium on Parallel Algorithms and Architectures, 2000, pp. 129-136.]]
[34]
{35} G.F Pfister, In Search of Clusters, Prentice-Hall, Englewood Cliffs, NJ, 1995.]]
[35]
{37} A.L. Rosenberg, On sharing bags of tasks in heterogeneous networks of workstations: greedier is not better, Third IEEE International Conference on Cluster Computing (Cluster'01), Newport Beach, CA, 2001, pp. 124-131.]]
[36]
{39} A.S. Tosun, A. Agarwal, Efficient broadcast algorithms for heterogeneous networks of workstations, 13th International Conference on Parallel and Distributed Computing Systems, 2000.]]
Index Terms
- Efficient trigger-broadcasting in heterogeneous clusters
Recommendations
A dynamic and reliability-driven scheduling algorithm for parallel real-time jobs executing on heterogeneous clusters
In this paper, a heuristic dynamic scheduling scheme for parallel real-time jobs executing on a heterogeneous cluster is presented. In our system model, parallel real-time jobs, which are modeled by directed acyclic graphs, arrive at a heterogeneous ...
A Parallel Computational Model for Heterogeneous Clusters
Heterogeneous clusters claim for new models and algorithms. In this paper, a new parallel computational model is presented. The model, based on the LogGP model, has been extended to be able to deal with heterogeneous parallel systems. For that purpose, ...
Energy-Aware Task Clustering Scheduling Algorithm for Heterogeneous Clusters
GREENCOM '11: Proceedings of the 2011 IEEE/ACM International Conference on Green Computing and CommunicationsHigh performance heterogeneous clusters have been widely used to commercial and scientific areas. However, huge energy consumption has prevented the further application of large-scale heterogeneous clusters. It is highly desirable to design energy-aware ...
Comments
Information & Contributors
Information
Published In
Copyright © Elsevier Inc. © 2005.
Publisher
Academic Press, Inc.
United States
Publication History
Published: 01 May 2005
Author Tags
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 08 Feb 2025