Design and Performance of a Heterogeneous Grid Partitioner

AbstractAn important characteristic of distributed grids is that they allow geographically separated multicomputers to be tied together in a transparent virtual environment to solve large-scale computational problems. However, many of these applications require effective runtime load balancing for the resulting solutions to be viable. In this paper we present a novel latency-tolerant partitioner, called MinEX, that dynamically balances processor workloads while minimizing data movement and runtime communication for applications that are executed in a parallel distributed grid environment. We also compare the performance of MinEX with that of METIS, a popular multilevel family of partitioners, using simulated heterogeneous grid configurations. A solver for the classical N-body problem is implemented to provide a framework for the comparisons. Experimental results show that the proposed MinEX partitioner provides superior quality partitions while being competitive to METIS in terms of execution speed.

[1]  Vipin Kumar,et al.  Parallel Multilevel k-way Partitioning Scheme for Irregular Graphs , 1996, Proceedings of the 1996 ACM/IEEE Conference on Supercomputing.

[2]  Alex Pothen,et al.  Graph Partitioning Algorithms with Applications to Scientific Computing , 1997 .

[3]  Anoop Gupta,et al.  Load Balancing and Data locality in Adaptive Hierarchical N-Body Methods: Barnes-Hut, Fast Multipole, and Rasiosity , 1995, J. Parallel Distributed Comput..

[4]  Ami Marowka,et al.  The GRID: Blueprint for a New Computing Infrastructure , 2000, Parallel Distributed Comput. Pract..

[5]  R. K. Shyamasundar,et al.  Introduction to algorithms , 1996 .

[6]  Sajal K. Das,et al.  MaTCH: mapping data-parallel tasks on a heterogeneous computing platform using the cross-entropy heuristic , 2005, 19th IEEE International Parallel and Distributed Processing Symposium.

[7]  Leonid Oliker,et al.  Parallelization of a Dynamic Unstructured Algorithm Using Three Leading Programming Paradigms , 2000, IEEE Trans. Parallel Distributed Syst..

[8]  Leonid Oliker,et al.  A Comparison of Three Programming Models for Adaptive Applications on the Origin2000 , 2000, ACM/IEEE SC 2000 Conference (SC'00).

[9]  Jiming Liu,et al.  Agent-based load balancing on homogeneous minigrids: macroscopic modeling and characterization , 2005, IEEE Transactions on Parallel and Distributed Systems.

[10]  Sajal K. Das,et al.  Graph partitioning for parallel applications in heterogeneous Grid environments , 2002, Proceedings 16th International Parallel and Distributed Processing Symposium.

[11]  Andrew S. Grimshaw,et al.  The Legion Vision of a Worldwide Computer. , 1997 .

[12]  Michael P. Mesnier,et al.  The Network-Enabled Optimization System (neos) Server , 1996 .

[13]  Pangfeng Liu,et al.  Experiences with parallel N-body simulation , 1994, SPAA '94.

[14]  Henri Casanova,et al.  Netsolve: a Network-Enabled Server for Solving Computational Science Problems , 1997, Int. J. High Perform. Comput. Appl..

[15]  Leonid Oliker,et al.  Message passing and shared address space parallelism on an SMP cluster , 2003, Parallel Comput..

[16]  Sajal K. Das,et al.  FastMap: a distributed scheme for mapping large scale applications onto computational grids , 2004, Proceedings of the Second International Workshop on Challenges of Large Applications in Distributed Environments, 2004. CLADE 2004..

[17]  David Abramson,et al.  Nimrod: a tool for performing parametrised simulations using distributed workstations , 1995, Proceedings of the Fourth IEEE International Symposium on High Performance Distributed Computing.

[18]  Martin G. Everett,et al.  Parallel Dynamic Graph Partitioning for Adaptive Unstructured Meshes , 1997, J. Parallel Distributed Comput..

[19]  Xin Liu,et al.  Traffic-based Load Balance for Scalable Network Emulation , 2003, ACM/IEEE SC 2003 Conference (SC'03).

[20]  Piet Hut,et al.  A hierarchical O(N log N) force-calculation algorithm , 1986, Nature.

[21]  Miron Livny,et al.  Condor-a hunter of idle workstations , 1988, [1988] Proceedings. The 8th International Conference on Distributed.

[22]  Bruce Hendrickson,et al.  A Multi-Level Algorithm For Partitioning Graphs , 1995, Proceedings of the IEEE/ACM SC95 Conference.

[23]  Andrew B. Kahng,et al.  Recent directions in netlist partitioning: a survey , 1995, Integr..

[24]  George Cybenko,et al.  Dynamic Load Balancing for Distributed Memory Multiprocessors , 1989, J. Parallel Distributed Comput..

[25]  Ian Foster,et al.  The Grid 2 - Blueprint for a New Computing Infrastructure, Second Edition , 1998, The Grid 2, 2nd Edition.

[26]  Andrew E. Johnson,et al.  Cavern: a distributed architecture for supporting scalable persistence and interoperability in colla , 1997 .

[27]  Michael S. Warren,et al.  Fast Parallel Tree Codes for Gravitational and Fluid Dynamical N-Body Problems , 1994, Int. J. High Perform. Comput. Appl..

[28]  M. S. Warren,et al.  A parallel hashed Oct-Tree N-body algorithm , 1993, Supercomputing '93.