Observing the Impact of Multiple Metrics and Runtime Adaptations on BSP Process Rescheduling

Process rescheduling is an useful mechanism to offer runtime load balancing, mainly in dynamic and heterogeneous environments. In this context, we developed a model called MigBSP which controls the process migration on BSP (Bulk Synchronous Parallel) applications. A BSP application is divided in one or more supersteps, each one containing both computation and communication phases followed by a barrier synchronization. Since the barrier waits for the slowest process, MigBSP's final objective is to adjust the processes location in order to reduce the supersteps' times. Its novel ideas are twofold. The former is represented by the combination of three metrics - Memory, Computation and Communication - in order to measure the Potential of Migration of each BSP process. The second idea consists in offering efficient adaptations that work on the rescheduling frequency. Both ideas turn MigBSP a viable model for getting performance on BSP applications. Meanwhile, it provides a low overhead on application execution when migrations do not take place. This paper presents MigBSP's algorithms, the parallel machine organization, some experimental results and related work.

[1]  Fabio Kon,et al.  InteGrade object-oriented Grid middleware leveraging the idle computing power of desktop machines: Research Articles , 2004 .

[2]  Uwe Schwiegelshohn,et al.  New Challenges of Parallel Job Scheduling , 2007, JSSPP.

[3]  Jin-Soo Kim,et al.  DynaGrid: An Adaptive, Scalable, and Reliable Resource Provisioning Framework for WSRF-Compliant Applications , 2008, Journal of Grid Computing.

[4]  Leslie G. Valiant,et al.  A bridging model for parallel computation , 1990, CACM.

[5]  Fabio Kon,et al.  InteGrade: object‐oriented Grid middleware leveraging the idle computing power of desktop machines , 2004, Concurr. Pract. Exp..

[6]  Sathish S. Vadhiyar,et al.  Self adaptivity in Grid computing , 2005, Concurr. Pract. Exp..

[7]  Olaf Bonorden,et al.  Load Balancing in the Bulk-Synchronous-Parallel Setting using Process Migrations , 2007, 2007 IEEE International Parallel and Distributed Processing Symposium.

[8]  Thomas L. Casavant,et al.  A Taxonomy of Scheduling in General-Purpose Distributed Computing Systems , 1988, IEEE Trans. Software Eng..

[9]  Jack Dongarra,et al.  Self adaptivity in Grid computing: Research Articles , 2005 .

[10]  Bruno Schulze,et al.  Heterogeneous Strong Computation Migration , 2006, ArXiv.

[11]  Laxmikant V. Kalé,et al.  Performance evaluation of adaptive MPI , 2006, PPoPP '06.

[12]  Michael Schmitz,et al.  Decentralized Dynamic Load Balancing: The Particles Approach , 1995, Inf. Sci..

[13]  Weiqin Tong,et al.  Agent based ServiceBSP Model with Superstep Service for Grid Computing , 2007, Sixth International Conference on Grid and Cooperative Computing (GCC 2007).

[14]  Bruno Schulze,et al.  State of the art in heterogeneous strong migration of computations , 2008 .

[15]  Rajeev Motwani,et al.  The load rebalancing problem , 2006, J. Algorithms.

[16]  Vasil P. Vasilev.  BSPGRID: Variable Resources Parallel Computation and Multiprogrammed Parallelism , 2003, Parallel Process. Lett..

[17]  Francine Berman,et al.  Models and scheduling mechanisms for global computing applications , 2002, Proceedings 16th International Parallel and Distributed Processing Symposium.

[18]  Laxmikant V. Kalé,et al.  Run-Time Support for Adaptive Load Balancing , 2000, IPDPS Workshops.

[19]  Philippe Olivier Alexandre Navaux,et al.  Applying Processes Rescheduling over Irregular BSP Application , 2009, ICCS.

[20]  Alfons Kemper,et al.  Migrating Autonomous Objects in a WAN Environment , 2000, Journal of Intelligent Information Systems.

[21]  Francisco Almeida,et al.  Dynamic Load Balancing on Dedicated Heterogeneous Systems , 2008, PVM/MPI.

[22]  Claudio Schepke,et al.  Performance Improvement of the Parallel Lattice Boltzmann Method Through Blocked Data Distributions , 2007, 19th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD'07).

[23]  Rafael Moreno-Vozmediano,et al.  Influence of Grid Economic Factors on Scheduling and Migration , 2004, VECPAR.

[24]  Fabio Kon,et al.  Portable checkpointing and communication for BSP applications on dynamic heterogeneous grid environments , 2005, 17th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD'05).

[25]  Load Balancing Strategies in a Web Computing Environment , 2005, PPAM.

[26]  Cong Du,et al.  Dynamic Scheduling with Process Migration , 2007, Seventh IEEE International Symposium on Cluster Computing and the Grid (CCGrid '07).