Automatic re-scheduling of dependencies in a RPC-based grid

ABSTRACT

RPC-based Grid infrastructures emphasize on the composition of services on a large number of computing resources. The key issue to reach high performance is to enable exploitation of parallelism on services invocations and communications. Moreover, this process should be transparent to reuse legacy codes. In this paper we present Homa an IDL compiler and a run-time support for automatic detection of the parallelism of invocations and their data dependencies on a set of CORBA objects. On homogeneous computational grids, such as clusters, Homa is accompanied by a predictable cost model. For instance, in the case of a application with a small parallel time, among p processors the speed up of Homa versus CORBA is asymptotically O(p). Also we describe how Homa can efficiently use data parallel objects. The illustrations on a case study in computational chemistry validate the cost model on a computational grid. For service-based Metacomputing, Homa offers high automation and transparency to detect parallelism for scheduling algorithms.

REFERENCES

• 1: P. H. Beckman, P. K. Fasel, W. F. Humphrey, and S. M. Mniszewski. Efficient coupling of parallel applications using paws. Technical report, Los Alamos National Laboratory, USA, 1998.
• 2: J.-P. Belaud, B. Braunschweig, and M. Pons. Open software architecture for process simulation: The current status of cape-open standard. In European Symposium on Computer Aided Process Engineering, ESCAPE-12, 2002.
• 3: P. E. Bernard and O. Coulaud. Parallel constrained molecular dynamics. INRIA Lorraine, Project NUMATH, Research report RR-3868, January 2000.
• 4: Randall Bramley , Kenneth Chiu , Shridhar Diwan , Dennis Gannon , Madhusudhan Govindaraju , Nirmal Mukhi , Benjamin Temko , Madhuri Yechuri, A Component Based Services Architecture for Building Distributed Applications, Proceedings of the Ninth IEEE International Symposium on High Performance Distributed Computing (HPDC'00), p.51, August 01-04, 2000
• 5: Denis Caromel, Toward a method of object-oriented concurrent programming, Communications of the ACM, v.36 n.9, p.90-102, Sept. 1993 [doi>10.1145/162685.162711]
• 6: Eddy Caron , Frederic Desprez , Frédéric Lombard , Jean-Marc Nicod , Laurent Philippe , Martin Quinson , Frédéric Suter, A Scalable Approach to Network Enabled Servers (Research Note), Proceedings of the 8th International Euro-Par Conference on Parallel Processing, p.907-910, August 27-30, 2002
• 7: H. Casanova and J. Dongarra. NetSolve: A network server for solving computational science problems. In Workshop of Vector and Parallel Computing, Manno, Switzerland, March 1997. SPEEDUP Society.
• 8: Gregory V. Chockler , Danny Dolev , Roy Friedman , Roman Vitenberg, Implementing a caching service a distributed COBRA objects, IFIP/ACM International Conference on Distributed systems platforms, p.1-23, April 03-07, 2000, New York, New York, United States
• 9: Dorian C. Arnold , Dieter Bachmann , Jack Dongarra, Request Sequencing: Optimizing Communication for the Grid, Proceedings from the 6th International Euro-Par Conference on Parallel Processing, p.1213-1222, August 29-September 01, 2000
• 10: Steven Fortune , James Wyllie, Parallelism in random access machines, Proceedings of the tenth annual ACM symposium on Theory of computing, p.114-118, May 01-03, 1978, San Diego, California, United States [doi>10.1145/800133.804339]
• 11: François Galilée , Jean-Louis Roch , Gerson G. H. Cavalheiro , Mathias Doreille, Athapascan-1: On-Line Building Data Flow Graph in a Parallel Language, Proceedings of the 1998 International Conference on Parallel Architectures and Compilation Techniques, p.88, October 12-18, 1998
• 12: T. Gautier and H. R. Hamidi. Homa: automatic re-scheduling of multiple invocations in corba. INRIA Rhône-Alpes, projet APACHE, Research Report, 2004.
• 13: T. Gautier, R. Revire, and J.-L. Roch. Athapascan: Api for asynchronous parallel programming. Technical Report RR-0276, INRIA Rhône-Alpes, projet APACHE, February 2003.
• 14: M. G. Hackenberg, P. Post, R. Redler, and B. Steckel. Mpcci, multidisciplinary applications and multigrid. In ECCOMAS 2000, CIMNE Barcelona, 2000.
• 15: IONA. Orbix Programming Guide IONA Technology Ltd., 1995.
• 16: Katarzyna Keahey , Dennis Gannon, PARDIS: A Parallel Approach to CORBA, Proceedings of the 6th International Symposium on High Performance Distributed Computing (HPDC '97), p.31, August 05-08, 1997
• 17: R. Kordale, M. Ahamad, and M. Devarkonda. Object caching in a corba compliant system. USENIX Computing Systems Journal, 9(4), 1996.
• 18: I. Lopez, G. J. Follen, R. Gutierrez, I. Foster, B. Ginsburg, O. Larsson, and S. Tuecke. Using corba and globus to coordinate multidisciplinary aeroscience applications. In Proceedings of the NASA HPCC/CAS Workshop, pages 15--17, February 2000.
• 19: Paul Martin , Martlesham Heath , Victor Callaghan , Adrian Clark, High Performance Distributed Objects Using Caching Proxies for Large Scale Applications, Proceedings of the International Symposium on Distributed Objects and Applications, p.110, September 05-07, 1999
• 20: Satoshi Matsuoka , Mitsuhisa Sato , Hidemoto Nakada , Satoshi Sekiguchi, Design Issues of Network Enabled Server Systems for the Grid, Proceedings of the First IEEE/ACM International Workshop on Grid Computing, p.4-17, December 17, 2000
• 21: OMG. Corba component model. Technical report, OMG, formal/2002-06-65, 2002.
• 22: OMG. Data parallel object. Technical report, OMG formal/2002-06-65, 2002.
• 23: M. Parashar, G. von Laszewski, S. Verma, J. Gawor, K. Keahey, and N. Rehn. A corba commodity grid kit. In Concurrency and Computation: Practice and Experience, John Wiley and Sons, 2002.
• 24: Christian Pérez , Thierry Priol , André Ribes, A Parallel CORBA Component Model for Numerical Code Coupling, Proceedings of the Third International Workshop on Grid Computing, p.88-99, November 18, 2002
• 25: Mitsuhisa Sato , Hidemoto Nakada , Satoshi Sekiguchi , Satoshi Matsuoka , Umpei Nagashima , Hiromitsu Takagi, Ninf: A Network Based Information Library for Global World-Wide Computing Infrastructure, Proceedings of the International Conference and Exhibition on High-Performance Computing and Networking, p.491-502, April 28-30, 1997
• 26: D. Schmidt, A. Gokhale, T. Harrison, D. Levine, and C. Cleeland. Tao: A high-performance endsystem architecture for real-time corba. In IEEE Communications Magazine feature topic issue on Distributed Object Computing, February 1997.

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.2 COMPUTER-COMMUNICATION NETWORKS
      C.2.4 Distributed Systems
          Subjects: Distributed applications

Additional Classification:
  D. Software
  D.1 PROGRAMMING TECHNIQUES
      D.1.3 Concurrent Programming
  D.2 SOFTWARE ENGINEERING
        D.2.12 Interoperability
  D.4 OPERATING SYSTEMS
      D.4.8 Performance
          Subjects: Modeling and prediction

General Terms:
Design, Performance

Keywords:
CORBA, distributed, interoperability, metacomputing, parallel