Advertisement

Optimizations Based on Hints in a Parallel File System

  • María S. Pérez
  • Alberto Sánchez
  • Víctor Robles
  • José M. Peña
  • Fernando Pérez
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3038)

Abstract

Existing parallel file systems provide applications a little control for optimizing I/O accesses. Most of these systems use optimization techniques transparent to the applications, limiting the performance achieved by these solutions. Furthermore, there is a big gap between the interface provided by parallel file systems and the needs of applications. In fact, most of the parallel file systems do not use intuitive I/O hints or other optimizations approaches. In this sense, applications programmers cannot take advantage of optimization techniques suitable for the application domain. This paper describes I/O optimizations techniques used in MAPFS, a multiagent I/O architecture. These techniques are configured by means of a double interface for specifying access patterns or hints that increase the performance of I/O operations. An example of this interface is shown.

Keywords

Parallel I/O optimizations caching prefetching hints 

References

  1. 1.
    Carns, P.H., Ligon III, W.B., Ross, R.B., Thakur, R.: PVFS:A parallel file system for linux clusters. In: Proceedings of the 4th Annual Linux Showcase and Conference, pp. 317–327 (October 2000)Google Scholar
  2. 2.
    Cho, Y.E.: Efficient resource utilization for parallel I/O in cluster environments. PhD thesis, University of Illinois at Urbana-Champaign (1999)Google Scholar
  3. 3.
    Corbett, P., Feitelson, D., Prost, J., Johnson, S.: Parallel access to files in theVesta file system. In: Proc. of the 15th. Int. Symp. on Operating Systems Principles, pp. 472–481. ACM, New York (1993)Google Scholar
  4. 4.
    DeBenedictis, E., del Rosario, J.M.: nCUBE parallel I/O software. In: Eleventh Annual IEEE International Phoenix Conference on Computers and Communications (IPCCC), pp. 117–124 (April 1992)Google Scholar
  5. 5.
    del Rosario, J.M., Choudhary, A.N.: High-performance I/O for massively parallel computers: Problems and prospects. IEEE Computer 27(3), 59–68 (1994)Google Scholar
  6. 6.
    Elford, C., Kuszmaul, C., Huber, J., Madhyastha, T.: Scenarios for the Portable Parallel File System (1993)Google Scholar
  7. 7.
    Kotz, D., Ellis, C.S.: Practical prefetching techniques for parallel file systems. In: Proceedings of the First International Conference on Parallel and Distributed Information Systems, pp. 182–189. IEEE Computer Society Press, Los Alamitos (1991)CrossRefGoogle Scholar
  8. 8.
    Pasquale, B.K., Polyzos, G.C.: Astatic analysis of I/O characteristics of scientific applications in a productionworkload. In: Proceedings of Supercomputing 1993, pp. 388–397 (1993)Google Scholar
  9. 9.
    Patterson, R.H., Gibson, G.A., Ginting, E., Stodolsky, D., Zelenka, J.: Informed prefetching and caching. In: Jin, H., Cortes, T., Buyya, R. (eds.) High Performance Mass Storage and Parallel I/O: Technologies and Applications, pp. 224–244. IEEE Computer Society Press and Wiley, NewYork, NY (2001)Google Scholar
  10. 10.
    Pérez, M.S., Carretero, J., García, F.,, J.M.P.: Aflexible multiagent parallel file system for clusters. In: International Workshop on Parallel I/O Management Techniques (PIOMT 2003), June 2003. LNCS (2003)Google Scholar
  11. 11.
    Pérez, M.S., Pons, R.A., García, F., Carretero, J., Robles, V.: A proposal for I/O access profiles in parallel data mining algorithms. In: 3rd ACIS International Conference on SNPD (June 2002)Google Scholar
  12. 12.
    Thakur, R., Gropp, W., Lusk, E.: Optimizing noncontiguous accesses in MPI-IO. Parallel Computing 28(1), 83–105 (2002)zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • María S. Pérez
    • 1
  • Alberto Sánchez
    • 1
  • Víctor Robles
    • 1
  • José M. Peña
    • 1
  • Fernando Pérez
    • 1
  1. 1.DATSI. FI. Universidad Politécnica de MadridSpain

Personalised recommendations