• DocumentCode
    3041430
  • Title

    Architecture of LA-MPI, a network-fault-tolerant MPI

  • Author

    Aulwes, Rob T. ; Daniel, David J. ; Desai, Nehal N. ; Graham, Richard L. ; Risinger, L. Dean ; Taylor, Mark A. ; Woodall, Timothy S. ; Sukalski, Mitchel W.

  • Author_Institution
    Adv. Comput. Lab., Los Alamos Nat. Lab., NM, USA
  • fYear
    2004
  • fDate
    26-30 April 2004
  • Firstpage
    15
  • Abstract
    Summary form only given. We discuss the unique architectural elements of the Los Alamos message passing interface (LA-MPI), a high-performance, network-fault-tolerant, thread-safe MPI library. LA-MPI is designed for use on terascale clusters which are inherently unreliable due to their sheer number of system components and trade-offs between cost and performance. We examine in detail the design concepts used to implement LA-MPI. These include reliability features, such as application-level checksumming, message retransmission, and automatic message rerouting. Other key performance enhancing features, such as concurrent message routing over multiple, diverse network adapters and protocols, and communication-specific optimizations (e.g., shared memory) are examined.
  • Keywords
    fault tolerant computing; message passing; open systems; workstation clusters; Los Alamos message passing interface; application-level checksumming; automatic message rerouting; concurrent message routing; message retransmission; network adapters; network protocols; network-fault-tolerant MPI; reliability; shared memory system; terascale clusters; Application software; Computer architecture; Fault tolerance; Fault tolerant systems; High performance computing; Laboratories; Libraries; Message passing; Protocols; Telecommunication network reliability;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Parallel and Distributed Processing Symposium, 2004. Proceedings. 18th International
  • Print_ISBN
    0-7695-2132-0
  • Type

    conf

  • DOI
    10.1109/IPDPS.2004.1302920
  • Filename
    1302920