• DocumentCode
    2229726
  • Title

    A Novel Intelligent Object Storage Controller for High Performance Computing

  • Author

    Guo Yu-feng ; Liu Guang-Ming ; Li, Luo ; Dou Qiang ; Luo Li

  • Author_Institution
    Sch. of Comput. Sci., Nat. Univ. of Defense Technol., Changsha, China
  • fYear
    2009
  • fDate
    26-28 Dec. 2009
  • Firstpage
    55
  • Lastpage
    58
  • Abstract
    The development of high performance computing has posed a tremendous demand on storage capacity and I/O performance, storage system has entered the Petabyte era. Application demands drive the change of storage architecture. Object-based storage is becoming the new direction of massive network storage systems. Object-based storage combines the advantages of SAN and NAS, adopts an object-based interface, and offloads some work of traditional file system to intelligent object-based storage devices, so greatly enhance the performance, reliability, and scalability of storage system. On the other hand, because the storage system scale of high performance computing is very large, the amount of storage nodes is very huge, so features such as power, size and weight have been the factors that cannot be neglected. This paper aims at the key component of object-based storage system - object storage controller, and proposes a novel design of object storage controller based on SOC. Measurement results show that the object storage controller designed by us not only optimize the performance of object-based storage system, but also has advantage in the respects of power, size and weight, so it´s very suit for building the object-based storage system for high performance computing.
  • Keywords
    input-output programs; object-oriented programming; software architecture; storage management; I/O performance; file system; high performance computing; intelligent object storage controller; object-based interface; object-based storage; petabyte era; storage architecture; storage capacity; storage system; Computer architecture; Control systems; File systems; High performance computing; Intelligent systems; Power measurement; Power system reliability; Scalability; Size measurement; Storage area networks;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Science and Engineering (ICISE), 2009 1st International Conference on
  • Conference_Location
    Nanjing
  • Print_ISBN
    978-1-4244-4909-5
  • Type

    conf

  • DOI
    10.1109/ICISE.2009.119
  • Filename
    5455413