• DocumentCode
    589586
  • Title

    Event-notification-based inactive file search for large-scale file systems

  • Author

    Takata, Masami ; Sutoh, A.

  • Author_Institution
    Yokohama Res. Lab., Hitachi, Ltd., Yokohama, Japan
  • fYear
    2012
  • fDate
    Oct. 31 2012-Nov. 2 2012
  • Firstpage
    1
  • Lastpage
    7
  • Abstract
    Finding inactive files is a challenging task in large-scale file systems. In a hierarchical storage management system or a cloud storage gateway system, inactive files are migrated to lower tiers to reduce the storage cost. Inactive files can also be applied the data reduction techniques such as deduplication and compression. Existing metadata search methods, using either file system crawling or metadata indexing, degrade file access performance because they require a large number of disk I/Os. The two-phase refining (TPR) we propose here searches for inactive files effectively by using file system event notification and pseudo-LRU lists. TPR consists of two processes: a listing process that lists the files created each day, and a searching process that refines files from the oldest pseudo-LRU list in order to identify files that can be migrated or subjected to data reduction technologies. Experimental results show that our method is superior to existing methods in terms of search time, file access performance, and space overhead.
  • Keywords
    data reduction; storage management; cloud storage gateway system; compression; data reduction technique; deduplication; event-notification; file access performance; file system crawling; file system event notification; hierarchical storage management system; inactive file search; large-scale file system; metadata indexing; metadata search method; pseudo-LRU list; search time; space overhead; two-phase refining; Local area networks; Logic gates; Metadata search; cloud storage systems; file storage systems; file systems;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    APMRC, 2012 Digest
  • Conference_Location
    Singapore
  • Print_ISBN
    978-1-4673-4734-1
  • Type

    conf

  • Filename
    6407531