• DocumentCode
    3090657
  • Title

    Workload Impact on Shingled Write Disks: All-Writes Can Be Alright

  • Author

    Le, Quoc M. ; SathyanarayanaRaju, Kumar ; Amer, Ahmed ; Holliday, JoAnne

  • Author_Institution
    Dept. of Comput. Eng., Santa Clara Univ., Santa Clara, CA, USA
  • fYear
    2011
  • fDate
    25-27 July 2011
  • Firstpage
    444
  • Lastpage
    446
  • Abstract
    The data density of magnetic disk has been growing at 30% to 50% per year. However, physical limitations will slow this growth unless creative recording techniques can be found. Shingled writing is a promising technique that trades off the inability to update in-place for narrower tracks and thus a much higher data density. In this paper, we look at a variety of potential workloads, drawn from real-world traces, and evaluate their impact on a shingled write disk model. Specifically, we demonstrate the dramatically different effects that different workloads can have upon the opposing approaches of remapping and restoring blocks, and how write-heavy workloads can result in a performance advantage for a shingled-write disk.
  • Keywords
    magnetic disc storage; data density; magnetic disk; shingled write disks; write heavy workloads; Data models; Layout; Magnetic recording; NASA; Tracking; Writing; SWD; model; performance; shingled-write disk; shingling; workloads;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Modeling, Analysis & Simulation of Computer and Telecommunication Systems (MASCOTS), 2011 IEEE 19th International Symposium on
  • Conference_Location
    Singapore
  • ISSN
    1526-7539
  • Print_ISBN
    978-1-4577-0468-0
  • Type

    conf

  • DOI
    10.1109/MASCOTS.2011.58
  • Filename
    6005414