• DocumentCode
    2664015
  • Title

    Leveraging big data to improve reliability & maintainability

  • Author

    Jun Li ; Reger, Brad ; Miller, Steve

  • Author_Institution
    NetApp Inc., Sunnyvale, CA, USA
  • fYear
    2015
  • fDate
    26-29 Jan. 2015
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    Big Data has brought tremendous opportunities for Reliability & Maintainability professionals. This paper shares how NetApp leverages big data from NetApp´s AutoSupportTM feature to deliver industry leading reliability and preventive maintenance. When AutoSupport is enabled, NetApp systems in the field periodically send selective system data to a NetApp corporate repository. This system data, consisting of configuration and log files with warnings, error messages and various sensor readings contains gold nuggets, which if mined and analyzed properly, can be of great benefit for improving system R&M. A few case studies are presented to show the effectiveness of the approach. These cases address the effect of cooling air temperature on hardware reliability, characteristics of DRAM memory errors, NAND flash field characteristics and preventive maintenance of defective memory modules. Results from each case also provide answers to some frequently asked questions in the computing industry.
  • Keywords
    Big Data; DRAM chips; integrated circuit reliability; preventive maintenance; production engineering computing; Big Data; DRAM memory errors; NAND flash field characteristics; NetApp AutoSupport; NetApp corporate repository; cooling air temperature; hardware reliability; maintainability; preventive maintenance; reliability; Error correction codes; Hardware; Random access memory; Reliability; Servers; Temperature distribution; Temperature sensors; big data; flash; memory; reliability; temperature;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Reliability and Maintainability Symposium (RAMS), 2015 Annual
  • Conference_Location
    Palm Harbor, FL
  • Print_ISBN
    978-1-4799-6702-5
  • Type

    conf

  • DOI
    10.1109/RAMS.2015.7105095
  • Filename
    7105095