• DocumentCode
    2290223
  • Title

    The design and performance evaluation of the RAID 5 controller using the load-balanced destage algorithm

  • Author

    Chang, Yun-Seok ; Kim, Chong-Sang

  • Author_Institution
    Dept. of Comput. Eng., Seoul Nat. Univ., South Korea
  • fYear
    1997
  • fDate
    10-13 Dec 1997
  • Firstpage
    28
  • Lastpage
    34
  • Abstract
    Write requests are written from the disk cache to the disks by the destage algorithm, and the response time of host read request dominates the performance of the disk. Since RAID is composed of multiple disks, although the performance at the disk level is important to service requests, effectively the performance at the disk array level is more important. In RAID, a host request cannot be completed until all the striped requests are completed and the response time of the host request is dependent on the response time of the disk with the heaviest load. However, existing destage algorithms do not take into consideration the overall performance of all disks in the RAID but destage write requests for optimal performance at each individual disk, and it may eventually lead to overload of a few disks. It may delay the service of some striped requests, and therefore, the response time of host request increases. The paper suggests a novel Load-Balanced Destage (LBD) algorithm adopted at the disk array level, and shows that the LBD algorithm has a higher performance than existing destage algorithms by evaluating their performance using a simulator
  • Keywords
    cache storage; magnetic disc storage; performance evaluation; resource allocation; storage management; LBD algorithm; RAID 5 controller; Redundant Arrays of Inexpensive Disks; destage algorithms; destage write requests; disk array level; disk cache; disk performance; host read request; host request; load balanced destage algorithm; optimal performance; performance evaluation; response time; service requests; simulator; striped requests; write requests; Algorithm design and analysis; Bandwidth; Computational modeling; Delay effects; Fault tolerance; Hard disks; Operating systems; Read-write memory; Redundancy;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Parallel and Distributed Systems, 1997. Proceedings., 1997 International Conference on
  • Conference_Location
    Seoul
  • Print_ISBN
    0-8186-8227-2
  • Type

    conf

  • DOI
    10.1109/ICPADS.1997.652526
  • Filename
    652526