Analysis of disk power management for data-center storage systems

Tom Bostoen, Sape J. Mullender, Yolande Berbers

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    9 Citations (Scopus)

    Abstract

    Up to one third of the electricity supplied to a data center is required for the operation of the storage subsystem. A typical data-center workload, characterized by short idle periods, prevents traditional disk power management (DPM) from saving energy. This paper starts with an analysis of DPM's traditional timer-based disk spin-down policy. Next, we examine how a multispeed disk adapts DPM to data-center workloads. Finally, we determine how to shape the workload to enable DPM on conventional server disks. All analysis is based on an analytical model for the energy consumed by a disk during an idle period. In addition, we assume standby power to be non-negligible. The results are that (1) the competitive ratio of a threshold-based disk spin-down policy depends on the ratio of standby power to idle power, (2) the notion of break-even time can be generalized for multispeed disks, and (3) DPM saves most energy when mean idle time and idle-time variance are maximized. With this analysis, the authors intend to stimulate the design of new data-center file and storage systems that optimally exploit DPM to save energy.
    Original languageUndefined
    Title of host publicationProceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012)
    Place of PublicationNew York
    PublisherAssociation for Computing Machinery (ACM)
    Pages2
    Number of pages10
    ISBN (Print)978-1-4503-1055-0
    DOIs
    Publication statusPublished - 11 May 2012

    Publication series

    Name
    PublisherACM

    Keywords

    • EWI-23155
    • Reliability
    • Performance
    • METIS-296349
    • IR-85289
    • Algorithms
    • Design

    Cite this

    Bostoen, T., Mullender, S. J., & Berbers, Y. (2012). Analysis of disk power management for data-center storage systems. In Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012) (pp. 2). New York: Association for Computing Machinery (ACM). https://doi.org/10.1145/2208828.2208830
    Bostoen, Tom ; Mullender, Sape J. ; Berbers, Yolande. / Analysis of disk power management for data-center storage systems. Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012). New York : Association for Computing Machinery (ACM), 2012. pp. 2
    @inproceedings{b9ace9418f7c48be94ee74ade831cafc,
    title = "Analysis of disk power management for data-center storage systems",
    abstract = "Up to one third of the electricity supplied to a data center is required for the operation of the storage subsystem. A typical data-center workload, characterized by short idle periods, prevents traditional disk power management (DPM) from saving energy. This paper starts with an analysis of DPM's traditional timer-based disk spin-down policy. Next, we examine how a multispeed disk adapts DPM to data-center workloads. Finally, we determine how to shape the workload to enable DPM on conventional server disks. All analysis is based on an analytical model for the energy consumed by a disk during an idle period. In addition, we assume standby power to be non-negligible. The results are that (1) the competitive ratio of a threshold-based disk spin-down policy depends on the ratio of standby power to idle power, (2) the notion of break-even time can be generalized for multispeed disks, and (3) DPM saves most energy when mean idle time and idle-time variance are maximized. With this analysis, the authors intend to stimulate the design of new data-center file and storage systems that optimally exploit DPM to save energy.",
    keywords = "EWI-23155, Reliability, Performance, METIS-296349, IR-85289, Algorithms, Design",
    author = "Tom Bostoen and Mullender, {Sape J.} and Yolande Berbers",
    note = "10.1145/2208828.2208830",
    year = "2012",
    month = "5",
    day = "11",
    doi = "10.1145/2208828.2208830",
    language = "Undefined",
    isbn = "978-1-4503-1055-0",
    publisher = "Association for Computing Machinery (ACM)",
    pages = "2",
    booktitle = "Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012)",
    address = "United States",

    }

    Bostoen, T, Mullender, SJ & Berbers, Y 2012, Analysis of disk power management for data-center storage systems. in Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012). Association for Computing Machinery (ACM), New York, pp. 2. https://doi.org/10.1145/2208828.2208830

    Analysis of disk power management for data-center storage systems. / Bostoen, Tom; Mullender, Sape J.; Berbers, Yolande.

    Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012). New York : Association for Computing Machinery (ACM), 2012. p. 2.

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - Analysis of disk power management for data-center storage systems

    AU - Bostoen, Tom

    AU - Mullender, Sape J.

    AU - Berbers, Yolande

    N1 - 10.1145/2208828.2208830

    PY - 2012/5/11

    Y1 - 2012/5/11

    N2 - Up to one third of the electricity supplied to a data center is required for the operation of the storage subsystem. A typical data-center workload, characterized by short idle periods, prevents traditional disk power management (DPM) from saving energy. This paper starts with an analysis of DPM's traditional timer-based disk spin-down policy. Next, we examine how a multispeed disk adapts DPM to data-center workloads. Finally, we determine how to shape the workload to enable DPM on conventional server disks. All analysis is based on an analytical model for the energy consumed by a disk during an idle period. In addition, we assume standby power to be non-negligible. The results are that (1) the competitive ratio of a threshold-based disk spin-down policy depends on the ratio of standby power to idle power, (2) the notion of break-even time can be generalized for multispeed disks, and (3) DPM saves most energy when mean idle time and idle-time variance are maximized. With this analysis, the authors intend to stimulate the design of new data-center file and storage systems that optimally exploit DPM to save energy.

    AB - Up to one third of the electricity supplied to a data center is required for the operation of the storage subsystem. A typical data-center workload, characterized by short idle periods, prevents traditional disk power management (DPM) from saving energy. This paper starts with an analysis of DPM's traditional timer-based disk spin-down policy. Next, we examine how a multispeed disk adapts DPM to data-center workloads. Finally, we determine how to shape the workload to enable DPM on conventional server disks. All analysis is based on an analytical model for the energy consumed by a disk during an idle period. In addition, we assume standby power to be non-negligible. The results are that (1) the competitive ratio of a threshold-based disk spin-down policy depends on the ratio of standby power to idle power, (2) the notion of break-even time can be generalized for multispeed disks, and (3) DPM saves most energy when mean idle time and idle-time variance are maximized. With this analysis, the authors intend to stimulate the design of new data-center file and storage systems that optimally exploit DPM to save energy.

    KW - EWI-23155

    KW - Reliability

    KW - Performance

    KW - METIS-296349

    KW - IR-85289

    KW - Algorithms

    KW - Design

    U2 - 10.1145/2208828.2208830

    DO - 10.1145/2208828.2208830

    M3 - Conference contribution

    SN - 978-1-4503-1055-0

    SP - 2

    BT - Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012)

    PB - Association for Computing Machinery (ACM)

    CY - New York

    ER -

    Bostoen T, Mullender SJ, Berbers Y. Analysis of disk power management for data-center storage systems. In Proceedings of the third International Conference on Future Energy Systems: Where Energy, Computing and Communication Meet (e-Energy 2012). New York: Association for Computing Machinery (ACM). 2012. p. 2 https://doi.org/10.1145/2208828.2208830