• DocumentCode
    652893
  • Title

    Longer Is Better: Exploiting Path Diversity in Data Center Networks

  • Author

    Fung Po Tso ; Hamilton, Gregg ; Weber, R. ; Perkins, Colin S. ; Pezaros, Dimitrios P.

  • Author_Institution
    Sch. of Comput. Sci., Univ. of Glasgow, Glasgow, UK
  • fYear
    2013
  • fDate
    8-11 July 2013
  • Firstpage
    430
  • Lastpage
    439
  • Abstract
    Data Center (DC) networks exhibit much more centralized characteristics than the legacy Internet, yet they are operated by similar distributed routing and control algorithms that fail to exploit topological redundancy to deliver better and more sustainable performance. Multipath protocols, for example, use node-local and heuristic information to only exploit path diversity between shortest paths. In this paper, we use a measurement-based approach to schedule flows over both shortest and non-shortest paths based on temporal network-wide utilization. We present the Baatdaat flow scheduling algorithm which uses spare DC network capacity to mitigate the performance degradation of heavily utilized links. Results show that Baatdaat achieves close to optimal Traffic Engineering by reducing network-wide maximum link utilization by up to 18% over Equal-Cost Multi-Path (ECMP) routing, while at the same time improving flow completion time by 41% - 95%.
  • Keywords
    routing protocols; telecommunication traffic; Baatdaat flow scheduling algorithm; data center network; distributed routing; heuristic information; measurement-based approach; multipath protocol; network-wide maximum link utilization; node-local information; optimal traffic engineering; path diversity; spare DC network capacity; temporal network-wide utilization; topological redundancy; Control systems; Hardware; Network topology; Ports (Computers); Schedules; Throughput; Topology; Data Center Networks; Multipath Scheduling; Software Defined Networking; Traffic Engineering; Unequal-cost Multipath;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Distributed Computing Systems (ICDCS), 2013 IEEE 33rd International Conference on
  • Conference_Location
    Philadelphia, PA
  • ISSN
    1063-6927
  • Type

    conf

  • DOI
    10.1109/ICDCS.2013.36
  • Filename
    6681612