Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17–18, 2017, Proceedings

Research Article

Reliability Evaluation of DCell Networks

Download
168 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-73317-3_26,
        author={Xi Wang and Funan He and Yuejuan Han and Lantao You},
        title={Reliability Evaluation of DCell Networks},
        proceedings={Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17--18, 2017, Proceedings},
        proceedings_a={ADHIP},
        year={2018},
        month={2},
        keywords={DCell network Reliability Extra connectivity Data center network},
        doi={10.1007/978-3-319-73317-3_26}
    }
    
  • Xi Wang
    Funan He
    Yuejuan Han
    Lantao You
    Year: 2018
    Reliability Evaluation of DCell Networks
    ADHIP
    Springer
    DOI: 10.1007/978-3-319-73317-3_26
Xi Wang,*, Funan He1, Yuejuan Han2, Lantao You3
  • 1: Suzhou Institute of Industrial Technology
  • 2: Soochow University
  • 3: Suzhou Industrial Park Institute of Services Outsourcing
*Contact email: wangxi0414@163.com

Abstract

Recently, the reliability evaluation of data center network (DCN) is important to the design and operation of DCNs. Extra connectivity determination and faulty networks structure analysis are two significant aspects for the reliability evaluation of DCNs. The DCell network is suitable for a massive data centers with high network capacity by only using cheap switches. A -dimensional DCell built with -port switches, denoted by , is an -regular graph. In this paper, we firstly prove that the extra- connectivity of for , if and , and if and , respectively. What’s more, for any faulty node set with , we obtain that there contains a large connected component in , and the rest of small connected components have not more than nodes in total if and (resp. and ). Our result can provide a proper measure for the reliability evaluation of the DCell network when it is used to model the topological structure of a large-scale DCN.