Abstract
By assuming a geographically distributed data center network (DCN) threatened by large-scale disaster, heterogeneous data backup is investigated for the threatened node (defined as the critical node) within a given early warning time, where different types of data may have their own candidate sets of safe nodes for data backup. By properly exploring the early warning time at the critical node, two backup schemes, maximum data backup and fair data backup, are proposed. The former maximizes the total amount of data that can be backed up. The latter takes fairness among different data types into account to maximize the same backup proportion (in terms of percentage of the total amount) for each data type. For each scheme, an integer linear program (ILP) and a heuristic are proposed by properly selecting a set of safe backup DCN nodes and corresponding backup routes to solve the resource competition issue. Extensive numerical results confirm the correctness of the ILP formulations and the efficiency of the heuristics.
© 2018 Optical Society of America
Full Article | PDF ArticleMore Like This
Lisheng Ma, Wei Su, Bin Wu, Tarik Taleb, Xiaohong Jiang, and Norio Shiratori
J. Opt. Commun. Netw. 9(6) 536-545 (2017)
Sifat Ferdousi, Massimo Tornatore, M. Farhan Habib, and Biswanath Mukherjee
J. Opt. Commun. Netw. 7(12) B163-B172 (2015)
Ning-Hai Bao, M. Farhan Habib, Massimo Tornatore, Charles U. Martel, and Biswanath Mukherjee
J. Opt. Commun. Netw. 7(5) 392-400 (2015)