Minimizing the Energy Cost of Servers in Data Center Networks

To alleviate the growing concern of energy waste in networked devices, we propose an energy saving strategy in the "server-centric" data center network. Focusing on servers, we introduce a basic scheme based on the transition of automata. By changing efficiently from sleeping state to acti...

Full description

Saved in:
Bibliographic Details
Published in2012 International Conference on Cloud and Green Computing pp. 311 - 318
Main Authors Jun Wu, Qin Jia, Jin Zhao, Xin Wang
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.11.2012
Subjects
Online AccessGet full text
ISBN1467330272
9781467330275
DOI10.1109/CGC.2012.51

Cover

More Information
Summary:To alleviate the growing concern of energy waste in networked devices, we propose an energy saving strategy in the "server-centric" data center network. Focusing on servers, we introduce a basic scheme based on the transition of automata. By changing efficiently from sleeping state to active state in different network traffic, servers save the energy. This scheme is self-adaptive and simple. Based on this prime scheme, we enhance the strategy by further taking advantage of the unique topology of BCube. We present a structure algorithm relying on the active substructure in BCube. Combining with the basic scheme and structure algorithm, our energy saving strategy can achieve a high proportion of energy saving without sacrificing service quality like packet delay in the simulation with real traffic trace.
ISBN:1467330272
9781467330275
DOI:10.1109/CGC.2012.51