Journal article

Simple and effective dynamic provisioning for power-proportional data centers

T Lu, M Chen, LLH Andrew

IEEE Transactions on Parallel and Distributed Systems | IEEE COMPUTER SOC | Published : 2013

Abstract

Energy consumption represents a significant cost in data center operation. A large fraction of the energy, however, is used to power idle servers when the workload is low. Dynamic provisioning techniques aim at saving this portion of the energy, by turning off unnecessary servers. In this paper, we explore how much gain knowing future workload information can bring to dynamic provisioning. In particular, we develop online dynamic provisioning solutions with and without future workload information available. We first reveal an elegant structure of the offline dynamic provisioning problem, which allows us to characterize the optimal solution in a divide-and-conquer manner. We then exploit this..

View full abstract

University of Melbourne Researchers

Grants

Awarded by China National 973 project


Awarded by University Grants Committee of the Hong Kong Special Administrative Region, China


Awarded by General Research Fund


Awarded by Australian Research Council


Funding Acknowledgements

The authors thank Minghong Lin for sharing the code of his LCP algorithm, and Eno Thereska for sharing the MSR Cambridge data center traces. A preliminary version of the paper appeared in IEEE CISS in 2012 [1]. The work described in this paper was partially supported by two China National 973 projects (grant No. 2012CB315904 and 2013CB336700), several grants from the University Grants Committee of the Hong Kong Special Administrative Region, China (Area of Excellence Project No. AoE/E-02/08 and General Research Fund Project No. 411010 and 411011), and two gift grants from Microsoft and Cisco. The work was also partially funded by the Australian Research Council grant FT0991594.