Energy-Aware Dynamic Server Provisioning and Frequency Adjustment in Multi-Tier Data Centers

Abstract

Data centers generally consume an enormous amount of energy, which not only increases the running cost but also simultaneously enhances their greenhouse gas emissions. This paper explores the benefits of Dynamic Voltage/Frequency Scaling (DVFS) and Vary-On Vary-Off (VOVO) for energy management in multi-tier data centers. Unlike previous work, which addressed the DVFS and VOVO separately. In this paper, an optimization problem is formulated to achieve energy benefits while maintaining performance in multi-tier data centers. DVFS and VOVO technology are combined to yield an optimal energy management strategy. From a user's perspective, only the end-to-end delay across the entire data center is relevant. The proposed optimization method offers controllable and predictable quantitative control of energy consumption with theoretically guaranteed end-to-end delay.