Spatio-Temporal Load Balancing for Energy Cost Optimization in Distributed Internet Data Centers

Cloud computing is powered by an engine known as Internet data center (IDC). As cloud computing flourishes, the energy consumption and cost for IDCs are soaring. The energy cost minimization problem for IDCs in deregulated electricity markets has generated growing interest. In this paper we study how to leverage both the geographic and temporal variation of energy price to minimize energy cost for distributed IDCs. We propose a novel architecture and two algorithms for unified spatial and temporal load balancing. Rigorous analysis shows that our algorithms have a low computational complexity, require a relaxed accuracy in electricity price estimation, and guarantee a service completion time for user requests. Using real-life electricity price and workload traces, extensive evaluations demonstrate that compared to the schemes using either spatial load balancing or temporal load balancing alone, the proposed spatio-temporal load balancing method significantly reduces energy cost for distributed IDCs.