Abstract

Abstract—With the rise of Internet-scale systems and cloud computing services, there is an increasing trend towards building massive, energy-hungry, and geographically distributed data cen-ters. Due to their enormous energy consumption, data centers are expected to have major impact on the electric grid and potentially the amount of greenhouse gas emissions and carbon footprint. In this regard, the locations that are selected to build future data centers as well as the service load to be routed to each data center after it is built need to be carefully studied given various environmental, cost, and quality-of-service considerations. To gain insights into these problems, we develop an optimization-based framework, where the objective functions range from minimizing the energy cost to minimizing the carbon footprint subject to essential quality-of-service constraints. We show that in multiple scenarios, these objectives can be conflicting leading to an energy-information tradeoff in green cloud computing. Index Terms—Green cloud computing, data centers, routing algorithms, electric grid, carbon footprint, price of electricity, price of bandwidth, carbon tax, renewable power, optimization. I.