Joint virtual machine allocation and power portfolio optimization for data centers in smart grid environment

Generally, data centers consume a great amount of electric power which incurs the major operating cost to a data center owner. Currently, smart grid whose one of the important features is the realtime pricing will be implemented by the public utility in a near future such that the owner could also encounter a risk of fluctuating electricity prices (i.e., spot prices) in electricity spot markets. To hedge against such a risk, the owner can sign forward contracts from electricity futures markets. In this paper, a stochastic programming model is formulated to jointly optimize the power cost in the electricity markets and the operating cost of virtual machine allocation in data centers. Numerical studies are performed to evaluate the model. The results clearly show that the proposed model can significantly reduce the cost of operating data centers under uncertainties of demand and power prices.