Energy saving virtual machine allocation in cloud computing


In the data center, a server can work in either active state or power-saving state. The power consumption in the power-saving state is almost 0, thus it is always desirable to allocate as many VMs as possible to some active servers and leave the rest to power-saving state in order to reduce the energy consumption of the data center. In this paper, we study such a VM allocation problem. Given a set VMs and a set of servers in a data center, each VM has a resource demand (CPU, memory, storage) and a starting time and a finishing time, and each server has resource capacity. There is an additional energy cost for a server to switch from power-saving state to active state. The servers are non-homogeneous. The problem of our concern is to allocate the VMs onto servers, such that the VMs resource demands can be met and the total energy consumption of servers is minimized. The problem is formulated as a boolean integer linear programming problem. A heuristic algorithm is proposed to solve the problem. Extensive simulations have been conducted to demonstrate our proposed method can significantly save the energy consumption in data centers. © 2013 IEEE.

Publication Title

Proceedings - International Conference on Distributed Computing Systems