Communication-Aware and Energy Saving Virtual Machine Allocation Algorithm in Data Center.

Data centers can provide users with reliable and on-demand cloud services. Efficient virtual machine allocation strategies are important for improving data center performance and increasing allocation efficiency. However, the existing allocation algorithms have limited consideration for the communication delay between virtual machines caused by the data center topology. Therefore, this study formalizes the network communication problem of virtual machines in data centers by studying these centers network topology. Accordingly, this paper proposes an energy saving allocation algorithm based on communication-aware, called ESBCA, which optimizes the data center traffic from the two stages of virtual machine migration and placement. In the migration phase, virtual machines with high traffic volumes are preferred to be migrated. In the placement stage, different targets are set and the destination physical machine is systematically determined. Experimental results show that the proposed algorithm evidently reduces the data traffic, improves the efficiency of the data center migration and decreases the energy consumption of data centers.