Load Similarity based Virtual Machine Placement in Cloud Computing Environment

Abstract

The emerging cloud applications run on physical machines in cloud environments in the form of virtual machines. Therefore, the placement strategy of virtual machines greatly affects the performance and efficiency of data centers. To solve the virtual machine placement problem in the scheduling process of the data center, this paper presents a virtual machine placement algorithm based on load similarity, by taking into account the load balance and physical resource utilization of the data center. Our algorithm manages an ideal map from virtual machines to physical machines by using the load similarity to determine the placement strategy that is the closest to the ideal case. The experimental results show that our method outperforms existing solutions by 10%-15% in load balance and resource utilization.