Virtual Machine Placement Strategy Research

Abstract

With the emergence of cloud computing, users begin to have more abstract concept for the physical machine. IaaS providers prohibits users request specify the physical machine to virtual machine resources. Users know nothing about underlying infrastructure resources. The data center carry out virtual machine placement decision-making. In this paper, we use particle swarm optimization theory to guide the placement of virtual machines in the data center resources modeling and analysis. According to user needs and current physical server load state, the algorithm quantifies each physical machine and the virtual machine's CPU uses, memory uses, and storage uses. By using multi-objective particle swarm optimization algorithm, the placement of virtual machines based on the physical server resource utilization and the number of virtual machine migration. Simulation results show that the algorithm can effectively improve the minimum number of virtual machine migration resource utilization.