基于架构负载感知的虚拟机聚簇部署算法

针对云计算中虚拟机部署问题，提出了一种基于架构负载感知的虚拟机聚簇部署算法。首先计算云数据中心各层架构的负载，并在架构内对主机进行聚簇。虚拟机进行部署时，先按照相应的规则进行虚拟机间的聚簇，并优先选择负载较低的架构进行部署，架构选择后，进行虚拟机簇与主机簇的匹配以选择最优的主机簇进行部署。最后通过CloudSim进行仿真验证，将其与贪婪算法及基于架构负载感知的基本部署算法进行比较，证明了算法在部署时间方面有明显的优越性，并提高了网络资源的利用率。

Deployment of virtual machines with clustering method based on frame load awareness

Concerning the deployment of virtual machines in the cloud computing, an algorithm on the deployment of virtual machines with clustering method based on frame load awareness was proposed. First of all, the load of each layer in datacenter was computed and the clustering of physical machines in each layer was constructed. In the process of deploying virtual machines, the clustering of virtual machines was first done according to some rules and then the frame with lower load was chosen preferentially. The last step was to match the virtual machines cluster and physical machines cluster in order to deploy the optimal physical machines cluster. The performance of the algorithm was validated with the experiments in CloudSim. The result was compared to that of the greedy algorithm and basic deployment algorithm with the frame load awareness, which shows that the algorithm proposed in this article has evident priority, and improves the utilization rate of network resources.