云环境下基于性能向量的虚拟机部署算法

Performance vector-based algorithm for virtual machine deployment in infrastructure clouds

针对云计算环境下虚拟机部署问题,提出充分考虑了系统负载均衡的PM-LB虚拟机部署算法。首先,采用性能向量,规范化地描述虚拟基础设施性能状况;然后,通过计算待部署虚拟机和服务器性能向量的相对距离,得到待部署虚拟机的匹配向量;最后,将匹配向量与系统负载向量综合分析,得到虚拟机部署结果。在CloudSim环境下进行了实验仿真,实验结果证明,使用所提算法可获得较好的系统负载均衡效果和较高的资源利用率。

Regarding the virtual machine deployment issues in cloud computing,the Performance Matching-Load Balancing（PM-LB） algorithm of virtual machine deployment was proposed.With performance vector,the performance standardization of virtual infrastructure was described.The matching vector was obtained by calculating the relative vector distance of virtual machine and the servers,then a comprehensive analysis of matching vector and load balancing vector was done to get the deployment result.The results of simulation in CloudSim environment prove that using the proposed algorithm can obtain better load-balancing performance and higher resource utilization.