一种基于资源使用阈值边界的虚拟机分配策略

虚拟机分配策略是提高云数据中心的物理主机利用率和降低能量消耗的关键技术。提出一种基于资源使用阈值边界的虚拟机(resource utilization thresholds virtual machine,RUT-VM)分配算法，建立了资源使用效率阈值边界和性能参数的数学模型； RUT-VM算法可以根据云端当前工作负载的状态来获得资源的使用效率情况，在工作负载高于或者低于设置的阈值上界与阈值下界时，将物理主机上的虚拟机进行迁移，放置到新的物理主机上。利用CloudSim3.0作为云数据中心的云端RUT-VM分配策略的测试环境。测试结果表明，RUT-VM分配算法可以减少云数据中心的服务等级协议(service level agreement,SLA)违规，保证云计算的服务质量， 与其它的虚拟机分配算法比较起来，可以明显的降低能量消耗。

A resource utilization threshold based virtual machine allocation strategy

Virtual machine allocation strategy is one of the most important techniques for physical resource usage and power cost reduction in cloud data center. A resource utilization thresholds based virtual machine allocation strategy called RUT-VM was proposed and described in this paper. Several common virtual machine allocation strategies were analyzed subsequently. The mathematic model of thresholds boundary and some performance factors were also constructed in RUT-VM. The host and virtual machine utilization status were obtained according to the current workload in RUT-VM and when the workload is higher than or lower than the set upper and lower thresholds, the virtual machine on the physical host is migrated and placed on the new physical host. RUT-VM was simulated in CloudSim3.0 tools by constructing a cloud data center. The evaluation results show that RUT-VM can reduce SLA violations and keep a good Quality of Service for data center. Fewer VM migrations occur compared to the existing techniques, the minimum energy consumption was also obtained in our test.