提出一种基于QoS的云负载均衡机制,即:构建QoS模型和云资源模型;建立资源度量与QoS属性之间的映射;对虚拟机实例负载状况和虚拟机集群资源利用状况进行量化评估;感知用户的QoS并对比所监控的云节点的资源度量情况,根据对比结果,通过任务调度算法和弹性伸缩算法分别实现任务的分发和虚拟机集群的弹性伸缩,最终达到优化的负载均衡的目的.通过模拟试验,结果表明本方法与Round robin算法相比,有更好的负载均衡效果.
This paper proposes the cloud load balancing mechanism based on QoS, namely: Constructing the QoS model and resource metric model; mapping resource metrics to QoS attributes; modeling the virtual machine instance load status and virtual machine cluster resource utilization quantitative assessment; based on comparison result of the user's QoS and the monitored resource metrics, the task scheduling algorithms and flexible scalable algorithm to achieve the task distribution and the elastic stretching of the virtual machine cluster, ultimately to achieve optimal load balancing purposes. By simulation test results show that this method has better load balance degree compared with the Round robin algorithm.