【摘要】 云计算拥有大量互连的服务器,多种物理和虚拟共享资源是动态配置和实时变化的,因此,及时准确地监测和获取资源信息,并减少监测开销是云计算管理的一个重要内容.本文以网络监测流量最小为目标,以监测响应时间和负载均衡为约束条件,建立动态划分监测区域的数学模型,并提出一种基于遗传算法的层次化云资源监测方法对该数学模型进行优化求解.该方法首先利用遗传算法确定各个区域监测中心的位置;然后,根据延迟最优策略动态划分监测区域.最后,构建了基于移动代理的层次化监测软件架构以适应动态的监测区域划分.仿真结果表明,所提出的资源监测方法能有效减少监测流量、缩短监测时间、保持区域之间的负载均衡,适用于动态变化的云计算环境.更多还原

【Abstract】 Cloud computing runs in use of interconnected computers,and various physical and virtual shared resources are configured with dynamical change in real time.Therefore,it is an important issue to monitor and obtain accurate resource information in time while reducing the monitoring traffic in the cloud management.In order to minimize the monitoring traffic and to ensure the monitoring response time and load balancing of monitoring centers,the mathematical model of dynamic region partition is established in this paper,and then a hierarchical cloud resource monitoring method based on genetic algorithm is presented to get the optimal solution of the mathematical model.In the proposed method,the genetic algorithm is introduced to select the region management center nodes,and then,the monitoring regions are divided according to the minimum delay strategy.Furthermore,a hierarchical monitoring software model based on mobile agents is established to adapt the dynamic monitoring region partitioning.Simulation results show that the resource monitoring method in this paper can reduce the monitoring traffic,shorten monitoring time and balance the load of the region centers.It is well suitable for dynamic cloud environment.更多还原