云数据中心基于温度感知的虚拟机迁移模型

TA-VMM: TEMPERATURE-AWARE VIRTUAL MACHINE MIGRATION MODEL FOR CLOUD DATA CENTERS

提出云数据中心基于温度感知的虚拟机迁移模型TA-VMM。TA-VMM迁移时着重考虑物理主机处理器的温度情况和物理主机负载均衡情况。在物理主机状态检测阶段寻找出候选迁移主机MigrationFromHosts;在虚拟机选择阶段寻找出候选迁移虚拟机列表VmstoMigrateList;在最后的虚拟机放置阶段完成候选迁移虚拟机的重新放置。CloudSim云计算模拟器仿真结果表明,TA-VMM中温度阈值对云数据中心的性能影响十分重要,TA-VMM比其他虚拟机迁移模型具有更低的能量消耗。

A temperature-aware virtual machine migration model for cloud data centers called TA-VMM is proposed in this paper. The temperature and load balance fators on physical host had been taken into account during the virtual machine migration in TA-VMM. In physical host status detection process, an opitimized algorithm was designed to obtain the candidated migrating host. In virtual machine selection process, an opitimized algorithm was designed to obtain the candidated migrating virtual machines. In final virtual machine placement process, the candidated virtual machines were relocated again. The experimental results on the CloudSim simulator show that the performance of cloud data centers significantly depended on the value of the temperature threshold, the total energy consumption of TA-VMM outperformed other kinds of virtual machine migration models.