基于DVFS感知与虚拟机动态合并的云数据中心能效策略

为了解决云数据中心资源分配时能耗与性能间的均衡问题，提出了一种基于DVFS感知与虚拟机动态合并的能效优化策略。首先，策略通过新的DVFS管理算法（DVFS-perf）在不降低系统性能的同时降低了数据中心功耗，然后，通过频率感知的虚拟机VM部署合并算法（Frequency-aware Placement）在实现DVFS最优配置的同时最小化总体能耗，同时确保了虚拟机映射时的QoS保障。最后，通过真实云负载数据流构建仿真实验进行了性能分析。结果表明，在动态负载条件下，策略可以在不降低QoS和不增加SLA违例的情况下，降低虚拟机迁移次数和数据中心的总体能耗，更好地实现能耗与性能的均衡。

Energy efficiency Strategy in Cloud Data Center Based on DVFS-aware and Dynamic Virtual Machines Consolidation

For solving the trade-off problem between energy consumption and performance during resource allocation in cloud data center, an energy-efficient optimizaiton strategy based on combined DVFS-aware and dynamic virtual machine consolidation is designed. First, through a new DVFS-performance algorithm, our strategy can reduce power consumption while preventing performance degradation. Then, a DVFS-aware consolidation algorithm is desinged to optimize consumption, considering the DVFS configuration that would be necessary when mapping virtual machines to maintain QoS. Finally, we perform an extensive evaluation on the simulation toolkit using real cloud traces. Experimental results show that our strategy can obtain less virtual machines migration and provide substantial energy savings of data center for scenarios under dynamic in workload conditions without reducing QoS and increasing SLA violation, which could get better trade-off between energy consumption and performance.