云计算环境下软硬件节能和负载均衡策略

发布时间：2018-09-11 12:22

【摘要】：针对云计算服务环境下软硬件节能和负载均衡优化问题,提出一种自适应的云计算环境下虚拟机(VM)动态迁移软节能策略。该策略采用常用的硬件能耗感知技术———动态电压频率调节(DVFS)来实现分段优化的系统部件静态节能,又通过VM在线迁移技术实现云平台的动态自适应软件节能。在CloudSim云仿真平台下对比实现DVFS静态节能和自适应负载均衡的软节能策略,经PlanetLab云平台监测数据验证,结果表明:软硬结合的自适应能耗感知策略能够高效节能96%;DVFS+MAD_MMT节能策略(采用平均绝对偏差算法判定主机是否超载,基于最短迁移时间(MMT)原则选择VM移出)在PlanetLab低负载云任务吞吐环境下比实验环境下能节约87.15%的能耗;安全阈值为2.5的MAD_MMT算法可以实现云环境下的高效软节能,有效支持自适应负载均衡的虚拟机动态迁移策略。
[Abstract]:Aiming at the problem of software and hardware energy saving and load balancing optimization in cloud computing environment, an adaptive dynamic migration strategy of virtual machine (VM) in cloud computing environment is proposed. The dynamic self-adaptive software energy-saving of cloud platform is realized by VM on-line migration technology. The soft energy-saving strategies of DVFS static energy-saving and adaptive load balancing are compared under CloudSim cloud simulation platform. The results of monitoring data on PlanetLab cloud platform show that the soft-hard self-adaptive energy-saving strategy can save 96% of energy efficiently. Energy-saving strategy (using mean absolute deviation algorithm to determine whether the host is overloaded or not, based on the principle of minimum migration time (MMT) to choose VM removal) can save 87.15% of energy consumption in the PlanetLab low-load cloud task throughput environment than in the experimental environment; the security threshold of 2.5 MAD_MMT algorithm can achieve efficient soft energy saving in the cloud environment, effectively supporting self-saving. Virtual machine dynamic migration strategy suitable for load balancing.
【作者单位】：新疆大学软件学院;69031部队;
【基金】：国家自然科学基金资助项目(61363083,61262088,61063042) 新疆高校科研计划项目(XJEDU2012I10) 新疆自然科学基金资助项目(2011211A011,2013211A011) 新疆维吾尔自治区青年科技创新人才培养工程资助项目
【分类号】：TP393.09

【参考文献】