基于人工鱼群算法的容器云资源低能耗部署方法

Low Energy Consumption Deployment Method for Container Cloud Resources Based on Artificial Fish Swarm Algorithm

目前容器云资源部署过程中能耗较高,直接增加了云服务提供商的经济成本。提出基于人工鱼群算法的容器云资源低能耗部署方法。首先,对容器平台的能量消耗模型展开详细分析,确定容器相关运行参数;通过制定的模型约束条件建立容器云资源的低能耗部署模型;然后,使用人工鱼群算法对模型求解,搜索出部署模型的全局最佳值,制定最佳部署方案;最后,依据制定的资源低能耗部署方案,实现容器云资源的低能耗部署。测试容器云资源低能耗部署时的平均最大等待时间、最大响应时间、最大平均任务队列长度和平均能量损耗。测试结果表明,容器云资源低能耗部署时的平均最大等待时间在检测次数为50次时,仍未超过70 ms;最大响应时间在检测次数为50次时未超过45 ms;随着部署时间增加,最大平均任务队列长度为85 mm,10次实验中平均能量损耗均在25672 kJ以内,由此可见该方法在容器云资源低能耗部署中具有较好的性能。

High energy consumption during the deployment of container cloud resources has increased the economic cost of cloud service providers.A low energy consumption deployment method of container cloud resources based on artificial fish swarm algorithm is proposed.Firstly,the energy consumption model of the container platform is analyzed in detail,and the relevant operation parameters of the container are determined.The low energy consumption deployment model of container cloud resources is established through the established model constraints.Then,the artificial fish swarm algorithm is used to solve the model,search the global optimal value of the deployment model,and formulate the optimal deployment scheme.Finally,the low energy consumption deployment of container cloud resources is realized according to the developed resource low energy consumption deployment scheme.The average maximum waiting time,maximum response time,maximum average task queue length,and average energy consumption in the low energy consumption deployment of the container cloud resources are tested.It is showed that the average maximum waiting time for low energy consumption deployment of container cloud resources does not exceed 70 ms when the number of inspections is 50.The maximum response time does not exceed 45 ms when the number of tests is 50.As the deployment time increases,the maximum average task queue length is 85 mm,and the average energy loss in 10 experiments is within 25672 kJ.Therefore,the proposed method has excellent performance in the low energy consumption deployment of container cloud resources.