基于花授粉算法的容器云资源低能耗部署方法

Low-energy Deployment Method of Container Cloud ResourcesBased on Flower Pollination Algorithm

论文提出了基于花授粉算法的容器云资源低能耗部署方法。首先构建容器云资源能耗模型,在此基础上迁移容器,提出最低增长法、最高增长法、混合增长法以及随机选择模式四种容器选择方式应用在迁移中。综合容器迁移以及能耗模型得出资源部署的最低能耗,利用花授粉算法实施容器云资源的部署,划分云资源任务个体适应度种群,得出每个种群最高适应度实施初始化和更新处理,输出花授粉算法的最优解,即资源的最优部署策略。实验结果表明,该方法的平台资源利用率高、部署完成时间短以及负载均衡好,降低了容器云资源部署过程中的能耗。

A low-energy deployment method of container cloud resources based on flower pollination algorithm is proposed.First,the container cloud and build the container cloud resource energy consumption model.On this basis,the container is migrat⁃ed,and four container selection methods are put forward,which are minimum growth method,maximum growth method,mixed growth method and random selection mode.The minimum energy consumption of the original resource deployment is obtained by in⁃tegrating the container migration and energy consumption model.The flower pollination algorithm is used to deploy the container cloud resources.Niche technology is added in the calculation process to divide the individual fitness population of cloud resource tasks,obtain the highest fitness of each population,implement initialization and update processing,and output the optimal solution of the flower pollination algorithm,that is,the optimal deployment strategy of resources.The experimental results show that the pro⁃posed method has high platform resource utilization,short deployment completion time and good load balancing,and reduces the en⁃ergy consumption in the process of container cloud resource deployment.