摘要
容器云平台部署过程中受服务器分配的影响,造成部署方案能耗较高,因此,引入粒子群算法实现容器云平台低能耗部署。根据松弛时间概念以及网络重要等级,将最低能量消耗和最短任务执行时间作为目标函数,构建容器云平台低能耗部署模型。以构建模型为基础,设置粒子参数为服务器和频率分配状态,获得单个粒子的局部最优解,将多个粒子优化组合成最优解集,然后在不同部署方案对应的最优解集合中搜索最优解,获取最佳部署方案。仿真测试结果表明,粒子群算法可以有效提升资源综合利用率,降低集群资源失衡度,能获取更加满意的容器云平台低能耗部署方案。
The deployment of container cloud resources is affected by server allocation,resulting in high energy consumption of the deployment scheme.Therefore,particle swarm optimization algorithm is introduced to realize low-energy deployment of con⁃tainer cloud platform.According to the concept of relaxation time and the importance level of the network,the lowest energy con⁃sumption and the shortest task execution time are taken as the objective function to build a low-energy deployment model of contain⁃er cloud platform.Based on the construction model,the particle parameters is set as the server and frequency allocation state,the lo⁃cal optimal solution of a single particle is obtained,multiple particle optimization is combined into the optimal solution set,and then the optimal solution in the optimal solution set corresponding is searched to different deployment schemes to obtain the optimal deployment scheme.Simulation test results show that the method can effectively improve the comprehensive utilization of resources,reduce the imbalance of cluster resources,and obtain a more satisfactory low-energy deployment scheme of container cloud plat⁃form.
作者
徐胜超
陈刚
毛明扬
XU Shengchao;CHEN Gang;MAO Mingyang(School of Data Science,Guangzhou Huashang College,Guangzhou 511300)
出处
《计算机与数字工程》
2023年第1期199-205,共7页
Computer & Digital Engineering
基金
国家自然科学基金面上项目(编号:61772221)
广州华商学院校级导师制科研项目(编号:2022HSDS16)
广东省哲学社会科学规划项目(编号:GD17XGL19)资助
关键词
粒子群算法
容器
云资源
低能耗部署
松弛时间
particle swarm algorithm
container
cloud resources
low-energy deployment
relaxation time
