计算机与数字工程2023，Vol.51Issue(10)：2219-2223,2241,6.DOI:10.3969/j.issn.1672-9722.2023.10.001
面向Kubernetes的容器混合伸缩方法
Kubernetes-oriented Container Hybrid Scaling Method
摘要
Abstract
The container-based microservice system in Kubernetes is a typical cloud native application.Most of the existing dynamic container scheduling algorithms only consider the adjustment of container quantity(horizontal scaling),but do not consid-er the adjustment of container configuration(vertical scaling),resulting in unreasonable allocation of cluster resources.Therefore,this paper proposes a Kubernetes-oriented container hybrid scaling method,which uses queuing model to calculate the processing capacity required by the application,and maps the processing capacity to a specific container hybrid scaling configuration scheme based on the container performance model based on deep neural network.Experimental results based on Kubernetes cluster show that this method has the lowest resource consumption cost compared with other methods under the premise of guaranteeing service level agreement(SLA).关键词
Kubernetes/容器/资源调度/垂直扩展Key words
Kubernetes/container/resource scheduling/vertical scaling分类
数理科学引用本文复制引用
柳雪妍,蔡志成,徐建..面向Kubernetes的容器混合伸缩方法[J].计算机与数字工程,2023,51(10):2219-2223,2241,6.基金项目
国家自然科学基金资助项目(编号:61972202,61872186,61973161,61991404)资助. （编号:61972202,61872186,61973161,61991404）
