The Design And Implementation Of Kubernetes Cluster Management System In Container Cloud Platform

In recent years,with the continuous popularization of microservices,users have more urgent needs for the rapid launch of services.The currently popular virtual machine nodes cannot meet the needs of such users due to some of their own problems,so Kingsoft Cloud decided to containerize some virtual machine nodes and create a new batch of container nodes.The Kubernetes cluster management system is developed to better manage container nodes and ensure the rapid launch and stable operation of user servicesThis thesis first describes the project background and significance of the cluster management system,analyzes the problems in the current public cloud system,and explains the necessity of container cloud development.Then it introduced the development overview and technical characteristics of similar products at home and abroad,and demonstrated the effectiveness of container cloud products for the company's service management.Then it summarizes the Docker and Kubernetes technologies required for system development,and details the characteristics and advantages of each technology.Then from the two aspects of functional requirements and non-functional requirements,the system analyzes the needs of administrator users and ordinary users in the system,and summarizes the three main functional modules of user management,resource management and resource release.On this basis,through the overall architecture design,user grouping and authority design,resource structure design,resource release logic design,etc.,the system is designed to outline,clear the structure and function of each module.Finally,using Go language and Gin highperformance Web framework,through the class diagram,E-R diagram,timing diagram,the above functional modules are designed and implemented in detail.The system has been basically developed,and a large number of test users have good feedback,which not only meets the needs of users to deploy various types of services such as scheduled tasks,load balancing,and daemons,but also solves the problem of slow virtual machine cluster startup and resource waste.It has ensured the full utilization of cluster resources and the experience of public cloud users when deploying services,and has made sufficient preparations for the upcoming 5G era.