Design And Implementation Of Container Cluster Management System Based On Docker

In recent years,container technology has become more and more widely used.At the same time,various container cluster management tools have been developed,such as kubernetes,Swarm et al.,however,through the research and analysis of the container cluster management tool,found that there are still many deficiencies,including:1.In the existing resource scheduling process,the layered principle of image is not fully utilized,and the resource reuse is not effective,resulting in the waste of resources;2.The resource scheduling is a static scheduling,and the scheduling priority is determined according to the existing resource requests,Without considering the actual resource usage of the node,and the different consumption of different resources by the application,it is easy to cause the unbalanced use of node resources,thus forming a performance bottleneck of a resource;3.At present,the expansion index of the automatic extender is relatively single,and a single expansion index can not meet the needs of the application in actual use,in order to ensure the application to provide high service quality and Stability,need to provide a self-defined expansion index of the auto extender.Aiming at the above problems and combining the functions of the container cluster management system,this paper focuses on the efficient use of IT resources and the demand of providing high reliable services for users,designs and implements the container cluster management system based on Docker,and makes a comprehensive description of it.Based on the problems existing in the current container cluster management tools,combined with Kbernetes container cluster management framework,aiming at the insufficient reuse of image resources,using the principle of image layering,the resource scheduling strategy based on basic image is proposed.To solve the problem of resource imbalance caused by uneven consumption of resources,a dynamic resource scheduling method is added.According to the actual use of resources and the performance of nodes,the priority of scheduling is calculated to reduce the consumption of different resources and reduce the performance bottleneck.Aiming at the single index of application dynamic scaling,the application dynamic scaling function based on user-defined index is designed and implemented.Using the development mode of front-end and back-end separation,using node.js to complete the back-end corresponding modules,Vue to complete the front-end corresponding modules,design and implement the container cluster management system based on Docker,to solve the problems of low resource reuse rate,unbalanced resource load and single dynamic scaling index in the current container cluster management.And according to the function of the system,design and implement each function module of the system.The specific work of the entire paper is as follows:1.Describe the background and significance of docker based container cluster management system,and combine the previous research and application status in container cluster management,put forward the problems to be solved and organize the paper structure.2.Introduce the related technologies of the container cluster management system in detail,including docker image layering principle,kubernetes and cluster monitoring tools.The scheduling module and dynamic scaling module of kubernetes are further studied and analyzed.3.Analyze the requirements of the system from the functional and non functional aspects.Finally,design the overall system and database according to the requirements of the system.4.According to the analysis of system requirements,design and implement each major functional module and its sub functional modules,mainly including resource monitoring module,resource scheduling module,operation and maintenance management module,application management module,image management module and template management module.This paper focuses on the resource monitoring module and resource scheduling module,combining with monitoring technology to provide the definition of self-defined scaling index,and in the resource scheduling module,it introduces two scheduling strategies based on basic image and load balance in detail.5.Based on the realization of system functions,test the system from the aspects of function,performance,browser compatibility and security,and record the test results,so as to provide basis for further improvement of the system and ensure the efficient and stable operation of the system.