Design And Implementation Of Docker-based Continuous Integration Platform

As the software development cycle become shorter, it will make project delayed or failed if problems occur in software integration. It needs a build, which includes compiling code, test and deployment, to verify software integration. Developers often spend a lot of time in the process of software integration. Continuous integration reduces the risk of software integration, and uses automatic build to let developers focus on the development process.Docker technology has the feature of “out of the box”. By combining Docker and continuous integration, it makes each step of build process in containers and makes continuous integration become modularization.Based on the research above, in order to solve the problems that there was high integration risk in SCUT’s energy cloud platform and lack of automated build, this paper raised a continuous integration solution based on Docker technology. It used Jenkins to undertake the task of distributing the schedule work. And it used Docker containers to accept the scheduled tasks and complete the build which included the code compiling, test and deployment.In this paper, the main research work was listed below:First of all, in this paper, the theory of continuous integration and continuous integration development status had been researched. Docker technology and the solution that combined Docker and continuous integration had been researched. According to the research result, the design proposal and goal of the platform had been summarized in this paper.Besides, according to the design proposal of the platform, the modules of build, test and deployment had been completed in this paper. This paper propose a test optimization strategy that can maintain the test result to be correct as well as shorten the test time. Two deployment environment, including the test environment and deployment environment, had been designed. And this paper solved the communication problems between Jenkins system and Docker. A schedule program had been designed to link the code building, testing and deployment together. It completed a monitoring function for real-time data from deployed project that users could have a better understand for the operational status of the deployed projects. Finally, energy transformation cloud platform development and deployment environment of South China University had been changed based on the solution above. This paper solved the problem encountered in the above-mentioned platform and verify the safety and reliability of the system.