preface

Due to the imperfect infrastructure of the R&D center, the project specifications are not unified, the process specifications are not unified, and the visualization degree is not high. In the case of the increasingly large system, the delivery efficiency, delivery quality and communication cost are greatly affected. In order to better support the rapid development of business, fast implementation. At this time, we proposed to build a continuous integration platform to solve this problem.

  • In order not to affect the normal operation of business, we should not bring new costs while solving problems — we need to be compatible with historical problems, smooth transition, and promote the implementation of engineering standardization
  • Reduce conceptual disunity — we need to make the user experience extremely streamlined, efficient and reduce the cost of understanding
  • Improve promotion progress and reduce communication cost — we need to be value-driven, select part of testing, operation and maintenance, development and PMO to jointly build, optimize the process and improve efficiency
Current problems encountered

Solve the above problems

Due to the above problems, we urgently need to create an automation and visualization platform for the whole R&D and testing process from the beginning of the project to the launch


Implementation scheme

  • Assembly line management assembly line (code scanning, unit testing, compilation and packaging, environment deployment, automated testing) platforming unified management One-click creation of assembly line, automatic integration of each node, promoting code scanning, single testing, automation implementation
  • Environment deployment Collect basic application information and environment configuration information. Manage code packages, image packages, and the environment in a unified manner. All information can be queried and the environment can be smoothly migrated
  • Project collaboration support project test, test report, demand management, iteration, automatic circulation, promote the project process specification automatically fall to the ground Automatically collect test automation, single, standard code scanning measurement, deployment situation, such as information, provide data source for quality operating platform, analysis of development process indexes submit code line trigger automatically, automatic management and demand, Reduce development and testing
Function summary diagram

Overall design scheme

The Pipeline mainly adopts Jenkins2.0 Pipeline as Code Pipeline: simply speaking, it is a set of workflow framework running on Jenkins, which connects tasks that originally run independently on a single or multiple nodes to achieve a complex release process that is difficult to be completed by a single task

Effect display diagram

Current Operating data