Continuous integration system: A Component of controlling executable algorithms’ composition for automated program code testing based on system and program load monitoring

Large IT companies, which are develop complicated software, use specific continuous integration systems, that are part of multi-component company infrastructure. Such systems are developed by these companies and differs from public analogs due to the complicated relationship between contained software projects. Because of it, some critical points of these systems development are: increasing stability and performance of automatic build and testing projects, facilitation and speeding up the system troubleshooting and reducing their impact on the continuous integration process. This work is performed for the continuous integration system that uses a monorepo, and "Trunk-based development" practice, which is developed by Yandex LLC. The purpose of the work is the development of the component of controlling executable algorithms’ composition for automated program code testing based on CI system load monitoring for more efficient available resources management. The issues of the work are: 1. The literature review; 2. The existing solutions review; 3. Analysis of ways to distribute the load on the CI system by determining the executable algorithms’ composition for automated testing; 4. Analysis and configuration of monitoring systems to determine the current load on the CI system; 5. Development of the component for deciding on the new executable algorithms’ composition for automated testing based on system monitoring. 6. Testing and integration the developed software product; 7. Development of the technical documentation; Keywords: resource management, continuous integration, monorepo, high load. The paper contains 38 pages, 3 chapters, 11 illustrations, 12 bibliography items, 5 appendices.