Implementasi Pipeline CI/CD dengan Github Actions dalam Mengurangi Downtime pada Aplikasi Berbasis Web

Abstract

This research focuses on optimizing the application deployment process to minimize prolonged downtime, which is often caused by the architecture that don’t supports automation of program code to the deployment server. To overcome this problem, Continuous Integration/Continuous Deployment (CI/CD) pipeline using GitHub Workflow is implemented by involving three main components: local development server, GitHub (repository and actions), and deployment server. The pipeline design includes three types of branches in the GitHub repository: dev, staging, and master. The workflow process starts from local development to the dev branch, then integrated to staging for integration testing, and finally to master which triggers automatic deployment actions to the server using the FTP protocol. From the results of the system testing, the metrics used are downtime and deployment time. The results show that the implementation of the pipeline successfully achieved zero downtime (0 seconds) both in the initial deployment and the deployment of changes. However, the deployment time is still relatively long due to the use of the FTP protocol. Although the application does not experience downtime, the long deployment time has the potential to make the application not optimally usable during the deployment process.