Skripsi
ANALISIS STATISTIK INFERENSIAL TERHADAP PERBANDINGAN PERFORMA CI/CD DAN DEPLOYMENT KONVENSIONAL PADA APLIKASI BERBASIS WEB SERVER
Continuous Integration and Continuous Deployment (CI/CD) is a widely used approach for automating software build and deployment processes in server-based applications. However, the implementation of CI/CD may increase server activity and affect resource utilization. This study aims to analyze and compare the performance of CI/CD and conventional deployment methods in a web-based server application. The experiments were conducted using two code change scenarios, namely minor changes and major changes, with build time, deploy time, CPU usage, and memory usage as evaluation parameters. Each combination of deployment method and scenario was tested for 300 iterations, resulting in a total of 1,200 deployment iterations. The collected data were analyzed using descriptive statistics, the Shapiro-Wilk normality test, and the Mann-Whitney U Test. The results show that in the minor change scenario, CI/CD achieved lower build time and CPU usage, while the conventional method achieved lower deploy time and memory usage. In the major change scenario, the conventional method outperformed CI/CD in build time, deploy time, CPU usage, and memory usage. The Mann-Whitney U Test indicated statistically significant differences for all parameters (p-value < 0.05). The findings suggest that CI/CD is not always faster or more resource-efficient than the conventional deployment method; however, it provides advantages in deployment automation, standardization, and process consistency.
No other version available