Sistem Evaluasi Layanan Lengkap Individu dan Catatan Aktivitas (SELLICA) was initially built using a monolithic architecture based on Next.js API routes. However, as the user scale increased, the system experienced performance and efficiency bottlenecks under heavy workloads. This study aims to conduct a comprehensive comparative analysis between an integrated Next.js backend architecture and a separated, service-oriented Go (Golang) backend for government-scale applications. The research method employs quantitative evaluation through performance testing (load testing with K6 and benchmarks) as well as qualitative evaluation of architectural complexity and development experience. The results indicate that migrating to a Go backend provides a response time improvement of 20 to 289 times faster, 20.25 times higher throughput, and memory usage efficiency up to 4 to 5 times lower compared to Next.js. Conversely, Next.js maintains advantages in rapid initial development and prototyping ease. In conclusion, for large-scale systems with stringent performance and concurrency requirements, a Go backend is highly recommended. A hybrid architectural approach, combining a static Next.js frontend with a Go backend, proves to be the most optimal solution in balancing high performance and cost efficiency.
Copyrights © 2026