%A NIM.: 22106050039 Fadilla Ma’sum Haqqoni
%O Ir. Muhammad Taufiq Nuruzzaman, S.T. M.Eng., Ph.D.
%T KOMPARASI PERFORMA, BIAYA, DAN SKALABILITAS CLOUD RUN DAN APP ENGINE PADA BACKEND REST API
%X The adoption of serverless Cloud Run and App Engine lacks sufficient head-to-head empirical data in the industry. This study evaluates the performance, cost efficiency, and scalability of both platforms running a RESTful API backend to provide an objective foundation for cloud architecture decisions. Quantitative experiments were conducted on a Go-based "Notes App" using K6 for average load (150 VUs) and stress test (3,000 VUs) scenarios with 10 replications. Results indicate that under moderate loads, both platforms perform equally, but App Engine is more economical (Rp1.45/1,000 requests) due to its daily free quota. Conversely, during stress testing, a cost reversal occurred: Cloud Run outperformed with a p95 latency of 325.59 ms (versus 1,646.15 ms for App Engine) and a 0.0027% error rate. This superiority is driven by Cloud Run's instant instance scalability, whereas App Engine's slow autoscaling triggered request queues leading to CPU utility saturation. Furthermore, Cloud Run proved more cost-efficient under high loads (Rp6.88/1,000 requests) due to its per-second billing scheme eliminating idle instance costs. In conclusion, Cloud Run is optimal for large-scale, fluctuating traffic, while App Engine is ideal for moderate daily loads.
%K Serverless Computing, Google Cloud Run, Google App Engine, Performance Testing, Go, RESTful API, Efisiensi Biaya
%D 2026
%I UIN SUNAN KALIJAGA YOGYAKARTA
%L digilib76449