Performance Analysis of Google Cloud Platform for Web-Based Applications

Abstrak

Selecting an appropriate cloud computing service remains a major challenge in web-based application development, as it directly affects performance, scalability, security, and operational costs. Google Cloud Platform (GCP) offers multiple computing services, yet empirical comparisons among its core services are still limited. This study aims to evaluate and compare the performance of Compute Engine, App Engine, and Kubernetes Engine in hosting web-based applications. A quantitative experimental approach was employed using a Node.js and PostgreSQL-based e-commerce application, tested under various workload scenarios using Apache JMeter. Performance metrics, including response time, throughput, latency, scalability, reliability, security, and cost, were analyzed. The results indicate that Compute Engine provides stable performance for predictable workloads, App Engine delivers low latency with higher operational costs, and Kubernetes Engine offers the best scalability and resource efficiency. Performance optimization techniques such as caching and CDN integration further improved API responsiveness. This study concludes that Kubernetes Engine is the most suitable choice for large-scale and dynamic web applications. Optimal GCP service selection should align with workload characteristics and organizational requirements.