Evaluating the Sustainability and Scalability of Java Virtual Threads and R2DBC in High Concurrency Web Portal Architecture

Java’s Virtual Threads, introduced through Project Loom, represent a significant shift in how Java handles concurrency. This study evaluates the performance of Virtual Threads in a Spring Boot web portal integrated with a reactive PostgreSQL database. A data-intensive student management dashboard was developed to benchmark the transition from the traditional thread-per-request model to a lightweight task-based concurrency model. Experimental results show a 75% reduction in memory overhead and noticeably improved system responsiveness under peak loads of 2,000 concurrent users. Compared to platform threads, Virtual Threads provided better throughput, stable response times, and improved memory efficiency without increasing development complexity. These findings demonstrate that Virtual Threads offer a practical and scalable solution for building modern high-performance web applications.