Performance improvement using reactive approach

[Arooba-git]

Hi 🙂,

Thank you for developing and maintaining this website, and for providing such clear instructions. We successfully executed the project using your guidelines.

With the rise in popularity of Java app development—especially with reactive programming and virtual threads—we are researching open-source projects to compare performance across three web development paradigms: imperative, reactive, and virtual threads. We assessed performance based on CPU and memory usage, as well as request concurrency.

Your clear guidelines allowed us to test your project in all three scenarios. Our findings indicate that reactive programming significantly enhanced performance, as shown in the table below:

Metric	Original	Reactive
Throughput (requests/sec)	19.7	83.2
Memory Usage (MB)	608.7	512.2
Latency (longest processing time)	35.8s	8.2s
CPU Usage (%)	5.8	23.8
Requests Passed (%)	99.6	99.9

As a proof of concept, we have created a simplified version of your project using webflux, featuring a single endpoint for demonstration. You can view and test this demo project here to compare its performance with the original.

Profiler used: JMeter

Best regards.