10 Best Practices for Jakarta EE Performance Optimization

With this article, we start a series where we compiled 10 best practices for performance optimizations and suggestions how to implement them using Jakarta EE & Eclipse GlassFish. Enjoy reading this initial overview and watching the video about performance tuning Java applications, and stay tuned for upcoming articles from our OmniFish team that will go into details about each best practice.

To make your Jakarta EE applications faster and more efficient, here are 10 proven best practices you can easily implement:

1. Optimize Database Operations: Use connection pooling, prepared statements, and statement caching to reduce query times and resource usage.
2. Implement Caching: Combine client-side, in-memory, and distributed caching to minimize server load and improve responsiveness.
3. Fine-Tune Runtime Settings: Adjust heap sizes, JVM options, Garbage collection options, thread pool sizes, or network listener settings
4. Leverage Asynchronous Processing: Handle tasks in parallel to boost system throughput and user experience.
5. Manage Components Effectively: Reduce overhead by optimizing resource allocation and disabling unused features.
6. Compress Network Data: Enable compression to reduce data transfer times and improve response speeds.
7. Use Monitoring Tools: Track performance metrics like memory usage, CPU load, and database connections to identify bottlenecks.
8. Test Performance Regularly: Spot and fix issues early with profiling tools like VisualVM or JProfiler, or using performance test tools like Apache JMeter, Gatling, or Bombardier
9. Streamline Code: Replace inefficient code patterns (e.g., use StringBuilder instead of string concatenation for better performance, use sets or maps instead of lists for big collections where appropriate, use asynchronous I/O and streaming data instead of blocking calls and request-response communication for bigger payloads)
10. Adopt New & Standard Java Features: Use features like Java Streams API for streamed and parallel processing and JCache for standardized caching.

These strategies directly impact user experience, scalability, and operational costs, helping you build applications that perform well under pressure.

---

Quick Comparison

Performance Area	Key Practice	Impact
Database Operations	Connection pooling, prepared statements, tuned Jakarta Persistence queries	Faster queries, fewer queries, less resource usage
Caching	In-memory, client-side, distributed caches	Reduced database load, reduce load on system components
Runtime Settings	Adjust heap sizes, JVM configurations, Server tuining	Improved stability and speed
Asynchronous Processing	Parallel task handling	Higher throughput
Component Management	Disable unused features	Lower memory overhead
Network Data Compression	Enable compression	Faster response times
Monitoring Tools	Gather metrics from JVM, OS, Server, Applications	Early bottleneck detection
Performance Testing	Regular testing and profiling with tools	Easier optimization
Streamline Code	Use efficient code patterns	Faster execution, lower memory use
New & Standard Java Features	Streams API, JCache	Enhanced responsiveness

Secrets of Performance Tuning Java on Kubernetes by Bruno Borges

Before we go into details about each best practice in the next articles, I’d like to leave this video for you get an interesting perspective on performance tuning of Java apps from Bruno Borges, an amazing Java developer advocate, Java Champion, and a person I deeply respect and admire.

Next Steps

To put these ideas into action, you should start with Continuously Monitoring your applications. Set up monitoring of your applications and collect metrics about what’s going on. Track CPU usage, memory, garbage collection, speed of processing requests, etc.. This data helps identify bottlenecks early and will prove invaluable to understand what’s happening once you have performance issues and how to address them. They will also show the impact of the optimizations you implement in the future

In the next article, we’ll go into more details about improving database performance, so stay tuned…