Payara Micro: Here’s How to Boost Speed and Efficiency

Payara

Payara Micro is designed from the ground-up to be nimble for lightweight, fast deployments. As such, it is ideal for microservices and cloud environments. However, there may be instances where you may want to squeeze every bit of performance out of it. In such cases, the minimalist configuration of Payara Micro comes in handy.

Thanks to this, you can cut Payara Micro’s footprint and speed up startup times, improving the efficiency of your app even more. In effect, depending on your application, you may be able to slash the startup time by up to a third! Here’s how.

Why Go Minimalist?

If you’re running Payara Micro in a container or a resource-constrained environment, every megabyte and millisecond counts. Minimalist configurations help you optimize resource usage and get your apps up and running quickly in such tight environments. 

Disabling Clustering and Hazelcast

Two common areas to streamline are clustering and Hazelcast. Clustering is great for load balancing and high availability, but it adds overhead. If you don’t need these features, you can safely disable clustering to reduce memory usage and improve startup time. Hazelcast provides a distributed data grid for caching and session replication. If these features are not useful within your application, you can disable Hazelcast too.

How to Do It

Here’s the command-line option to disable both clustering and Hazelcast:

java -jar payara-micro.jar --noHazelcast

You can also pass these options in a domain.xml file or through the MicroProfile Config API.

As an example, the start time for Payara Micro with Hazelcast is about 4.4 seconds the first time.

Payara Micro 6.2024.5 #badassmicrofish (build 18) ready in 4,432 (ms)

Starting the same instance without Hazelcast using the above command, the startup time drops to about 1.5 seconds.

Payara Micro 6.2024.5 #badassmicrofish (build 18) ready in 1,578 (ms)

Other Optimizations

JVM Options: Experiment with JVM options, such as garbage collection settings and class data sharing, to optimize memory usage.

Wrap-up

Minimalist configurations are a great way to squeeze the most out of Payara Micro, especially in resource-constrained environments. By carefully choosing which features to enable, you can further optimize startup speed and efficiency without sacrificing the core functionality of your application. Download Payara Micro now to get started! Happy Coding!

 

 

Comments (0)

Post a comment

Your email address will not be published. Required fields are marked *

Payara needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, please review our Legal & Privacy Policy.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Related Posts

Webinar banner for “High-Frequency Trading on Jakarta EE: GC Stress Testing with Azul C4 and Payara Micro,” March 25, 2026, 2 PM GMT. Features Azul and Payara Micro logos and speaker photos of Luqman Saeed, Jakarta EE Specialist, and Simon Ritter, Deputy CTO and Java Champion. 1 minute
Cloud & Microservices

High-Frequency Trading on Jakarta EE: Join Our Upcoming Live Webinar

Modern high-frequency trading (HFT) platforms operate under extreme performance constraints, processing tens of thousands of messages per second while […]

Illustration promoting the Payara Platform Community Survey, featuring bold text on a blue background alongside a clipboard with a checklist, star ratings, and check marks, with coral and fish graphics in an underwater theme. 1 minute
Community

Help Shape the Future of Payara Platform Community – Take Our 2026 Survey

Earlier this week, we’ve launched the 2026 Payara Platform Community Survey and we’d love to hear from you. If […]

Cut Jakarta EE Startup Times from Seconds to Milliseconds with CRaC 8 minutes
Jakarta EE

Cut Jakarta EE Startup Times from Seconds to Milliseconds with CRaC 

Jakarta EE applications can take anywhere from several seconds to over a minute to start, depending on their size […]