Azul Optimizer Hub

A free, optional component of Azul Prime, Optimizer Hub offloads compilation and warm-up work to external services so JVMs start faster and hit peak performance sooner.

Download the Data Sheet

Optimizer Hub runs as a shared service alongside your JVM fleet, not inside it. Two components split the work: Cloud Native Compiler and Ready Now Orchestrator. Cloud Native Compiler takes over JIT compilation, so JVMs can run in smaller, lower-cost instances. ReadyNow Orchestrator manages and distributes ReadyNow profiles across the fleet, simplifying deployment at scale. Together, they let every JVM start faster, warm up faster, and perform better.

Cloud Native Compiler

Cloud Native Compiler shifts JIT compilation from individual JVMs to a centralized, scalable service, slashing CPU workloads and caching compilation for reuse. The resulting code runs faster, handles more traffic, and reaches full speed more quickly and smoothly. By eliminating CPU overhead, each JVM can be deployed in smaller instances and containers, reducing cloud costs. Compilation caching speeds up Java fleet warm-up times by 2x – 5x compared with standard OpenJDK installations.  

Learn more

ReadyNow Orchestrator

ReadyNow, a feature of Azul Prime, reduces Java warm-up time by logging and reusing optimization profiles between JVM runs. ReadyNow Orchestrator delivers intelligent curation so each application uses the best profile to ensure consistent responsiveness, fewer errors, and higher SLA attainment during critical times like a retail rush, gaming traffic spikes, or market open.

Learn more

JVMs Accept Traffic and Warm Up Faster

With optimization work offloaded to Optimizer Hub, JVM resources go entirely toward executing your code. There’s no more splitting cycles between compilation and execution, and new JVMs warm up faster when they join the fleet. Benefits:

Better customer experience when autoscaling to meet spikes in traffic

Smoother deployment of new JVMs, for autoscaling or planned fleet restarts

Auto scale with confidence and stop over-provisioning statically scaled fleets

Deliver real savings on your cloud costs

Improve confidence in CI/CD continuous deployment of small fixes – fleet restarts are smoother and less complex

Application Resource Usage on OpenJDK vs Prime with Optimizer Hub

Before

Warms up code as traffic starts
Splits CPU: execution vs. compilation

Result

Slower code
Stalls, errors, and missed SLAs
Increased CPU overhead

After

Optimization done before traffic starts
Moves compilation to shared service

Result

Faster time to full speed and faster code
Fewer errors and missed SLAs
JVMs run in smaller containers

With Optimizer Hub, one of Azul’s largest Prime customers achieved a history first with 10,000+ JVMs collaborating and sharing performance optimizations

Challenge: Inability to dynamically scale with confidence due to errors, timeouts during warm-up, and over-provisioned statically scaled fleet driving up cloud costs.

Solution: Optimizer Hub cut start-up and warm-up time dramatically, making scaling smoother and more predictable.

Benefit: Significantly lower compute requirements and 20%+ cloud cost savings.

Request a Trial of Prime

Azul Prime Stream Builds are free for development and evaluation.

Request a Trial