How fast is LavinMQ? It's time to find out
The LavinMQ benchmark page tracks throughput and latency across releases and AWS instance types. The test setup is open and the numbers are updated with every release.
What the benchmark page shows
Two metrics are measured on the benchmark page:
- Throughput — how many messages per second LavinMQ can handle under load
- Latency — how long it takes for a message to travel from publisher to consumer
Both matter. A broker that moves messages fast but with unpredictable latency is not useful in every workload.
The test setup
The benchmarks run on AWS using open-source tools and custom scripts. Nothing is hidden or proprietary. The hardware varies by instance type so it is easy to compare how LavinMQ scales with different resources.
The full setup is on GitHub. Clone the lavinmq-benchmark repository and run the same scenarios against a local or hosted instance.
A few things worth knowing
No benchmark tells the whole story. Synthetic tests do not reflect every real-world workload, and network and disk performance can skew results. AWS network burst limits also make the largest messages unreliable on smaller instances, so those runs are skipped.
These are real constraints worth keeping in mind when reading the numbers.