The scenario is as follows:
I have a Main-Failover broker setup of apache-artemis-2.33.0.redhat-00016. See broker-main.xml and broker-failover.xml for my broker configurations.

I have built a demo Quarkus app to demonstrate the issue I am having.
Run the Quarkus app using ./mvnw quarkus:dev

The demo consists of two REST resources, /test/normal and /test/large-header. The resources will return the response time of the request.

A `curl localhost/test/normal` call will work fine after the module has started
A `curl localhost/test/large-header` call will send a message to a queue with a header that is too large, so it will trigger a JmsException.

After startup, you can try a few calls to /test/normal to see that the module is working.
Next, call /test/large-header once
You will get an Exception as response with a message that the header is too large.

After this, when you try to call /test/normal a few times, you will see that the call will not respond.
Only after killing the failover broker, the call will resolve

Note: The issue does not occur when i only run the main broker.