RedHat AMQ 6.3 for Openshift doesn't restart pod when broker application is hung-up

Issue

• When testing if RedHat AMQ 6.3 detects if the AMQ broker is hung-up, Openshift didn't detect or restart the AMQ broker POD.

We simulated a bug in code that causes a deadlock by pausing the AMQ java process on the Openshift node, i.e.:
sudo kill -STOP

The RedHat AMQ comes with a template providing a readinessProbe (running /opt/amq/bin/readinessProbe.sh) but there is no livenessProbe.

The readinessProbe didn't detect the hung-up broker pod, i.e. traffic was not stopped, neither did it restart the pod.

Is there some recommendation from RedHat how to add a liveness probe?