Slave Readiness Probes faling about activemq-artemis-helm

[Namphibian]

Slaves wont be ready till there is a failover. This is by design.

[bendahl]

Hi,

I just stumbled upon the same issue and was wondering why the slaves' readiness probes kept failing. As @Namphibian already mentioned, this is by design. In case anyone is wondering why this is, take a look at this:

https://activemq.apache.org/pure-master-slave

and:

https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/

Reading through those documents should clarify a) why the salves will refuse connections and b) why it makes sense to use a readiness probe for that sort of scenario. So all in all a feature, not a bug.

That leaves the question whether this issue should be closed...

Cheers

[vromero]

You are right. This is a very bothersome problem, as it does create problems in regular usage, e.g: helm install ... --wait

I'm still thinking about what to do with this, it might easily happen that the readiness is handled by a script and JMX and the loadbalancer is removed and the clients exptected to configure manually. Not sure yet. I did create an issue to have office hours but still hoping (waiting) for @DanSalt to appear as he has contributed multiple times.