mqtt-pulsar-forwarder currently separates client ID of the replicas by randomizing the suffix within the instance, similar to e.g. pulsar-forwarder-jRnvjs4. Instead, use a static client ID, e.g. pulsar-forwarder-1-<SECRET> for MQTT protocol v3.1.1 (<SECRET> due to protocol DoS bug) or maybe just pulsar-forwarder-1 for MQTT protocol v5 (need to check).
But how to succinctly configure Kubernetes to run otherwise identical replicas in separate AZs using different secrets or configmaps?
mqtt-pulsar-forwarder currently separates client ID of the replicas by randomizing the suffix within the instance, similar to e.g.
pulsar-forwarder-jRnvjs4
. Instead, use a static client ID, e.g.pulsar-forwarder-1-<SECRET>
for MQTT protocol v3.1.1 (<SECRET>
due to protocol DoS bug) or maybe justpulsar-forwarder-1
for MQTT protocol v5 (need to check).But how to succinctly configure Kubernetes to run otherwise identical replicas in separate AZs using different secrets or configmaps?