Investigate Kafka Configuration merge to Solve Non-existent Leader Issue

[felix-seifert]

Currently, the infrastructure is started with a docker-compose up in the infrastructure folder. This spins up, amongst others, a strimzi Kafka container. After the infrastructure startup, no topic is created automatically and the topics sensors and sensors-sma are only created when respective services access them.

When a sensor is started after the infrastructure is spun up, the sensor logs show a warning that a leader is not available.

WARN  [org.apa.kaf.cli.NetworkClient] (kafka-producer-network-thread | kafka-producer-sensors) [Producer clientId=kafka-producer-sensors] Error while fetching metadata with correlation id 1 : {sensors=LEADER_NOT_AVAILABLE}

As another warning message of the sensor tells the the merge configuration is not known, the issue might be that sensors have to use the configuration merge because there can be multiple producers for the same topic (see application.properties.

WARN  [org.apa.kaf.cli.pro.ProducerConfig] (main) The configuration 'merge' was supplied but isn't a known config.

Try to figure out why these warnings appear and how to remove them.