Per Kateryna Senchenko, this may not be required once the other stabilization work is done. Leaving it as P4, but changing it from a spike to a task to check and see whether all is fine, or if additional work is needed.
Modules interacting with kafka show excessive usage of CPU even in when there is no data-import job running. It might be related to the fact that each Kafka Consumer is configured in a separate Verticle and we have multiple Consumers for some topics. Need to investigate whether it is beneficial to have multiple Consumers or it is enough to have just one to reflect the partition strategy.