Separate consumer to work as worker nodes in Kafka

Currently there is only one worker node which is handling git, gist, pasties etc. If the github has massive data then it takes hours to analyze for secrets.

The performance needs to be optimized.