Hi team, In which use case we will use push based Kafka ?

Hello, Are you talking about push based kafka from this image: ? If yes, push based Kafka is meant to be used for resilience, to handle scenarios where GMS might crash, become unavailable or to handle large loads. Kafka provides a means to support eventual consistency of metadata ingestion and scaling capabilities. You simply right the data kafka and will now that it will be eventually processed even if GMS can’t handle your load immediately. Another use-case is to allow for a scalable endpoint if you create a custom ingestion source that is able to push metadata to DataHub rather than have scheduled crawlers. This enables realtime ingestion of metadata.