There is a mismatch between the distinct network_userids in elasticsearch and redshift. The network_userid count in redshift (batch pipeline) is almost 70-80% higher than the distinct network_userids in elasticsearch (real time pipeline).
Am I missing out on something?