Pipeline has lesser fields than Filebeat directly to Elasticsearch

Hi team,

I have a pipeline like this: filebeat =>kafka =>logstask=>elasticsearch (ELK7.12)

Why I have a smaller amounts of fileds than filebeat directly =>elasticsearch (ELK7.11.1)

Whenever I check the fields from ECS seems that even the default dashboards from Filebeat -Zeek are not finding correctly all the fields from ECS.

The new version 7.12 ships with smaller fields synchronisation for modules?

Thank you!