When AWS DMS Replication Task runs in full load mode and lets say oracle is the source, does it go and read all the database objects from source (like including views, materialized views, tables etc.) or only the tables? If yes to this question, then is there a way to limit the extraction to tables only?
If there is long list of tables to be excluded form the extraction, is there an easy way to filter those tables out from source, rather than coming up with selection rules for filter them out?