Currently if new tables are created during CDC the existing inclusion/exclusion filters that were set when the task was created are not honored. This means that we will begin tracking changes on a newly created table during CDC that we do not want to track. This leads to major issues for our company in particular because we are explicitly filtering out tables with columns that are not supported in other tasks (Kafka).
Example.
- Task created and full load completed with exclusion filter: 'AUDIT[_]%'
- During full load no tables that begin with 'AUDIT' are loaded
- A new table is created for a new application feature with name 'AUDIT_MYNEWFEATURE'
- Qlik task begins CDC for that new table even though exclusion filter is setup
- Other task begin failing because our Audit tables have data types not supported by them