Hey folks, I brought an issue up in office hours yesterday, and I’ve done a little more digging, and created a test project to easily reproduce the issue. The issue is that when using logical replication against postgres, if you add any new tables to an existing jobs, their state isn’t correctly saved and it’lll do an initial sync (full extract) against the new tables every time the job runs. This is my `meltano.yml`:

version: 1
plugins:
  extractors:
  - name: tap-postgres
    variant: transferwise
    pip_url: pipelinewise-tap-postgres==1.8.0
    config:
      filter_schemas: public
      default_replication_method: LOG_BASED
      logical_poll_total_seconds: 60
      dbname: test
    select_filter:
    - public-a
    # - public-b

  loaders:
  - name: target-snowflake
    variant: transferwise
    pip_url: pipelinewise-target-snowflake==1.13.1
    config:
      dbname: MELTANO
      warehouse: MELTANO
      user: MELTANO_ROBOT
      file_format: FILE_FORMATS.CSV
      role: MELTANO