Snowflake connector currently does an automatic switch when it finds the source table in snowflake has

1. Change tracking disabled.

2. Change tracking metadata is deleted.

In these scenarios , we want an ability to fail , connector currently automatically switches from Time Travel to Teleport .. restating the whole target table. There are two issues with this

1. It switches to allow performant option without us knowing.

2. It often fails after that trying to perform a full re-sync using teleport.

It is a vulnerable solution, so we require a feature to force once one type of replication and define our recovery processes based on that.