My Team originally setup a Connector for Workday HCM, but at the time we were choosing objects that clearly had no PII related data.  As our project has grown we have recently had the need to add other objects that DO have PII data and we wanted to pick/choose columns, but after the initial SYNC this is not possible until we SYNC those objects at least once - which violates are requirement to NOT load PII data.

We have worked with Fivetran Support on this for over a week validating this.

In Summary: When a connector is already configured but has not yet loaded all the objects, users are unable to access the column list in the Schema tab.
 
Feature request— enable loading of individual tables into Fivetran Caching (even for already established connectors). Basically, a version of "I need to customize data before syncing option" for existing Connectors.

This would allow users to inspect and select columns for hashing or blocking before the data reaches the destination.

At this time, we are forced as a work-aroud to Restrict all access to the Destination Schema in Snowflake, SYNC the object(s) in question, Pause the Connector and drop the objects from Snowflake, Un-Pause the Connector and Un-Restrict access in Snowflake again.

A workaround that will work, but technically violates our NO PII Load policy for that particular Connector.