To remain internally and externally compliant, companies must ensure that they aren’t ingesting any PII (personally identifiable information) into their own data warehouses.
In this article, we will discuss the available options for blocking, omitting, and deselecting this data.
Considerations & Recommendations
There are dependencies on the types of connectors you use in Fivetran that determine whether you have the ability to block PII data or not.
For databases and some other connector types listed below, you can deselect the PII data from the relevant schemas:
For application connectors, there are nuances to every connector type where users cannot completely guarantee that PII data can be deselected (see the screenshots below).
Suitable for standard Fivetran customers or those with a small number of connectors and schemas to modify.
In your Fivetran dashboard, navigate to the “Schema” tab of your connector and simply uncheck any schemas/tables/columns that include PII.
Greyed-out checked tables cannot be disabled.
Suitable for all Fivetran customers who have access to the REST API and a large number of connectors to modify.
Fivetran’s REST API allows users to control this process using the Modify a Connector Schema Config endpoint, where you can enable or disable the appropriate schemas, tables, and sometimes columns.
If you intend to use the Fivetran REST API, you can see that field `“allowed”: false` in the Retrieve a Connect Schema Config response which means a table/column is a system table/column you cannot block from being ingested.
For more information on Fivetran’s REST API, please see our Getting Started section.
For many connectors, column-level data blocking is not yet supported, which is why you are not able to click a table to expand it and see the columns within it.
On certain connector schemas (see the screenshots above), you’ll see some tables with grayed-out selected checkboxes, which means these are disabled. These are system tables. Users do not have the ability to deselect them today. System tables are brought over by default to ensure the consistency of our relationships (outlined in their respective ERDs).
If data has already landed in your destination warehouse, please read our support article for removing such data: How do I delete Personally Identifiable Information(PII) data for GDPR Compliance.