Skip to main content

Row-Level Security (RLS) | Access Management

📘

Implementing RLS in your organization's dashboards requires custom implementation. Contact the Sales Team (comercial@dadosfera.ai) to request it through Professional Services.

What It Is​

Dadosfera supports row-level security to determine which rows are returned in query results. Each user can access only the records permitted for their queries.

Currently, Dadosfera provides RLS only for dashboards and questions created in Metabase, the natively integrated Visualization module. Users see only the data they are authorized to access.

The permission list is created from the filter field.

👍 Use Case

A company sells throughout Brazil and has sales managers in each region. Each manager can see only sales results for their own region. A general manager, however, can see both aggregate and regional results in the same dashboard.

How to Use​

If your organization has implemented RLS, open the asset normally in Dadosfera to see only the records you are authorized to access. The filter defining record access is not available for selection in the interface.

🚧

To query a dashboard correctly configured with RLS, access it only through the Dadosfera catalog. Users with RLS restrictions must not have direct access to the Visualization module. Grant them permission only to access the catalog and the required data assets.

Current Limitations​

The current RLS data access security solution has some limitations:

  • Only assets created in Metabase support RLS.
  • Only one parameter can be provided: for a single field, only one value can be defined for a user's access. In the use case above, a user can access one region or all regions, but not two specific regions, for example.
  • Not all fields support implementation. Each case must be tested to determine whether RLS can be implemented.
  • CLS (Column Level Security) cannot be implemented to control whether users can access records in a specific column. This would be useful for columns containing sensitive data, for example.