Current Restriction in Microsoft Purview Unity Catalog Scanning

As of now, Microsoft Purview only supports scoped scans at the catalog level when working with Azure Databricks Unity Catalog. This means:

• You cannot directly filter scans by schema or table within Unity Catalog.

• The scan setup UI does not offer schema-level or table-level filtering.

• Custom scan rule sets do not support table filters for Unity Catalog scans.

Workarounds and Recommendations

While schema-level filtering is not natively supported, here are some practical workarounds:

1. Split Catalogs Strategically

• Consider splitting your Unity Catalog into smaller, purpose-specific catalogs. This allows you to scope scans more narrowly and reduce scan volume.

2. Use Managed Access Controls

• Apply access controls within Unity Catalog to restrict visibility and access to only relevant schemas and tables. This doesn’t limit scanning but helps manage exposure.

3. Automate Filtering via Scripts

• Use automation or scripting to post-process scan results and isolate the tables of interest. This can help simulate filtering even though the scan itself is broad.

4. Leverage Lineage Tracking

• Utilize lineage tracking to monitor how data flows across tables and schemas. This helps in governance and identifying critical data paths.

5. Use Hive Metastore for Schema-Level Scans

• If schema-level granularity is essential, consider using the Hive Metastore connector instead of Unity Catalog. Hive supports schema-level filtering, unlike Unity Catalog.