Clean Up Source Tags

Alation Cloud Service Applies to Alation Cloud Service instances of Alation

Customer Managed Applies to customer-managed instances of Alation

Applies from Alation version 2025.1.3

You can clean up unused or outdated tags using the connector capabilities.

While the unified source tag model standardizes tag behavior across data sources, Data Source Admins can still control tags across source systems using schema deselection in metadata extraction (MDE) and the Clean Up Source Tags toggle.

Use the Clean Up Source Tags toggles to control the clean-up of tags. For Snowflake, tag deletion is driven by schema deselection. For Unity Catalog, tag deletion is driven by source-side tag deletion or asset deselection during metadata extraction. Find more details in Snowflake and Databricks Unity Catalog.

Important

Alation doesn’t currently provide a UI-based preview of where a tag is applied across catalog pages prior to deletion using the Clean Up Source Tags toggle. Always exercise caution when enabling tag clean-up, particularly in environments where tags are used across data sources or schemas. We recommend that you manually review where tags are applied before deselecting schemas or activating the tag deletion toggle.

Consider the following example to understand tag deletion:

A user extracts metadata from Schema A in Snowflake. The Schema A defines a tag INDUSTRY with values OIL and TECH. The user applies the INDUSTRY tag with value OIL to:

  • A table in Schema A (tag applied in Snowflake via SQL)

  • A table in Schema B (tag applied in Snowflake via SQL)

  • A table in Databricks Unity Catalog (tag applied in Databricks Unity Catalog)

After applying the tags, this user performs these actions in the connector settings:

  • Removes Schema A from extraction in the metadata extraction configuration.

  • Enables the Clean Up Source Tags toggle

Note

Deletion applies only to tags extracted through the same data source connection identified by the JDBC URL from which the schema was deselected. When you remove a schema from extraction while the Clean Up Source Tags toggle is enabled, Alation performs targeted deletion. It removes only the tag definitions (corresponding custom Picker fields) and tag links (association between tag values and objects to which they are applied) that were specifically extracted through the connector associated with that schema. Let’s assume a tag with the same name exists in a schema that is part of a different connector, such as another Snowflake connection pointing to a distinct account or region. In such a case, that tag and its associated tag links won’t be affected.

Since the tag INDUSTRY originated from Schema A, and that schema is currently not selected for extraction:

  • The tag definition (custom Picker field) INDUSTRY is deleted from Alation

  • All assignments (links) of INDUSTRY with value OIL are deleted from the catalog pages of relevant objects, including those on Schema B in Snowflake and the objects in Databricks because the source tag no longer exists in Alation.

Note

This deletion behavior applies to any object where a source tag value was applied. These links are removed if the tag definition is deleted.