Unity catalog and lakehouse governance

Unity Catalog And Lakehouse Governance

Overview

Use this skill when Unity Catalog is the core governance surface for a Databricks platform. It helps agents design governed boundaries across catalogs, schemas, storage locations, shares, and access paths for lakehouse data products.

When to Use

• designing Unity Catalog hierarchy and ownership
• defining external locations, credentials, and storage boundaries
• governing shared Delta tables, views, volumes, or models
• reviewing cross-workspace or multi-team lakehouse governance
• planning secure sharing, publish approvals, or governed consumer access

Do not reduce Unity Catalog to permissions only. It is also a publish, ownership, and operating-boundary decision.

Workflow

1. Define the governance hierarchy. Decide:

   • catalog boundaries
   • schema boundaries
   • producer versus consumer separation
   • ownership and stewardship paths

2. Define storage and credential boundaries. Cover:

   • external locations
   • managed versus external tables
   • storage credentials
   • workspace and environment separation

3. Define access and sharing behavior. Include:

   • roles and groups
   • row or column restrictions where needed
   • governed views
   • cross-team or external sharing expectations

4. Align governance with lakehouse delivery. Require:

   • clear publish layers
   • lineage visibility
   • certification or trust signals
   • handling for schema and contract change

5. Validate day-2 operations. Review how promotions, new workspaces, external consumers, and object-store boundaries behave under the governance model.

Common Rationalizations

Rationalization	Reality
"We already use Databricks, so Unity Catalog design is obvious."	Poor catalog and location design creates long-term friction in ownership, sharing, and environment boundaries.
"We can give broad workspace access and tighten later."	Over-broad access becomes hard to unwind once shared data products are adopted.
"Lineage comes for free."	Platform lineage is useful only when publish boundaries and ownership are designed intentionally.

Red Flags

• catalogs and schemas do not map to ownership or publish boundaries
• external locations and credentials are inconsistent or ad hoc
• workspace separation and environment promotion are vague
• sharing or external-consumer behavior is undocumented
• governed lakehouse objects are published without trust or lineage signals

Verification

• Catalog, schema, and ownership boundaries are intentional
• Storage and credential boundaries are explicit
• Access and sharing rules fit the consumer model
• Publish and lineage expectations are visible
• Promotions and day-2 operations work with the governance design