Unity Catalog is the governance layer of Databricks. In Azure Databricks it is described as the unified governance layer for data and AI. Once it is enabled for a workspace, it works underneath all data and AI interactions there. It provides centralized access control, auditing, lineage and discovery across workspaces, so that rules about who may see or change data are defined in one place instead of being repeated in each workspace.
The documentation describes how this works in practice. Unity Catalog controls access whenever a table is queried or a model is called. It tracks lineage automatically and records activity for auditing. Access to the catalog is available through Catalog Explorer, SQL, the Databricks CLI and REST APIs. Everything that is governed is a securable object, and privileges on these objects are granted to users, service principals or groups. Tables, views, volumes, functions and models are organized in a three-level namespace made up of a catalog, a schema and the object itself.
A distinction that matters for storage decisions is the one between managed and external objects. Tables and volumes can be managed, in which case Unity Catalog handles the storage lifecycle as well as governance. They can also be external, in which case Unity Catalog handles governance only and the storage stays outside its lifecycle management. Choosing between the two is a typical design decision when data already lives in existing storage locations.
The functional areas named in the Microsoft Learn documentation go beyond basic permissions. They include access control with attribute-based policies and row and column filters, discovery, lineage, audit logs exposed as a system table, data classification, data quality monitoring, data sharing across organizations and clouds, and AI governance. This means one layer covers both analytical data and AI assets such as models, rather than treating them as separate governance problems.
On availability, Unity Catalog is turned on automatically for workspaces created after November 2023. Workspaces created earlier must be upgraded to use it. The documentation also notes that an open-source implementation of Unity Catalog exists, and it points to a separate page for requirements and limitations that should be checked before planning a rollout.
The M365.FM episode "Azure Databricks - Simply Explained" places Unity Catalog within its overview of the platform. Its outline covers what Azure Databricks is, why the lakehouse idea exists, Apache Spark, notebooks, clusters and Delta Lake, and then groups "Workspaces, Accounts, and Unity Catalog" together as one topic. The episode also closes with guidance on when to use Databricks, Fabric or Synapse. In that framing, Unity Catalog is the part that explains how several workspaces under one account can share a consistent governance model, which is relevant when comparing Databricks with other Microsoft data platforms.
For people searching for the term, the short version is that Unity Catalog answers the questions of who can access which data or model, where that data came from, who used it, and how it can be found or shared. Because it is enforced at the point where tables are queried or models are called, and because lineage and audit records are generated automatically, it forms the basis for governance in Azure Databricks rather than an optional add-on.
Overview compiled with AI from M365.FM episodes and the facts on this page. You can improve it with the edit button below.
Type
Product
Key points about Unity Catalog
Key points about Unity Catalog from the episodes, each with its source episode:
Learn what Azure Databricks actually is, how Apache Spark powers it, and when to use Lakehouse architectures over traditional data warehouses. Apache Spark started as a research project at a university in 2009, and it grew into the engine most big data…
Frequently asked questions
Questions marked with AI were written with AI from the facts on this page.
What is Unity Catalog in Azure Databricks? (AI)
Unity Catalog is the unified governance layer for data and AI in Azure Databricks. It controls access, tracks lineage automatically and records activity for auditing across workspaces.
How are objects organized in Unity Catalog? (AI)
Tables, views, volumes, functions and models use a three-level namespace of catalog, schema and object. Privileges on these securable objects are granted to users, service principals or groups.
Is Unity Catalog enabled by default? (AI)
According to Microsoft Learn, it is on automatically for workspaces created after November 2023. Workspaces created before that must be upgraded to use it.
Short info above, community knowledge below: add practical tips, examples or corrections. New contributions are reviewed first; trusted members publish directly.
Experts and companies for this topic
Nobody has listed this topic yet. Add it to your profile.