Introduction
Databricks Unity Catalog provides a unified governance layer for managing data access and organizing datasets efficiently. One of its core principles is the three-level namespace, which ensures consistent data organization across different workspaces.
In this article, we'll focus on.
- Understanding the Metastore and its role.
- Explaining the three-level namespace structure.
- Querying data using SQL in a Databricks notebook.
Challenges
Before the Unity Catalog, Databricks stores the data within the workspace's Hive Metastore. This led to significant data-sharing challenges across teams and departments.
- Data Silos Across Workspaces: Each Databricks workspace had its own Hive Metastore, meaning data was not accessible across different workspaces.
- Complex and Inefficient Data Sharing: Alternatively, teams resorted to external storage solutions (e.g., mounting an external Data Lake), which required additional permissions and configurations.
- Lack of Centralized Governance: There is no single source of truth for auditing and governance, making it difficult to track who accessed what data.
Unity Catalog Metastore and three-level namespace

The top-level container is the Metastore. There are two key differences between Unity Catalog's Metastore and the Hive Metastore.
- This Metastore is defined at the regional level.
- It is the centralized metadata and governance layer for organizing and securing data across multiple workspaces.
Below the Metastore, Unity Catalog structures data in a three-level namespace:
- Catalog: The first-level container. You can define multiple catalogs in your Metastore.
- Schema: The next level is schema, and it's often referred to as Database. You can have multiple schemas within a Catalog.
- For the third level, we have Table, View, Function, and Volume.
- The table stores data in delta format.
- Volume allows you to store data in any format. This ranges from structure data (db table, XML), semi-structure (JSON), or binary file (Zip file).
Let's dive into the article on how to utilize the three-level namespace.
In this article, we will use the NYC Taxi sample data from Databricks.
We will be using Databricks' SQL Query for our demo.
- After you log in to Databricks, click on the 'SQL Editor' menu.
- A New Query will be created automatically.
- You can create additional queries by clicking on the '+' next to the query tab.





Join the conversation! Your thoughts help the community grow.