Mirror OneLake in Cosmos DB database in Microsoft Fabric (preview)

Every Cosmos DB in Microsoft Fabric database is mirrored into OneLake in the open-source Delta Lake format. This feature doesn't require any extra configuration or setup and is automatically enabled when the database is created. This tight integration eliminates the need for ETL (Extract, Transform, Load) pipelines and ensures that Cosmos DB data is always analytics-ready.

This automatic mirror support enables scenarios including, but not limited to:

• Ad-hoc queries using the Transact SQL (T-SQL) query language
• Integration with Apache Spark
• Analytics over real-time data using notebooks
• Data science and machine learning workflows

Mirroring status

You can check the status of replication by navigating to the replication section for the database in the Fabric portal. This section includes metadata about replication including the status of the last sync.

SQL analytics endpoint queries

The mirrored database can be queried directly using the SQL analytics endpoint experience in the Fabric portal. At any point, you can use the portal to switch between the NoSQL-native data explorer and the T-SQL-native SQL analytics endpoint explorer.

Within the SQL analytics endpoint, you can query data using common T-SQL query language expressions like:

SELECT
  category,
  COUNT(*) AS quantity
FROM
  [<database-name>].[<container-name>]
GROUP BY
  category

Related content

• Learn about Cosmos DB in Microsoft Fabric
• Frequently asked questions about Cosmos DB in Microsoft Fabric
• Review limitations of Cosmos DB in Microsoft Fabric