Microsoft Fabric Blog explains the launch of schema-enabled lakehouses, outlining new organizational options, Spark compatibility, security features, and ongoing plans for improvements.

Lakehouse Schemas Now Generally Available in Microsoft Fabric

Schema-enabled lakehouses introduce a new way to organize and manage your tables in Microsoft Fabric. When creating new lakehouses, schema-enabled options are now the default, though users can still opt for classic non-schema lakehouses if preferred.

Key Features

  • Improved Table Organization: Schemas act like folders, helping users find and segment data efficiently.
  • Schema Shortcuts: Create pointers to internal tables in other schemas, or to external data sources such as ADLS Gen2.
  • Enhanced Spark Interoperability: Reference lakehouses across multiple workspaces and join them in a single query. Both schema and non-schema lakehouses are supported through naming conventions; future migration scenarios are planned.
  • Security and Performance: Support for OneLake security with RLS/CLS and Fabric Materialized Views.
  • Notebook Compatibility: To use schema-enabled lakehouses in a Notebook, ensure you have it pinned, or none pinned. Spark currently requires this configuration, but improvements for more flexible workspace modes are coming soon.

Upcoming Enhancements and Limitations

Several Spark-related features for schema lakehouses are rolling out soon:

  • Support for Spark Views
  • Shared lakehouse support across workspaces
  • User-Defined Functions (UDFs) for lakehouses
  • External ADLS table support
  • Workspace Private Links
  • Outbound Traffic Protection
Workarounds are currently available for these, with more detail at [Lakehouse schemas – Microsoft Fabric Microsoft Learn](https://learn.microsoft.com/en-us/fabric/data-engineering/lakehouse-schemas).

Existing Non-Schema Lakehouses

Non-schema lakehouses continue to be supported, and Microsoft is working towards full feature parity and interoperability. Tools will be introduced for a smooth transition from non-schema to schema-enabled lakehouses, ensuring users can benefit from new features without migration downtime.

References

This post appeared first on “Microsoft Fabric Blog”. Read the entire article here