If multiple tables are present in the destination, create one shortcut per table. The following table contains the scenario best-practices for each item type when using it over shortcuts. Microsoft Fabric Lakehouse supports tables defined over OneLake shortcuts, to provide utmost compatibility and no data movement. Auto discovery also applies to tables defined over OneLake shortcuts. The table references are displayed under the Tables section of the Lakehouse explorer user interface. It has a key capability of discovering and displaying tables that are described in the metadata repository and in OneLake storage. The Lakehouse explorer provides a tree-like view of the objects in the Microsoft Fabric Lakehouse item. The following table contains the configuration differences between Azure Synapse Analytics and Microsoft Fabric Runtime for Apache Spark. Configuration differences with Azure Synapse Analytics Currently, the Lakehouse explorer only renders table objects. The Lakehouse explorer user interface experience varies depending on table type.

Tables defined using PARQUET, CSV, AVRO, JSON, and any Apache Hive compatible file format work as expected. Microsoft Fabric Lakehouse and the Apache Spark engine support all table types, both managed and unmanaged this includes views and regular non-Delta Hive table formats.

Advanced Apache Spark users may revert configurations to previous values to better align with specific scenarios. In Microsoft Fabric, key performance features are turned on by default. Microsoft Fabric Runtime for Apache Spark uses the same foundation as Azure Synapse Analytics Runtime for Apache Spark, but contain key differences to provide a more streamlined behavior across all engines in the Microsoft Fabric service. Big data, Apache Spark and legacy table formats For a more comprehensive introduction to the Delta Lake table format, follow links in the Next steps section.

Lakehouse design