Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Shortcut transformations convert raw files into Delta tables that stay always in sync with the source data.
The transformation is executed by Fabric Spark compute, which copies the data referenced by a OneLake shortcut into a managed Delta table so you don’t have to build and orchestrate traditional ETL pipelines yourself.
Note
Shortcut transformations are currently in public preview and are subject to change.
Why use shortcut transformations?
- No manual pipelines – Fabric automatically copies and converts the source files to Delta format; you don’t have to orchestrate incremental loads.
- Frequent refresh – Fabric checks the shortcut every 2 minutes and synchronizes any changes almost immediately.
- Open & analytics-ready – Output is a Delta Lake table that any Apache Spark–compatible engine can query.
- Unified governance – The shortcut inherits OneLake lineage, permissions, and Microsoft Purview policies.
- Spark based – Transforms build for scale.
Prerequisites
| Requirement | Details |
|---|---|
| Microsoft Fabric SKU | Capacity or Trial that supports Lakehouse workloads. |
| OneLake shortcut | A shortcut targeting a folder that contains homogeneous CSV files. |
| Workspace role | Contributor or higher. |
Supported sources and destinations
| Source file format | Destination | Notes |
|---|---|---|
| CSV (UTF-8, comma-delimited) | Delta Lake table in the Lakehouse / Tables folder | Additional formats (Parquet, JSON) will be added during preview. |
Set up a shortcut transformation
- In your lakehouse, select + New > Shortcut transformation (preview).
- Choose shortcut – Browse to an existing OneLake shortcut that points to the folder with your CSV files.
- Configure transformation:
- Table name – Provide a friendly name; Fabric creates it under /Tables.
- Delimiter – Select the character used to separate columns (comma, semicolon, pipe, tab).
- First row as headers – Indicate whether the first row contains column names.
- Select Create.
Fabric Spark compute copies the data into a Delta table and shows progress in the Manage shortcuts pane.
How synchronization works
After the initial load, Fabric Spark compute:
- Polls the shortcut target every 2 minutes.
- Detects new or modified files and appends or overwrites rows accordingly.
- Detects deleted files and removes corresponding rows.
Monitor and troubleshoot
- Open the lakehouse and select Shortcuts in the left pane.
- Choose the shortcut that feeds your transformation.
- In the details pane, select the Manage shortcut tab to view:
- Status – Last scan result and current sync state.
- Activity log – Chronological list of sync operations with row counts and any error details.
From this tab you can also Pause or Delete the transformation if needed.
Limitations (preview)
- Only CSV sources are supported.
- Files must share an identical schema; schema drift isn’t yet supported.
- Transformations are read-optimized; MERGE INTO or DELETE statements directly on the table are blocked.
- Available only in Lakehouse items (not Warehouses or KQL databases).
Clean up
To stop synchronization, delete the shortcut transformation from the lakehouse UI.
Deleting the transformation doesn’t remove the underlying files.