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.
Note
The Spark Synapse to Fabric Migration Assistant is currently in public preview.
Use the Spark Synapse to Fabric Spark Migration Assistant to migrate Spark workloads from Azure Synapse Analytics to Microsoft Fabric through a guided workflow.
In public preview, the assistant migrates Spark pools, notebooks, Spark job definitions, and lake databases.
Migration doesn't move your data. The assistant copies and transforms supported Spark items for Fabric and generates a migration report for each run.
The assistant supports migration of the following items:
- Spark pools are migrated to Fabric Pools and corresponding environments.
- Notebooks and associated environments are migrated.
- Spark job definitions (SJDs) are migrated with associated environments.
- Lake databases are mapped to Fabric schemas, and managed Delta tables are migrated by creating OneLake catalog shortcuts.
For lake database migration details, the default Synapse database maps to the dbo schema in Fabric, and more databases are migrated as schemas in the same Lakehouse.
How migration works
To migrate Apache Spark pools, notebooks, Spark job definitions, and lake databases to Fabric, the assistant guides you through a short workflow.
Copy and transform items
First, the assistant helps you locate the source workspace that contains the data engineering items you want to migrate. It then copies and transforms supported items so they can run in a Fabric workspace.
Monitor migration
During migration, you can monitor progress and review a summary of migrated data engineering items. When migration is complete, the new items are available in your Fabric workspace.
Get started with the migration assistant
Open a Fabric workspace, and then select Migrate from the toolbar.
In the wizard, select Data engineering items.
Specify the source Azure Synapse workspace details. You can search for the source workspace by subscription. After you select the subscription, you can select the source workspace from that subscription.
- Enter a migration Name.
- Select the Subscription and source Source workspace.
Select the target Fabric workspace where items should be migrated.
The source workspace and target Fabric workspace that you select in the wizard determine the migration direction.
Review available data gateway connections that you can access. Add a new connection if needed.
To create a new connection, select Add new.
The server and path are generated from the default Azure Data Lake Storage Gen2 account and container for the source workspace. This value is required because lake database migration creates shortcuts to the managed-table locations.
Optionally select the data gateway connection to use for this migration run.
Note
If you don't select a gateway connection, databases aren't migrated.
Review migration details, and then select Next.
Closing the dialog at this point allows you to cancel the migration:
Select Migrate to start migration.
Monitor migration progress. During migration, you can see the status and summary of migrated items.
The spinner indicates that migration started.
A notification at the top-right corner indicates the progress.
A link to the Migration Report is available.
After a few seconds, you're automatically redirected to the report.
Note
The migration report can also be accessed from the Monitoring Hub. You can search for the migration job by its migration name or filter by the Spark Job Definition item type.
Select the Migration Report link from the previous step to review results.
Note
Select Details to view additional information about errors or warnings encountered during migration.
Public preview limitations
- Synapse workspaces under a virtual network (VNET) can't be migrated.
- Spark configurations, custom libraries, and custom executor settings aren't migrated.
- Non-Delta table formats aren't supported.