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.
Your Azure Data Factory pipelines already power critical workflows. In this article, you learn how to bring them into Fabric to unlock a more integrated, analytics-ready experience. This built-in migration experience helps you modernize your existing Azure Data Factory workloads in a few simple clicks.
The migration experience helps you:
- Assess pipeline readiness directly in Azure Data Factory.
- Understand compatibility gaps at the pipeline and activity level.
- Migrate supported pipelines to a Fabric workspace.
- Plan next steps for items that need updates or that are coming soon.
This assessment-first approach helps ensure migrations are intentional, transparent, and incremental. You can upgrade pipelines at your own pace and validate results before switching production workloads.
Prerequisites
Before you start, make sure you have:
- An existing Azure Data Factory instance with pipelines.
- Access to a Microsoft Fabric tenant.
- A Fabric workspace in the same Microsoft Entra ID tenant as the Azure Data Factory instance.
Step 1: Assess your pipelines for migration
To run the migration assessment, in your Azure Data Factory authoring canvas, select Migrate to Fabric (Preview) > Get started (preview) to evaluate pipelines and activities for migration readiness.
Step 2: Review and understand assessment results
Both the factory and individual pipelines are categorized with a readiness status: Ready, Needs review, Coming soon, or Unsupported. You can also export your assessment results to a CSV file to support offline review and remediation planning.
Each pipeline and activity is assigned one of the following statuses. Use these results to plan your migration.
| Status | Meaning |
|---|---|
| Ready | Fully supported and safe to migrate. |
| Needs review | Requires minor updates, such as parameter or configuration changes. |
| Coming soon | Support is planned; migrate later. |
| Not compatible | No Fabric equivalent; redesign required. |
Step 3: Select a Fabric workspace and mount your Azure Data Factory
After you review the assessment, select Next to mount your Azure Data Factory to a Fabric workspace and continue the migration flow in Fabric. Mounting lets you reference your Azure Data Factory (ADF) instance inside a Fabric workspace without migrating, copying, or altering the Azure Data Factory environment.
After mounting completes, select Continue in Fabric to proceed with migration steps.
Step 4: Migrate pipelines
Continue migration from the Fabric experience by selecting Migrate to Fabric (Preview).
Select the pipelines you want to migrate.
Step 5: Map linked services to Fabric connections and complete migration
Select Review connections to map Azure Data Factory linked services to Fabric connections and then select Confirm.
The migration experience attempts to automatically create connections for authentication methods that can be safely and reliably mapped from Azure Data Factory to Fabric’s managed identity and security model without requiring customer‑managed infrastructure or network configuration.
Connections automatically created during migration (supported only)
| Connector | Azure Data Factory authentication | Fabric authentication |
|---|---|---|
| Azure Blob Storage | Account key; Shared access signature (SAS); Service principal; System-assigned managed identity | Account key; Shared access signature (SAS); Service principal; Workspace identity (system-assigned managed identity) |
| Azure Data Lake Storage Gen2 | Account key; Shared access signature (SAS); Service principal; System-assigned managed identity | Account key; Shared access signature (SAS); Service principal; Workspace identity (system-assigned managed identity) |
| SQL Server | Basic authentication (SQL authentication); Service principal; System-assigned managed identity | Basic authentication; Service principal; Workspace identity (system-assigned managed identity) |
| Azure SQL Database | Basic authentication (SQL authentication); Service principal; System-assigned managed identity | Basic authentication; Service principal; Workspace identity (system-assigned managed identity) |
| Azure Data Explorer (Kusto) | Service principal; System-assigned managed identity | Service principal; Workspace identity (system-assigned managed identity) |
| Azure Cosmos DB for NoSQL | Account key | Account key |
| Azure Cosmos DB for MongoDB | Basic authentication | Basic authentication |
| Azure SQL Managed Instance | Account key; Service principal | Basic authentication; Service principal |
| Azure Database for PostgreSQL | Basic authentication | Basic authentication |
| Azure Database for MySQL | Basic authentication | Basic authentication |
| MySQL | Basic authentication | Basic authentication |
| PostgreSQL | Basic authentication | Basic authentication |
For other connections, either select an existing Fabric connection or create new connections by using the modern Get Data experience or from workspace settings. Then select Confirm.
This action starts the migration of the selected pipelines to the root folder in the Fabric workspace. A confirmation message appears when the migration completes successfully.
After migration completes, go to your Fabric workspace to review the migrated pipelines. Each pipeline is created under the workspace and prefixed with its source factory name. You can open each pipeline to review and validate it before you continue with further configuration or testing.
Note
If you don't map any connections during this step, pipelines still migrate. Activities within those pipelines are deactivated, and you can configure them later in Fabric.
After migration completes, validate the pipelines in the Fabric Data Factory experience.
Migration behavior
- Pipelines migrate into a Fabric Data Factory workspace.
- Pipeline names must be unique within a workspace.
- If a pipeline with the same name already exists, the migration tool skips that pipeline.
- To ensure uniqueness, migrated pipelines use the following naming format:
<Source factory or workspace name>_<Pipeline name>. - The migration flow includes a mounting step that lets you view your existing factory structure in Fabric before migration.
Post-migration validation
After migration, complete the following tasks:
- Validate all connections and credentials.
- Recreate global parameters as variable libraries.
- Re-enable and configure triggers (disabled by default).
- Run end-to-end tests to confirm pipeline behavior.
- Validate migrations in a nonproduction environment before you migrate production workloads.
What's out of scope
The following items aren't supported in the UX-based migration experience today. Pipelines that use these features require redesign or alternate migration approaches.
| Category | Out-of-scope item | Details |
|---|---|---|
| Integration runtimes | Self-hosted integration runtime (SHIR) | Self-hosted integration runtimes can't be migrated. Replace with the Fabric on-premises data gateway (OPDG). |
| Managed virtual network integration runtime (Managed virtual network IR) / Virtual network–injected integration runtime (VNet – Virtual network) | Fabric doesn't support migrating managed virtual network integration runtimes. The Fabric virtual network gateway uses a different model and requires reconfiguration. | |
| SQL Server Integration Services integration runtime (SSIS IR) | Infrastructure migration, including SQL Server Integration Services integration runtimes, isn't supported. | |
| Workload types | Azure Data Factory change data capture (CDC) | Change data capture workloads are out of scope and don't migrate. |
| Apache Airflow assets | Directed acyclic graph (DAG)–based orchestration from Apache Airflow can't be migrated to Fabric. | |
| Unified Structured Query Language (U‑SQL) / Azure Data Lake Analytics | Deprecated services and not supported in Fabric. | |
| Cross‑cloud or Azure Machine Learning refresh workloads | Workspace identity support is in progress. These workloads don't migrate. | |
| Connectors | Long‑tail connectors (for example, SAP ERP Central Component (ECC), SAP Business Warehouse (BW), Multidimensional Expressions (MDX), SAP Core Data Services (CDS)) | Fabric has no equivalent connectors. Redesign is required. |
| Marketing and finance software‑as‑a‑service connectors (HubSpot, Google Ads, QuickBooks, Shopify, Xero) | Not supported today. | |
| Triggers and orchestration | Custom event triggers | Custom event triggers can't be migrated. |
| Storage event triggers | Support is coming soon. | |
| Tumbling window triggers | Known as Interval‑based scheduling in Fabric. Watermark and backfill workloads must be redesigned. | |
| Chaining or dependency triggers | Chaining and dependency trigger semantics aren't supported yet. | |
| Security and authentication | Advanced configurations (customer‑managed keys (CMK), dual tokens, federated identity credential (FIC) flows) | Unsupported workspace identity or service principal authentication models don't migrate. |
| Certificate‑based authentication (Web activity) | Unsupported and requires redesign. | |
| User‑assigned managed identity (UAMI) support | Use workspace identity (WI) as a workaround. | |
| Parameterization and metadata | Global parameters | Support is coming soon. Recreate by using Fabric variable libraries. |
| Dynamic linked services (parameterized connections) | Not supported. Each permutation must be a separate connection and can't migrate. | |
| Metadata‑driven pipelines | Highly dynamic linked service or dataset‑driven patterns can't migrate. | |
| Activities and compute | Azure Synapse Spark job definition (SJD) or notebook | Partially supported. Requires redesign into Fabric notebooks or Spark jobs. |
| Mapping data flows (MDF) | Support is coming soon. | |
| Web, webhook, or HTTP activities with custom authentication or headers | Complex authentication scenarios must be rebuilt manually. | |
| Notebook pool environment settings | Not supported. Migration is blocked. | |
| Batch or custom activity workspace identity support | Missing workspace identity support blocks migration for these activities. | |
| Copy activity upsert into Lakehouse tables | Not supported. Requires copy to staging and a notebook MERGE operation. |
FAQ
Does the assessment change my factory?
No. The assessment is read-only. It scans your factory configuration and surfaces findings in the side pane without modifying pipelines, activities, or settings. You can safely run it to understand migration impact before taking any action.
Can I rerun the assessment or migration after making changes?
Yes. You can rerun the assessment at any time during validation. If you rerun migration for the same pipelines, you must first delete the previously migrated pipelines in Fabric, because pipeline names must be unique within a workspace.
Does mounting Azure Data Factory migrate my pipelines?
No. Mounting is just a snapshot of your existing Azure Data Factory in a Fabric workspace. No pipelines are migrated until you explicitly start migration by selecting the Migrate to Fabric (Preview) button from your mounted data factory in Fabric.
Will triggers migrate automatically?
Schedule triggers are migrated automatically but disabled after migration by design. You must manually re-enable them in Fabric. All other triggers must be manually reconfigured and re-enabled after you validate the migrated pipelines.
Do unsupported items block the entire migration?
No. Unsupported activities affect only the pipelines that contain them. Other supported pipelines can migrate independently. The assessment clearly identifies which pipelines require redesign.
Can I migrate without mapping connections?
Yes. Pipelines still migrate, but activities that depend on unmapped connections are deactivated. You must configure the required Fabric connections and re-enable those activities before running the pipelines.
Can I validate migrations before moving production workloads?
Yes. Microsoft recommends validating migrations in a nonproduction environment, confirming connections, triggers, and end-to-end execution before migrating production pipelines.