Compare
SSIS to Microsoft Fabric: what are the options?
Microsoft offers three routes: run packages with the Invoke SSIS Package activity in Fabric Data Factory (in preview), keep them on the Azure-SSIS integration runtime and call them from Fabric, or rebuild them as Fabric pipelines and Dataflow Gen2. Running packages as they are needs the fewest changes when they fit the preview's limits, and conversion is for teams that want Fabric code without a manual rebuild. Here are the facts, sourced from Microsoft Learn.
Side by side
How do the SSIS to Fabric routes differ?
Microsoft's three routes, next to Novuz as one example of a third-party conversion accelerator.
| Invoke SSIS Package (Fabric) | Azure-SSIS IR (Azure Data Factory) | Rebuild in Fabric Data Factory | Novuz | |
|---|---|---|---|---|
| What you end up with | Your existing .dtsx packages, stored in OneLake and run from a Fabric pipeline. | Your existing packages in SSISDB on Azure SQL Database or SQL Managed Instance, run by Azure Data Factory and called from Fabric. | Fabric pipelines and Dataflow Gen2 (Power Query or Spark-based MDF transforms), built by your team. | Converted SSIS objects for Microsoft Fabric, with a reconciliation report for every object. |
| Changes to packages | Microsoft describes it as moving SSIS workloads into Fabric with minimal changes. | Deployed with the project deployment model. Packages that pass validation are upgraded to the latest package format. | Each package is redesigned as pipelines and dataflows. | Fixed, versioned rules convert about 95% of objects with no model calls. A resolver handles the rest, and an engineer approves each change. |
| Status | Preview. Subject to change before general availability. | Available in Azure Data Factory. Not yet available in all Azure regions. | Pipelines available. MDF transforms in Dataflow Gen2 are in public preview. | Converts SSIS today. Typically 95% or more of objects are converted with no manual edit. |
| Published limits | OneLake-stored packages only. No on-premises or private-network sources or destinations. No custom or third-party components. | Additional drivers and licensed third-party components can be installed. Microsoft documents connecting to on-premises data sources and file shares. | See Microsoft documentation | Scope is fixed at assessment. Objects the rules do not convert go to the resolver for engineer approval. |
| Writing to Fabric Data Warehouse | See Microsoft documentation | Supported with Microsoft Entra ID authentication, staging in ADLS Gen2 and COPY INTO. Fabric Data Warehouse supports a subset of T-SQL. | Pipelines work with Fabric lakehouses and warehouses directly. | Microsoft Fabric is a supported target. |
| Where it runs | Fabric workspace, with 4 vCores per workspace, fixed during preview. | Azure-SSIS integration runtime hosted by Azure Data Factory. | Fabric capacity. | Conversion runs inside your environment, as a Docker container or VM. |
| Billing | 1.35 CU hours per vCore under the SSIS in Fabric meter, with a 30-minute time-to-live after the last activity. | Azure Data Factory pay-as-you-go, per activity run, data movement and compute. | Fabric capacity (F SKU), charged for activity runs and pipeline data movement. | Fixed price per scope, set after a $3,000 assessment delivered in three business days. |
| Equivalence evidence | Package execution logs written to OneLake when logging is enabled. | See Microsoft documentation | See Microsoft documentation | A reconciliation report for every object. Reconciliation is typically above 95% on first pass. |
Re-verified against Microsoft Learn documentation. Last checked: 2026-10-30.
When to choose
Which SSIS to Fabric route should I choose?
- Choose the Invoke SSIS Package activity if your packages use cloud sources only, need no custom or third-party components, and you can run on a preview feature.
- Choose the Azure-SSIS integration runtime, called from Fabric, if packages need on-premises sources or third-party components today. Microsoft's upgrade guide describes this pattern.
- Choose a rebuild in Fabric pipelines and Dataflow Gen2 if the package count is small and your team wants to redesign the logic.
- Choose a conversion accelerator such as Novuz if you want packages converted to Fabric code rather than rebuilt by hand, with a reconciliation report for every object.
- Third-party options fall into two groups: tools that convert packages to code, and services firms that rebuild them. Ask each what it produces, where it runs and how it shows equivalence.
FAQ
Frequently asked questions
Can I run SSIS packages in Microsoft Fabric?
Yes, in preview. The Invoke SSIS Package activity runs .dtsx packages stored in OneLake from a Fabric pipeline. During preview it cannot reach on-premises or private-network sources and does not support custom or third-party components.
What happens to the Azure-SSIS integration runtime when we move to Fabric?
Microsoft's upgrade planning guide says to execute SSIS packages using Azure Data Factory pipelines and call them from Fabric. Microsoft's comparison page lists Azure-SSIS integration runtimes as a future capability in Fabric.
Can existing SSIS packages write to a Fabric Data Warehouse?
Yes. Microsoft's guide says to switch authentication to Microsoft Entra ID, stage data in ADLS Gen2 and load it with COPY INTO from an Execute SQL Task. Fabric Data Warehouse supports a subset of T-SQL, so some packages may fail on unsupported features.
Sources and further reading
- Microsoft Learn, Use the Invoke SSIS Package activity in a pipeline (Preview)
- Microsoft Learn, Upgrade planning for Azure Data Factory to Fabric Data Factory
- Microsoft Learn, Differences between Data Factory in Fabric and Azure
- Microsoft Learn, Mapping data flow transforms in Dataflow Gen2 (Preview)
- Microsoft Learn, Deploy and run SSIS packages in Azure
- Microsoft Learn, Tutorial: Integrating SSIS with Fabric Data Warehouse
Written by Ted Ghose, Co-founder and CTO, Novuz
Last reviewed:
Find out what your logic debt is costing you.
Every object counted, lineage, dead code and a fixed price in three business days. $3,000, credited.
Not sure which door? The scorecard takes ten minutes and tells you.