Forum Discussion
Migrate Dataflow
Hi Pugtastic99_,
Thank you for reaching out to the Microsoft Fabric Forum Community.
Please review the response provided by SamsonTruong Poojara_D12 below, as it addresses your issue. It details the steps to manually migrate a dataflow, especially when Deployment Pipelines aren’t an option, including handling existing dataflows and Dataverse connections.
Solved: Moving dataflows from one environment to another e... - Microsoft Fabric Community
If this post helps, then please give us ‘Kudos’ and consider Accept it as a solution to help the other members find it more quickly.
Thank you.
Hi Pugtastic99_
To migrate an updated dataflow (especially one sourcing from Dataverse) from a development workspace to a test workspace—where an older version of that dataflow already exists—you can manually export and import the dataflow JSON definition. Start by going to the dev workspace, select the dataflow, and choose the "Export .json" option. This exports the M code and metadata for the dataflow. Then, go to the test workspace, locate the existing dataflow, and either delete it (if a clean overwrite is acceptable) or open it for editing and replace the queries manually using the JSON. Alternatively, create a new dataflow in the test workspace and use the "Import .json" option to upload the updated definition. Make sure that any referenced parameters, linked services, or gateway connections are also configured in the test workspace to match the dev environment, especially since Dataverse connectors rely on specific environment and table mappings. Finally, once the dataflow is updated or recreated, refresh it to validate that the structure and data load as expected. This manual approach is necessary when deployment pipelines are not usable due to limited stage support (e.g., when more than dev/test/prod are required).