Forum Discussion
Dataflow lineage when deployed not updated
Hi mcolbert - This issue is a common scenario when working with multiple environments (e.g., UAT and PRD) in Power BI, especially when dataflows reference each other.
- After deploying Dataflow B to PRD:
- Open Dataflow B in the PRD workspace.
- Edit the queries in Dataflow B.
- Update the references to point to Dataflow A in PRD.
- Save and refresh Dataflow B in PRD.
This ensures that Dataflow B in PRD references the correct Dataflow A in PRD.
To avoid manual updates, you can use parameters to dynamically define the workspace and dataflow references:
Step 1: Create a parameter (e.g., WorkspaceID) in both Dataflow A and B.
In Dataflow A, the parameter is not directly used but ensures consistency.
In Dataflow B, use the parameter to reference Dataflow A dynamically.
Step 2: During deployment, update the parameter value to match the PRD workspace's ID.
This approach requires some initial setup but streamlines deployments.
If you deploy frequently, you can automate the process using the Power BI REST API or tools like PowerShell:
- Use the API to update the connections in Dataflow B after deployment.
- Automate the reference update to point to Dataflow A in PRD.
Dataflows - Get Dataflow - REST API (Power BI Power BI REST APIs) | Microsoft Learn
Dataflows - Update Dataflow - REST API (Power BI Power BI REST APIs) | Microsoft Learn
Thanks for the reply. Yes, I could use parameters like I do for other items, but this should just work.
I did a quick test this am and confirmed my suspicion that if you allow the entities from Dataflow A to be load enabled (making it a linked entity) in Dataflow B, then the deployment fixes the linkage. If the entity is not loaded and just referenced, the linkage is not fixed. This seems like a bug. I can't think of a reasonable explanation that this would be by design.
Microsoft, if you are listening, please fix this.