Get certified for free when you join Fabric Data Days 2026 and dive into Fabric, Power BI, SQL, AI, and other essential data skills.
Join nowData Days is here! Join us now for 60+ days of learning, challenges, and connection. Learn more
I've been getting this error with DataFlow Gen2 and have narrowed it down to a pivot step I am doing. I have a super simple table with 3 fields, Key, Name, and date; I'm pivoting on name. Any ideas how to troubleshoot?
Solved! Go to Solution.
So after a bunch of trial and error, I finally found the issue; after the pivot I had some rows in [error] and once I removed them it loaded successfully.
So after a bunch of trial and error, I finally found the issue; after the pivot I had some rows in [error] and once I removed them it loaded successfully.
Hi @mklevemann ,
Thanks for reaching out to Microsoft Fabric Community.
The error code 104100 in Dataflow Gen2 usually comes up as a generic evaluation error. This can happen in situations where staging is enabled on intermediate queries or when the query structure introduces conflicts. Disabling staging on the related queries or simplifying the transformation has helped resolve this in some situations, as discussed here:
Solved: Re: Dataflow Gen2 (CI/CD) Non-deterministic failur... - Microsoft Fabric Community
Solved: DataFlow Gen2 104100 Couldn't refresh the entity b... - Microsoft Fabric Community
Since you’ve narrowed it down to the pivot step, the error could be related to duplicate or inconsistent Name values that generate duplicate column names after pivoting, or multiple Key–Date values under the same Name which require an aggregation. A mismatch in the data type of the column being aggregated can also lead to this issue.
If the error persists after verifying the above points, I recommend raising a support ticket to have this investigated further.
Here is the guide on how to raise a support ticket: How to create a Fabric and Power BI Support ticket - Power BI | Microsoft Learn
Hope this helps. Please reach out for further assistance.
Thank you.