Forum Discussion
Dataflow is not refreshing
Hi RadkaProefo ,
We can't check what the error is with the IDs etc you provided, are there other error messages?
You can provide specific error messages via refresh history so that we can figure out the error. I have a related datafolw refresh troubleshooting article here that you can refer to.
Troubleshooting dataflow issue - get data from dataflow - Power Query | Microsoft Learn
If it does not help, please provide more details with the error.
Best Regards,
Yifan Wang
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
I got the exact same problem. It's been like this for weeks now. I tried everything, and even though the on-prem data loads into Power Query/DF Gen2, the automatic refresh that starts after you click publish ALWAYS FAILS!
What's the solution?
Hello,
I still try to find a solution. I had a call with Power BI Microsoft support and they told me, that they cannot solve it, because it is a problem of Excel and I have to asi Excel Microsoft support.
I tried something else yesterday. I created 2 Dataflows Gen2: DF1 and DF2. In DF1, connect to your data source, do transformations, etc, AND DO NOT specify a destination. Then click Publish. This should create a table in the DataflowsStagingLakehouse (in my case all the source column names are lost and replaced by generic column names such as Column1, Column2, etc..., but I get them back at the end of this process). Wait til the refresh is finished.
Connect DF2 to DF1 AND, this time, specify a destination in Fabric (could be the default staging lakehouse or... actually I haven't tested with another type of destination other than the default staging lakehouse, so don't know if it'll work too for other destination types). Click Publish. Wait til the refresh is finished.
Go to your destination and you should find a table with the data and all the correct column names. In my case, the default staging lakehouse contains 3 copies of the same data: once because of DF1 (even though no destination was specified, the Enable staging option was on), twice because of DF2 (same reason as DF1), and a third copy being the final result, which is the output from DF2. That's a lot of waste storage if you ask me, but so far the only solution I found to ingest data from an on-prem DB.
