Incremental Refresh in Fabric Dataflow with MySQL ODBC

[martinacahya]

Hi everyone,

I’m trying to implement the built-in incremental refresh feature in a Dataflow (Gen2) in Microsoft Fabric.
My data source is MySQL connected via ODBC.

However, when I try to publish the dataflow, it fails — it can’t even be published.

Has anyone successfully implemented incremental refresh in Fabric Dataflow with a MySQL ODBC source?
Would love to know if there’s any workaround or specific configuration needed.

Thanks in advance!

[v-pnaroju-msft · Accepted solution]

Thankyou, @tayloramy and @lbendlin for your responses.

Hi martinacahya,

We appreciate your inquiry submitted via the Microsoft Fabric Community Forum.

Based on my understanding, the issue arises because Fabric Dataflow (Gen2) incremental refresh validation requires full query folding to the data source. Although the Power Query Editor displays a green folding indicator, ODBC based MySQL connectors do not guarantee complete pushdown of the incremental filter to the backend. Consequently, the service blocks the Dataflow publish and displays the “query with incremental refresh is not folding” error.

Please follow the steps below, which may help resolve the issue:

1. In your Dataflow Gen2 → Incremental Refresh settings → Advanced options, uncheck “Require incremental refresh query to fully fold.” This relaxes the validation check and allows publishing, although refresh performance may be slower.
2. Replace the ODBC connection with the native MySQL connector, if your environment supports it. Native connectors provide better folding support and are fully compatible with incremental refresh.
3. Load your MySQL data (via ODBC) into a Fabric Lakehouse or SQL endpoint first. Then configure incremental refresh on a secondary dataflow that reads from this Fabric native source, ensuring full folding and partition support.

Additionally, please refer to the links provided below:
Incremental refresh in Dataflow Gen2 - Microsoft Fabric | Microsoft Learn
Power Query MySQL database connector - Power Query | Microsoft Learn
Understand and optimize dataflows refresh - Power BI | Microsoft Learn

If the issue persists after performing the above steps, and since you have already raised a support ticket, we request that you monitor updates from the support team regarding this matter.

Should you have any further queries, please feel free to contact the Microsoft Fabric community.

Thank you.

[tayloramy]

Hi @martinacahya, 

Can you show us the details of that error? 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution. 

[lbendlin]

Says "Please contact Customer Support".  Have you done that?

If you have a Pro license you can open a Pro ticket at https://admin.powerplatform.microsoft.com/newsupportticket/powerbi
Otherwise you can raise an issue at https://community.fabric.microsoft.com/t5/Issues/idb-p/Issues .

[martinacahya]

Yes, I have opened ticket as you suggested. Thanks!

[tayloramy]

Hi @martinacahya, 

Can you share the details about the failure? 

Have you tried using a created date for the filter rather than a modified date? 

Can you tell us the structure of your source table and the behaviour? Are records only added, or are records modified? The existance of a lastModifiedDate leads me to assume that they can be modified, and for the incremental refresh, when records can be modified as well as added, a seperate created date and modified date are required. 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution.

[martinacahya]

Hi @tayloramy 

I've tried using a created date for the filter instead of the modified date, as you suggested. However, the result is still the same — the dataflow can’t even be published. Here are the details about the failure: it says that the query isn’t folding. 

However, as you can see in the dataflow, the query is fully folding (indicated by the green icon on the right side of the Applied Steps), which makes me confused.

The source table contains daily sales order data. It basically consists of delivery address, delivery cost, customer data, salesman, and several date-related columns, as shown below.

The records are added for new sales orders and modified for existing ones whose status changes — for example, from Shipped to another status.

Hope this helps give you a clearer picture of the issue.