"When a new row appears" report alert not firing on a Direct Lake (OneLake) model - refresh vs. reframe?

Hi @4EverAStudent ,

Thank you for sharing the detailed information.

Based on your description, your theory is certainly plausible. However, the fact that the new record is visible in the report confirms that the semantic model can see the data change, but it does not necessarily confirm that Activator received a new observation or evaluation event for that change.

As a troubleshooting step, I recommend testing with an explicit semantic model refresh, for example by using a Semantic Model Refresh pipeline activity, immediately after loading the new row. If the alert consistently fires after the explicit refresh but not after automatic Direct Lake updates, this would help indicate a difference between the automatic update behavior and the event path used by Activator.

For production scenarios involving batch-loaded OneLake data, you may also want to evaluate alternatives such as OneLake event-driven processing, SQL query-based alerts where applicable, or pipeline-driven notifications, as these provide a more explicit triggering mechanism.

If the behavior persists, please consider opening a Fabric support request and include the Activator rule ID, workspace details, and timestamps from both the automatic-update and explicit-refresh tests. This information can help the Fabric product team determine whether the behavior is expected or if there is a gap in the Direct Lake/Activator integration.

For additional information, please refer to the following Microsoft documentation:

• Direct Lake automatic updates and framing: https://learn.microsoft.com/en-us/fabric/fundamentals/direct-lake-how-it-works
• Direct Lake considerations and behavior: https://learn.microsoft.com/en-us/fabric/fundamentals/direct-lake-overview
• Using the Semantic Model Refresh activity in a Fabric pipeline: https://learn.microsoft.com/en-us/fabric/data-factory/semantic-model-refresh-activity

I hope this helps. Please let us know what you observe from the explicit refresh test.

If you have any further questions, please feel free to ask.

hi 4EverAStudent & thx for sharing urs issue here at Q&A portal,

ur theory is plausible, but I wouldn’t treat it as confirmed from the current docs. Direct Lake uses framing/reframing to point the semantic model at the latest Delta table state. A semantic model refresh triggers that reframe, but the report becoming current doesn’t necessarily prove that the Activator rule received a fresh evaluation event. Microsoft documents that report alerts can detect a new row in a table visual, but I couldn’t find a documented guarantee that every automatic Direct Lake reframe triggers that rule path. I’d test a manual or pipeline-triggered semantic model refresh immediately after loading the new row. If the alert fires consistently after that, it strongly suggests the automatic Direct Lake update isn’t triggering Activator evaluation the same way. The Semantic Model Refresh pipeline activity can be used for that test.

For a production design, I wouldn’t rely only on a report visual alert for batch-loaded OneLake data until Microsoft confirms the behavior. OneLake events, a Warehouse SQL query alert, or a pipeline-driven notification will usually give u a more explicit event boundary. Microsoft documents both OneLake event-driven pipelines and SQL query-based Activator alerts.

The cleanest next step is to reproduce it once with automatic Direct Lake updates and once with an explicit semantic model refresh, then send both timestamps and the Activator rule ID to Fabric support. That should tell the product team whether this is expected behavior or a Direct Lake/Activator gap.

rgds,

Alex

&

If my answer was helpful pls mark it and additional thx if u follow me at Q&A portal

and at my blog https://ctrlaltdel.blog/