Data Flow (Gen1) Refresh fails

[sgeheeb]

Hi,

I have a dimension table which is used in multiple data models, so I decided to move it to a data flow (for the ovious reasons).

In the Preview, everything looks finde, but the dataflow fails to refresh. When i download the .csv, the error message is empty (Only Activity-ID and Request-ID are provided).

The dataflow is located in a Pro Workspace and using PPU is not an option for this particular customer.

I suspect this has to do with the computed entities not being available for Pro workspaces.

The dataflow uses multiple tables, which are joinded, merged, etc..  However, only one table (the final result) has its "Load" activated. 

When I activate the "Load" for multiple tables, I get the "Bolt" Icon and the error message, stating that this won't fly in a Pro Workspace. 

In my current setup, only one table is loaded and both the bolt and the warning disappear. 

Can anyone confirm that the computet tables are still the issue, and the waning simply fails to display? Or could there be other reasons (which I might acutally be able to fix :))

Best,

Stephan

[Streamline]

Any update?

We experience similar problems with refreshing Gen1 dataflows on Business Central Saas.

[sgeheeb]

No, this seems to be the designed bahaviour. I am guessing that you are creating a computed entity (bolt icon) at some point, which requires a PPU license. My solution was to perform as many transformations as possible in the dataflow (without creating a computed entity) and take it from there in Power BI Desktop. Still not ideal, but better then nothing.

[DoubleJ]

hi, 

have you found a solution to this issue?

[sgeheeb]

Unfortunately, no 😞

[DoubleJ]

thanks for your quick reply.

[Coldfish]

Is it the source Oracle On premise ?

[sgeheeb]

Hi @Coldfish , no, the data source is Business Central SaaS, so basically REST-APIs.

[Anonymous]

Hi @sgeheeb ,

Checking the refresh history might provide more insights. You can do this by navigating to the dataflow settings and reviewing the refresh history. This might help you identify specific steps or entities causing the failure.

If computed entities are essential for your dataflow, consider breaking down the dataflow into simpler steps or using Power Query in Power BI Desktop to perform complex transformations before loading the data into the dataflow. And please ensure that your dataflow is optimized for refresh.

This includes minimizing the number of transformations and ensuring that only necessary tables are loaded.

Understand and optimize dataflows refresh - Power BI | Microsoft Learn

Best regards.
Community Support Team_Caitlyn

[sgeheeb]

Hi,

thanks for the reply. As I said, the refresh history of the dataflow does not contain any information on the nature of the error. This is the content of the .csv:

Requested on,Name des Dataflows,Aktualisierungsstatus für Dataflow,Tabellenname,Partitionsname,Status aktualisieren,Startzeit,Endzeit,Dauer,Verarbeitete Zeilen,Verarbeitete Byte (KB),Max. Commit (KB),Prozessorzeit,Wartezeit,Compute-Engine,Fehler,
2024-10-07 11:10:50,Artikel,Fehler,Artikel,NV,Fehler,2024-10-07 11:10:50,2024-10-07 11:13:21,00:02:31.7530,NV,NV,NV,NV,NV,NV,Error:  Request ID: 76f9ec38-f294-d8a2-ecd8-fbb205f3c66e Activity ID: df72a89c-316c-4347-95c8-9f8ad465fb23

The computation of the dimension table is currently done in Power BI Desktop (In four different models). I am trying to move the creation of the table into a dataflow, so it only needs to be done once. This involves many stepts (mostly joining data from different source tables), which cannot be simplified any further.

Best,

Stephan