Fabric Workspace - Semantic Model - Fabric Warehouse - Model Update Issues

[royclack]

Hi,

We have a direct lake Semantic Model pointing at a Fabric Warehouse and have altered some tables with additional columns and want to expose them and update the Semantic Model. Please note this is a custom Semantic Model and I am not referring to the Default Semantic Model on the Fabric Warehouse item itself and that model shows the new columns. For the custom Semantic Model the refresh and confirm of the existing tables in the model returns the following error:

{"error":{"code":"ExternalServiceFailed","pbi.error":{"code":"ExternalServiceFailed","parameters":{},"details":[{"code":"ModelingServiceError_Reason","detail":{"type":1,"value":"Error in ModelingEngineHost"}},{"code":"ModelingServiceError_Location","detail":{"type":1,"value":"ModelingEngineHost"}},{"code":"ModelingServiceError_ExceptionType","detail":{"type":1,"value":"InvalidOperationException"}},{"code":"ModelingServiceError_Message","detail":{"type":1,"value":"Sequence contains more than one matching element"}}]}}}

Has anyone else seen this issue and/or have a workaround resolution?

Regards

Roy

[BhavinVyas3003]

The error occurs because your custom Semantic Model is out of sync with the updated Fabric Warehouse schema likely due to duplicated or ambiguous column references. To fix it, open the model in Power BI Desktop, remove and re-add the affected tables to refresh metadata, and check for duplicate column names in the warehouse using INFORMATION_SCHEMA.COLUMNS.

[royclack · Accepted solution]

Thanks for any input.

As an update I have sorted myself, in realising that the Semantic Model had added to it in a current iteration of development an aliased table following a migration approach from a on-premise solution and that you can't operate with aliased tables in Direct Lake models and they need to physicalized - in my case in the Fabric Warehouse.

I worked this out through a process of elimination and after deleting the aliased reference from the model I had no issues.