Find everything you need to get certified on Fabric—skills challenges, live sessions, exam prep, role guidance, and more.
Get startedGrow your Fabric skills and prepare for the DP-600 certification exam by completing the latest Microsoft Fabric challenge.
Is it a recent change to Power BI service that new columns added to the table / view in the source database or warehouse will not flow through to the semantic model ? As a concequence we have to update on power BI desktop and re-publish the model for it to include the source table changes ?
I am pretty sure this wasn't the case 3 months back.
Solved! Go to Solution.
Hi @Hoping ,
I'd be surprised if columns were added automatically upon refresh. In the tabular model there is a static binding between semantic model columns and source columns. The columns you see is a static definition as part of the model, not coming dynamically from the source.
The reason is that data refreshes are technically partition refreshes, even if Power BI tables have only one partition except for incremental refresh tables. Since different partitions of the same table cannot have diffrent schemas, schema won't update upon refresh. There is no refresh type that includes a schema update. https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/refresh-dataset#datasetrefreshtype
If you refresh data in Power BI Desktop this includes a schema update and an update of all partitions (including a full table, all partitions update of tables for which incremental refresh is configured). With Premium workspaces you could implement this yourself, writing a script that first does the schema update through XMLA and then triggers the data refresh. Adding a new column is always an explicit schema update of the tabular model. Update Tabular Model when the underlying Relational Database schema changes | Microsoft Learn
Kind regrds,
Martin
Hi @Hoping ,
I'd be surprised if columns were added automatically upon refresh. In the tabular model there is a static binding between semantic model columns and source columns. The columns you see is a static definition as part of the model, not coming dynamically from the source.
The reason is that data refreshes are technically partition refreshes, even if Power BI tables have only one partition except for incremental refresh tables. Since different partitions of the same table cannot have diffrent schemas, schema won't update upon refresh. There is no refresh type that includes a schema update. https://learn.microsoft.com/en-us/rest/api/power-bi/datasets/refresh-dataset#datasetrefreshtype
If you refresh data in Power BI Desktop this includes a schema update and an update of all partitions (including a full table, all partitions update of tables for which incremental refresh is configured). With Premium workspaces you could implement this yourself, writing a script that first does the schema update through XMLA and then triggers the data refresh. Adding a new column is always an explicit schema update of the tabular model. Update Tabular Model when the underlying Relational Database schema changes | Microsoft Learn
Kind regrds,
Martin