Forum Discussion
Upload from Sharepoint - Dataset
A question concerning upload PBI files from Sharepoint: The underlying database has changed (some columns removed). In Power Bi Desktop, the file has been adapted accordingly by updating the dataset definitions.
However, when saving the PBIX file to the Sharepoint, only the report is updated in app.powerbi.com, not the dataset. So the dataset is still referring to fields that do not longer exist, causing the report refresh to fail.
From Power Bi desktop, all looks fine. How to push or regenerate the dataset in the cloud environment ? As the dataset is generated during publication, I don't have a separate Dataset file. Anyone encountered something similar ?
Hello,
Thanks for thinking along and trying to find a solution.
I finally found the problem, and happy to share here: the problem was in the "Data view": it still contained a reference to the removed columns, allthough in the transformation layer this was corrected. By removing them in the "Data view", an extra step appeared in the transformation layer. Removing this step from the transformation layer finally launched some kind of regeneration of the complete dataset.
2 Replies
Hi Ike369 ,
Data refresh on the Power BI service will fail when the source column or table is renamed or removed. It fails because the Power BI service doesn't also include a schema refresh. To correct this error, a schema refresh needs to happen in Power BI Desktop and the dataset republished to the service.
Best regards,
Community Support Team_Binbin Yu
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.Hello,
Thanks for thinking along and trying to find a solution.
I finally found the problem, and happy to share here: the problem was in the "Data view": it still contained a reference to the removed columns, allthough in the transformation layer this was corrected. By removing them in the "Data view", an extra step appeared in the transformation layer. Removing this step from the transformation layer finally launched some kind of regeneration of the complete dataset.
