Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
5 years ago
Solved

Editing a published dataset (SQL)

Hi, 

We currently have a dataset published in the PBI service that sources data from multiple tables in Azure Postgre via SQL. This dataset is used across multiple reports as well. There now exists a requirement to make changes to the underlying SQL for the dataset. What option would be the most efficient way to sort out this?

TIA !

 

  • Hi. You should take the dataset.pbix and edit it with Power Bi Desktop. Change the query and publish replacing the one in Service. That way the dataset will take the change in SQL for all reports connected to that one.

    It's a good idea to keep versioning of the dataset. You can save it in sharepoint in order to keep the changes you make like this one versioned. to roll back things.

    Hope that make sense.

5 Replies

  • Hi. You should take the dataset.pbix and edit it with Power Bi Desktop. Change the query and publish replacing the one in Service. That way the dataset will take the change in SQL for all reports connected to that one.

    It's a good idea to keep versioning of the dataset. You can save it in sharepoint in order to keep the changes you make like this one versioned. to roll back things.

    Hope that make sense.

    • Anonymous's avatar
      Anonymous
      Not applicable

      Thank you! 
      I tried downloading the dataset PBIX file but for some reason (as it has been for the last month or so) the "we couldn't export to .pbix format" error keeps popping up. 








      • ibarrau's avatar
        ibarrau
        Super User

        That's terrible. Too bad. Can't you reach the original file? someone has made that pbix for the first time and they might have saved in a sharepoint or something like that.

        I hope you find it, because Power Bi desktop is the only way to edit sources in a non-premium dataset.

        Regards,

    • Anonymous's avatar
      Anonymous
      Not applicable

      Hi v-luwang-msft

      Switching off scheduled refresh and large dataset format solved the issue!

      Much thanks