Changing source for reports SQL Server to Databricks

[GavC]

Hi,

I have a report that is pulling from a SQL Server data source.
In Azure Databricks, I have shaped the tables to be exactly the same as those from SQL Server.

I now want to switch the data source for the reports from SQL Server to Azure Databricks, and then with little or no additional work, for the report to seamlessly work. The only thing that has altered is the source of the data, so I see no reason for this not to work.
It would be the same as just changing an excel sheet for example.

However, I have tried and failed...
It seems like I have to rewrite the report rather than just swap over the source, I was hoping that someone has atried this also and has found a solution that I am missing.

Many thanks

[v-agajavelly]

Hi @GavC ,

Just checking in once more were you able to review the details shared regarding swapping from SQL Server to Databricks in Power BI? If you still have questions or need further clarification, feel free to share. Happy to assist further.

Regards,
Akhil.

[v-agajavelly · Accepted solution]

Hi @GavC ,

Thanks for sharing your scenario, this is a common use case when transitioning from SQL Server to Databricks.

Since you’ve maintained the same table structures, it's logical to expect a smooth switch. However, Power BI doesn’t allow you to directly swap a data source like SQL Server to Databricks without some rework, especially since the data connectors and query folding behaviors differ.

Here’s what you can do to retain your existing report and minimize rework:

1. Open Power BI Desktop.
2. Go to Transform Data > Advanced Editor in Power Query.
3. Copy the entire query for each table (from SQL Server).
4. Connect to your Databricks source and load the same tables.
5. In Power Query, for each Databricks-loaded table.
• Replace the default query with the one you copied from SQL Server.
• Just update the Source step to point to Databricks (adjust the connector and database path only).
6. Ensure all columns/data types remain consistent.

This way, you’re reusing the transformations but simply changing the backend. If names and schemas are identical, Power BI visuals should work without breaking.

You can also right-click your original SQL Server queries and use “Reference” instead of “Duplicate,” so updates can cascade if needed.

Let me know if you’d like a sample structure or a mock walk-through! You don’t need to rebuild your report from scratch just align the source steps.

If this response helps, consider marking it as “Accept as solution” and giving a “kudos” to assist other community members.

Thank's,
Akhil.

[v-agajavelly]

Hi @GavC ,

Just checking in were you able to try the steps I shared regarding swapping from SQL Server to Databricks in Power BI? I know it can feel a bit tedious at first, especially when connectors behave differently, but reusing your existing queries and just updating the source path usually keeps things running smoothly without breaking your visuals. If you've run into any hiccups or need a quick walk-through on adjusting the source step, feel free to ping me happy to help.

Regards,
Akhil.

[Cookistador]

How did you try to achieve this ?

I never used databrick to build a report with Power BI, but it should work

My suggestion would be to export databricks, making the same transformation that you did in your SQL Server and then compare the difference betwween the both in Power Query with advanced editor