Forum Discussion

rho20's avatar
rho20
New Member
7 months ago
Solved

Excel Get Data from Fabric Connection, Getting Prompted Multiple Auths for Single Connection

I have data stored in a Fabric Lakehouse that I am connecting to through Excel 'Get Data' -> 'From Database' -> 'From Analysis Service', then connecting via the PowerBI Api Connection Link. After I m...
data hub
workspaces
  • V-yubandi-msft's avatar
    V-yubandi-msft
    7 months ago

    Hi rho20 ,

    Thank you for the clarification. You are correct Excel behaves the same way with a Fabric semantic model.

    Each PivotTable creates its own Analysis Services (OLAP) session, and authentication is handled per session rather than per workbook or connection. As a result, Excel prompts for sign in for each PivotTable when reopening the workbook or when the connection or schema changes.

     

    Currently, Excel does not allow a single authentication to be shared across multiple PivotTables. There is no setting or workaround to prevent multiple sign in prompts. The practical solutions are to reduce the number of PivotTables, reuse existing ones, or manage frequent schema changes in Power BI Desktop.

     

    Regards,

    Yugandhar.

  • deborshi_nag's avatar
    deborshi_nag
    7 months ago

    Hello rho20 

     

    When Excel connects to Microsoft Fabric (or Power BI) through:

    Get Data > From Database > From Analysis Services
    (using the XMLA endpoint / Power BI API link)

    Excel treats each PivotTable as its own OLAP session.

     

    Because of this architectural design:

    • Each PivotTable requests its own authentication token.
    • Reloading the workbook causes multiple prompts — one per session.
    • Changing the connection string (e.g., pointing to a new schema) invalidates all sessions, so each one has to authenticate again.

    This behaviour is not specific to Fabric — it also happens with:

    • Power BI Datasets
    • Azure Analysis Services
    • SQL Server Analysis Services

    It’s simply how Excel’s OLAP client works today.

     

    Workaround:

     

    If you load the data into the Excel Data Model (Power Pivot) instead of connecting via Analytics Services:

    • You get a single workbook-level model session
    • All PivotTables connect to the same internal model
    • You authenticate once
    • You can change schemas by editing the Power Query connection instead of dozens of OLAP sessions

    Workaround:

    1. Data > Get Data
    2. Choose the SQL endpoint of your Fabric Lakehouse or Warehouse
    3. Load to: Only Create Connection + Add to Data Model
    4. Build PivotTables from the Data Model

    This eliminates repeated authentication prompts!

     

    Hope this helps - please appreciate by leaving a Kudos or accepting as a Solution! 

MJParikh's avatar
MJParikh
Super User
7 months ago

There isn't a secret setting that you're missing. This is how Excel works right now when you connect to Fabric through the Power BI XMLA endpoint.

Excel sees each PivotTable as its own Analysis Services (OLAP) session, even if they all use the same "Existing Connection." Authentication is limited to the session, not the workbook. If you change the connection or reload the workbook, each PivotTable will ask for its own Entra ID token. This will cause multiple sign-in prompts.
This doesn't just apply to Fabric. Azure Analysis Services and Power BI datasets act the same way.


No. Excel does not share authentication tokens between different OLAP sessions in the same workbook. There is no way to force a single sign-in per connection in Excel or Fabric.


You can review and try below alternate.

  • Instead of making new PivotTables, copy the ones you already have. This can cause extra prompts to show up later, when the workbook is opened again.
  • Cut down on the number of PivotTables. Instead of separate PivotTables, use slicers, measures, and calculation logic.
  • Don't change schemas and keep the connection endpoint stable. Let the model deal with changes by using filters or measures.
  • Instead of connecting directly to the Lakehouse, use a Fabric semantic model as the source. Metadata stability makes things run more smoothly.
  • Power BI Desktop is the better tool if you switch between schemas or datasets often. It uses the same authentication for all visuals.
  • What doesn't help: "Use this connection for multiple PivotTables" and "Existing Connections."
  • Turning off background refresh
  • Changing the connection string by hand


One OLAP session is equal to one PivotTable, and each session needs to be authenticated. At this time, Excel does not allow for shared authentication across PivotTables. The only real way to fix this in Excel is to either lower the number of PivotTables or move the analysis to a semantic model or Power BI Desktop.