Forum Discussion

sergs's avatar
sergs
Advocate I
3 years ago

Embedding a report with DirectLake dataset.

Hi

 

We're investigating options for migrating our reports to Microsoft Fabric. Everything (data ingest, report etc) works fine until we attempt to embed this report. PowerBiEmbed client GenerateTokenAsync fails with "Embedding a DirectLake report is not supported"

So the question is - will it be supported when Fabric is GA or is there another way of embedding reports based on DirectLake dataset?

 

Thank you!

Need Help

6 Replies

  • vladm-datanaut's avatar
    vladm-datanaut
    Frequent Visitor
    3 years ago

    Mind you. If you use the Power BI Embedded Playground you can embed a report with a DirectLake connection. So it must be available soon - hopefully.

  • jasonn's avatar
    jasonn
    Frequent Visitor
    2 years ago

    Has anyone heard anything definite about this?
    I've tried several different combinations but still get the "Embedding a DirectLake report is not supported"

     

    In respect to embedding (app owns data) I've found the documentation lacking clarity on what is actually supported.

     

    I've gone over this link several times https://learn.microsoft.com/en-us/power-bi/enterprise/directlake-overview

    For example - what does this even mean? 

    • Embedded scenarios that rely on embedded entities are not yet supported.

  • jasonn's avatar
    jasonn
    Frequent Visitor
    2 years ago

    I ran into this - what a pain.

    Here's how I ended up solving this (workaround)

     

    1. Data is in a fabric Lakehouse
    2. Use PowerBI Desktop to import the data from the lakehouse using the SQL endpoint. (Azure SQL Datasource)
    3. Publish the Semantic Model to your workspace.
    4. Build reports off of that Semantic Model - you can set it to refresh from the Lakehouse as needed.

    It seems like a pain, but the most reliable way to get embedded reports to render. There was a surprising upside to this approach. Importing the Lakehouse into the Desktop Power BI application and then publishing made managing schema drift wayyyyyy easier. I found that when I had to handle schema drift in the lakehouse itself deleting of the table before republishing just destroyed any schemas I had mapped out. Managing the schema in the desktop app allowed me to delete a table in the lakehouse then refresh the file - and nothing in the schema would get broken. New fields would simply appear (reconcile or be deleted) as you would expect them to.