Fix Mixed Model Paginated Report

I have a Power BI semantic model that used to pull exclusively from Dataverse and the linked paginated report rendered flawlessly. After adding several Azure Databricks columns to the same model, every RDL built on it now opens blank. Dataverse data still flows through; the break happens the moment the report tries to touch anything that originates from Databricks.

My priority is simple: display all data correctly—including those new Databricks fields—while keeping the existing Dataverse logic intact. The need is urgent, so I’d like to resolve this ASAP.

You’ll probably need to dive into composite-model limitations, DirectQuery vs. Import settings, data source credentials, and how the XMLA endpoint is exposing the model to the paginated layer. Whether it means refactoring relationships, creating a dedicated semantic layer for Databricks, or rewriting queries in DAX/M language, I’m open to the cleanest path that gets the report fully operational.

Deliverables
• Updated semantic model (PBIP/PBIX) that supports both Dataverse and Azure Databricks columns
• Fixed paginated report (RDL) that renders every field without blanks or errors
• Brief walkthrough or notes so I understand the changes and can maintain them moving forward

Acceptance criteria
– Report loads instantly with no blank visuals
– Databricks columns display accurate values verified against the source notebook/table
– No regression on existing Dataverse measures or visuals

If you know Power BI Composite Models, Paginated Reports, Dataverse, and Azure Databricks inside out, I’m ready to start right away.