For larger datasets with many measures, relationships, and tables, a dedicated metadata report can be constructed using Power BI. In this approach, the Analysis Services data connector is used to access the DMVs of the Power BI dataset and this data is transformed via M queries. Finally, a set of report pages are created to visualize the primary objects of the model and support common ad hoc questions, such as which relationships use bidirectional cross-filtering?

Implementing the DMV-based Power BI report consists of the following four steps:

1. Obtain the server and database parameter values of the Power BI dataset
2. Query the DMVs of the Power BI dataset from a separate Power BI Desktop file
3. Integrate and enhance the DMV data to support the visualization layer
4. Develop the report pages