Forum Discussion
Microsoft Employeerelationships in ssas cube
Advocate IAs a best practice I wouldn't recommend importing the data from the cube as you would have significant performance issues when you bring in all the data (lowest level of granularity) for larger cubes. The import cube data feature is relevant only if you are working on a smaller subset of the data and it generates one single mdx query based on your selection criteria for attributes and measures in which case there will be just one table in Power BI holding the dataset of the mdx query generated.
The best option would be to use direct query feature available for Analysis Services (available for both Multi-Dimensional and Tabular). This is very similar to connecting the cubes using Excel. You don't need to worry about importing dimensions, measures etc. and creating relationships in Power BI. You just connect to the cubes and start building the charts.
Hope this makes sense.
I have the same issue here. I have to import data from SSAS cube since some new measure and calculation needed, which cannot be done in live connection. There is only one object (dimension+measures) in relationship view. When I need to import addtional columns that missed before, it displays in separate object and doesn't work to build relationship between them. Is that means I have to create from scratch every time or make sure ALL required columns included one time?
- karthik
I'd suggest you to add all new measures in the cube if you are using the direct query connection. Import works well only when you want to extract a small sub-set of the data from the cube.