Wanted to get some help in understanding datasets and if my implementation is "best practice" or inefficient. I fear it may be inefficient.
We have a SQL Server Analysis Services model based on a data warehouse and a few additional datasets. Its one main SSAS model that we primarily use. Whenever we create a new Power BI report, we connect to this SSAS model via live connection. When we post the report to the Service, it creates a new dataset. Thus we have several datasets which technically refer back to the one SSAS model.
I feel like this is potentially redundant data in the datasets. Our main SSAS model is kind of an 'all encompassing model' which has many data tables in it and not every report uses every data table, maybe 3-5 tables out of 20 depending on the purpose of each report. Some reports overlap a bit.
How does this sound to you guys? Is there a more efficient way? Not sure if it's possible but one thought of mine is to create 1 dataset made up of our main SSAS model and then each PBI report we create connects to this dataset/model in the Power BI Service; but not sure if that's ideal or possible.
Would love feedback and thank you!
Can someone please tell me if the sub-report functionality for Power Bi would be available anytime soon? Or is there a work around to be able to do what the sub report would do in Power BI Report builder? Thank you!