I have 15 Power Bi Reports in 17 workspaces and all the reports use the same dimension tables it is only the fact tables that change but every time I bring in a dimension table I do transformations on those queries. My question is what is the best practice in having dimension tables in an Oracle DB for Power BI reporting. The company I work for already have the tables separated into fact and dim tables in the DB however I need the tables to be transformed into how I use those tables by removing columns or doing a group by. Now I have heard people use views in there Oracle DB is that a good solution for me and can someone recommend whether I should ask my manager for my own dedicated space in the Oracle DB server as there are already views created in our live server. Is there an alternative I have heard of data warehouses are they built on top of a companies DB by creating staging tables. Any help or advice is highly appreciated.

OPTIONAL QUESTION: When bringing in dim tables how do you filter them or do you filter them like say a customers table has 200,000 rows do you bring all those rows in? and if you filter it how do you know your not filtering a row that is required in the fact table?

CodePudding user response：

If you can't get access to the DB, then you should take a look at data flows to create your conformed dimension and let your reports import from there.