Forum Discussion
Multiple tables from same DB vs Query
Hi, I saw across forums and tutorials several users loading different tables from a database and then creating relations on Power BI to work.
With this you have to use formulas with the "RELATED" operation and have a specific treatment when joining tables.
Any advantage in loading simple tables (when the source is a single DB) and creating the realtions Vs loading a query with all the fields you need to work? either via direct query or import.
The query approach seems simple when working with the data.
Thank you!
It's actually a pretty common to load multiple tables from the same DB and create the relationships in Power BI data model. If you have a straight-up database, then all of the tables are disconnected technically. You may have keys defined that "tie" the tables together but those relationships really only exist in the SQL statements that either insert, retrieve and delete data from the database.
Now, if you have a cube, either MDX or Tabular, then those relationships exist in the cube as actual defined relationships between tables and you could connect Power BI to the cube directly and that's probably a better way to go.
When people build relationships in Power BI, they are generally taking the disconnected tables that are imported, either from the same database or multiple data sources and are creating that tabular cube within the Power BI data model.
3 Replies
That's a very broad question with no single "one-size-fits-all" answer.
Please see this post regarding How to Get Your Question Answered Quickly: https://community.powerbi.com/t5/Community-Blog/How-to-Get-Your-Question-Answered-Quickly/ba-p/38490
Thank you Greg_Deckler that's true, I was for the broader answer, I think that I could wrote the question better.
But I think I will understand if you could tell me an example (very simple) when load individual tables from same database is better.
I'm thinking about it and I don't have ideas of why would yo load individual tables from same database and make the joins in PowerBI.
Maybe is better in the case when I have one to many to avoid the "distinct" in every formula/measure?
thank you!
It's actually a pretty common to load multiple tables from the same DB and create the relationships in Power BI data model. If you have a straight-up database, then all of the tables are disconnected technically. You may have keys defined that "tie" the tables together but those relationships really only exist in the SQL statements that either insert, retrieve and delete data from the database.
Now, if you have a cube, either MDX or Tabular, then those relationships exist in the cube as actual defined relationships between tables and you could connect Power BI to the cube directly and that's probably a better way to go.
When people build relationships in Power BI, they are generally taking the disconnected tables that are imported, either from the same database or multiple data sources and are creating that tabular cube within the Power BI data model.