Star schema or SQL script

[odzelashvili1]

Hello everybody ^^

I have one simple question, which is better a star schema, or one table which is made by SQL script?

In my database I have fact table, and some dimension tables, for power BI I'm importing data by SQL Script, so I have only one table and only needed columns, is this correct? or it is better to import fact table and also dimension tables and than make relationships in my model?

Thanks a lot 🙂

[amitchandak]

@odzelashvili1 , Star Schema is always better. If you have say 5 tables, bring them into one is not a good idea.

Having 3 or 5 in a star schema is better.

Also refer

https://www.sqlbi.com/articles/the-importance-of-star-schemas-in-power-bi/

[odzelashvili1]

Thank u for your reply 🙂

So, to be more clear:

I have 1 fact table, with 20 column, where is only IDs and I need only 10 columns for my report. Also in my DB there is 19 dimension tables.

I have two ways: 1) Write SQL Script, where I get 10 column with values, not IDs and import 1 table in my data model, or 2) Import 10 tables, 1 Fact table (only needed columns), and 9 (also only needed columns) dimension tables and make relationships between them.

As u told, its better to use second way, am i right?

thanks again 🙂

[Sabo256w · Accepted solution]

Hi 

@amitchandak is absolutely right, it is best to use the Dimension tables and Fact tables then relate them in PowerBi instead of 1 Flat table. You will run into issues later with your data model if it's just one table.

example of star schema: