Hello,
Nowadays, I see many B.I tools on the market advertising that their tools are plug'n play, connect with relational databases, excels, and "instantly generate B.I systems."
This gave me some doubts, as I already had a session in the B.I area and I remember that we always used a mix of Snowflake Schema and Star Schema, we did not use the relational. But I was never really told why.
What comes to mind as to why it would be:
- Can do drill ups and drill downs because of the Snowflake Schema framework
- Can break the value by any part (dimensions), what could only be possible in the relational by means of "tabelões" in the tables of values
- To have a piece of information only up to a certain level, without exaggerated granularity (Prevents the user from trying to execute a query or graph on something giant and very granular). .
Disadvantages:
- Having to transform data from the relational world to the dimensional model
Are there other advantages / disadvantages I'm not remembering? Do you think that for a medium to large company it would be better to adopt a dimensional model instead of the relational model?
Thank you !!