Data Warehousing - Starry Scheme vs Flat Table

I am trying to create a data warehouse for one store of commonly required data, ranging from financial systems, project planning systems and many scientific systems. That is a lot of different data marts.

I read Data Warehouse and popular methods such as Star Schemes and Kimball, etc., but one question I cannot find an answer to is:

Why is it better to design your DW Data Mart as a star schema rather than a single flat table?

Surely there is no match between facts and attributes / dimensions faster and easier than having many small joins to all dimension tables? Disk space is not a problem, we will just throw more disks into the database, if necessary. Is the star layout a bit dated these days, or is it still an architect of architecture?

+4
source share
3 answers

Your question is very good: the Kimball mantra for dimensional modeling is designed to increase productivity and improve usability.

But I do not think this is outdated or dogma - this is a reasonable practical situation for many situations and platforms.

, , , , , ..

3NF ( ) - , OLTP-, - . .

" ", , , - . . , , .

, , , "wordier", , , .

, , ? - , , ? , , - . .

+5

, , , .

, - , - . , , ? , , , ?

, .

+1

.

, (, ). , , SCD .

0

Source: https://habr.com/ru/post/1679124/


All Articles