I have a fact table with 100 millions+ rows and with 4 dimensions and everything works fine from query performance point of view and from processing cube perspective (~it takes around 5-6 minutes to process the cube).
1/ Now I have to add a new dimension with 6 millions rows. The cube processing time will explode?
2/ Is that a good practice? Do I have other alternatives to model this dimension?
3/ What about if my fact table raises to 300 millions rows and my dimension to 15 millions rows?
Any thought is welcome. Thanks in advance.