{Posted from a BI-Training Hyperion instructor}
As an instructor teaching Hyperion Explorer, I often will get a question about someone's database at work. Unfortunatly, I cant answer those questions. Thats called 'consulting'. And I do accept Paypal for those kind of questions. (Just dont tell my boss!)
So, how does one approach an understanding of massive database within a data warehouse? Well, Hyperion Intelligence is a good discovery tool. By that I mean, its easy to peek inside the various tables to figure out whats important.
Take a look at the Detail view in the Query section. It uncovers the first ten rows by default. So if your'e afraid to create a runaway query by Processing a query with everything in that table, this will give you a peek into it. From there, determining the joins is the next challange. Typically, they are the firsts colulms in a table. The join between these colulmns and the colulmns in another table can be done automatically, or you can delete and re-join as required.
The auto join is on by default, but if you see a spider web of joins, its best to turn it off und ther datamodel options. From there, experimentation is in order to achieve what business question you want answered, whether it be outer joins, or equal joins.
Good luck!