Hi,
I'm working on a fairly complicated SSAS project at the moment:
- SSAS provides information to 6 teams, which I have modelled as 6 separate cubes. Each team should not be able to access the other team's cubes.
- Each cube has 4-5 measure groups, with no cross-over with other teams.
- Each cube has approx. 8 dimensions, 5 of which are conformed dimensions, common across all cubes, and 3 dependent on the specific measure groups.
- Each cube has a couple of perspectives set up to simplify the presentation.
- The underlying data comes from a single data warehouse database.
- It's possible that another team may shortly want to develop their own cube. The data would be completely separate however they would need to use the same conformed dimensions as above.
I'm at the starting stage, so I have set up one cube along these lines and am about to work on a second. My question is, is this the best structure for the project, in terms of both working with it in VS (the data source view is already
horribly messy, but I don't think it can be split up?) and also avoiding processing the conformed dimensions multiple times. If not, what is the best approach to follow to avoid duplication? Using linked dimensions sounded like the right
approach, until I read about how they were then co

**View Complete Post**

## how ssas generate sql queries when processing dimensions and partitions?