on 10-26-2009 3:56 AM
Can someone explain the pros and cons of developing multiple universes and then linking them up vs. developing a single monolithic universe?
We have 4 major modules within our enterprise software, each one can be individually considered a product on its own. We have some large scale customers who are interested in being able to run ad-hoc queries which might span data from these multiple modules for analysis.
We do have some (20%) common dimensions between these modules.
BO does not provide us the capability that would allow multiple people in multiple teams (each working on a different module) to work on the Universe at the same time.
Given the above two challenges, we are contemplating if it would make sense to have people from different modules work on their Universe and either link them at the end of each iteration or if it would make sense to have the overhead of merging the Universe everytime a change has been made by another team.
What are your thoughts? Suggestions? Experiences .
From what I read it seems like it makes sense to go with a concept of a Kernel Universe and Derived Universe given these problems described above.
Any issues with this approach - other than having to redefine custom hierarchies, contexts, universe params in each of the derived Universes...
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
90 | |
10 | |
10 | |
10 | |
7 | |
7 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.