Do we create relational and dimensional model for one project

[inu]

Hi
I dont have idea about real time application / reports. what is done actually. I m little bit confused, that, if we are creating relational model in project., and authoring the reports in report studio. Do we create even dimensional model in the same project in real time application. (OR) generally we are not assigned the job for creating dimensional model, if we are creating relational model.

Please justify it...THanks buddy

Regards
Inam

[MFGF]

Hi
I dont have idea about real time application / reports. what is done actually. I m little bit confused, that, if we are creating relational model in project., and authoring the reports in report studio. Do we create even dimensional model in the same project in real time application. (OR) generally we are not assigned the job for creating dimensional model, if we are creating relational model.

Please justify it...THanks buddy

Regards
Inam

What you build is as a direct result of the business requirements. If a relational model is all that is required to fix the business pains of the client, than that's all you would build. If, however, the client also needs the ability to analyze data and drill down/up etc, then a dimensional model would typically answer these requirements. Whether you build DMR structures in your FM model or whether you create some kind of OLAP cube (PowerCube or Dynamic Cube or TM1 cube or other OLAP) is really driven by the requirements and the data. Each OLAP technology has its own sweet spot. PowerCubes are good for small data volumes where the data is often from disparate sources and up-to-the-minute data isn't important. TM1 cubes are good for medium data volumes and situations where writeback may be an advantage (ie Planning/Budgeting/Forecasting is on the agenda). Dynamic Cubes are good for large data volumes where the data is held as star or snowflake schemas in a dta warehouse. DMR is good for smaller data volumes where latest data values are important.

There is no "one size fits all" answer to your question. It all depends on the business requirements and the data.

Cheers!

MF.