OLAP vs Dimensional Relational Tables
We’ve recently experienced full meaning of merging different business intelligence solutions. In one company was installed Business Objects and in other Microsoft Analysis Services.
Problem was to chose one solution and in time switch all reporting only on one platform. There is a great post about differences between OLAP and Dimensional Relational. This is actually difference between Microsoft and Business Objects in this case. Both have advantages and disadvantages and there is no simple answer with which solution should be continued. Business problems are not obvious from pure listing of advantages and disadvantages and that is why we are talking more from users side.
OLAP cubes were preferred by power analysts because they were relatively independent in analysis from IT and analysis was very fast. On a contrary this approach has technological problem because OLAP can not support all required calculations (up to certain) limits and requires creation of more OLAP cubes. More OLAP cubes cause reporting instability and controlling department was in favor for relational tables by Business Objects.
This only one example what can happen in Management Information Systems practice. Also wrong moves can be made due to partial understanding of problems of this highly complex environment. Risks are in not using synergy and in generating additional costs for maintenance.
What can be done?
One approach is to use both solutions and to force synergy where ever it is possible. What is very important if both solutions will continue to work i n parallel, they have to be INTEGRATED.
Second option is political decision, quite often in business intelligence industry, what shall be used. Quick decision to terminate one solution might cause partial collapse of reporting without proper replacement.
There is no easy and simple answer.
Related posts, articles and documents:
loading...
