Tuesday, April 2, 2013

MultiDimensional Vs Tabular...

With the release of SQL Server 2012, SSAS has two modes one with MultiDimensional and the other with tabular. Depending on the business needs of the organisation one of the modes can be chosen or the two modes can exist on different servers. Well decisions would need to be made on what mode suits the organisation, one of the ways to approach this problem is that one could do a technological evaluation of the two modes and decide which one to choose. A better way to handle this would be is to use a decision matrix which would combine a good mix of business and technology needs. I would list them here, thanks to Devin Knight(From Pragmaticworks). Here are the following points.
1. The Time required to Develop the solution.
2. The type of Data Sources (What are the data sources is it going to be a relation database  or flat files or web or xml).
3. The Ease of Query Language ( When using Multidimensional one needs to know MDX, when one uses tabular one needs to know DAX, this is kind of similar to excel based functions. Depending on the resources, what would be the learning curve, does the budget allow for staffing new resource who are skilled with either of the languages.)
4. Security (What are the security considerations for the solution to be developed, is security a top priority or not).
5. Presentation Options (What type of reports are required and how are they to be presented. What are the delivery modes for the reports for the end users)
6. Model Compatibility (How does either of the modes adhere to the data model available at hand, is there a good star schema database/dimensional model available or is the modelling of data very varied).
7. Scalability ( How scalable the solution is going to be, what is the data growth going be, by how much the source database going to grow, how much value the data is going to provide).

I find these 7 points cover a good range of questions that can be answered, it tries to incorporate a mix of business and technical factors. Hope the readers find these very helpful.

No comments:

Post a Comment