ROLAP versus MOLAP Implementations

OWB separates the logical design of dimensional objects from their storage. The logical design, which consists of business rules, first focuses on the structure and the content of the dimensional object. You can then choose to store the dimensional object in a relational ROLAP or an analytic MOLAP implementation.

• ROLAP and relational implementations store the dimensional object in a relational schema in the database.

• A MOLAP implementation stores the dimensional object in analytic workspaces in the database.

OWB enables you to use the same metadata to create and manage both your relational and multidimensional data stores. Separating the logical design from the physical implementation has the advantage of making design of ETL logic transparent. Regardless of the physical storage implementation, the logic for loading dimensions and cubes is identical.

Uses for OLAP

Dimensional objects provide complex analytic power to your data warehouse. After you load data into dimensional objects, you can run complex analytical queries that answer your business questions. These analytic queries include time-series analysis, inter-row calculations, access to aggregated historical and current data, and forecasts. Multidimensional objects are more effective in answering these types of queries quickly.