RIDW contains three physical areas, all defined and residing within one database:
Configuration area: This area is identified in the physical model by the CONFIG schema. It contains and maintains warehouse configuration information and serves as a reference point for auditing and tracking the extract, transform, and load (ETL) activity.
Operational data store: This area is identified in the physical model by the RIODS schema. It is the area where data from various operational sources is integrated. The process by which data arrives and is stored in this area is known as the ETL process. Data residing in this area is manipulated and cleansed to meet a certain quality level. The purpose of the area is to maintain pertinent information needed in analytics. It is not intended to be a copy of the operational source. The area is implemented as a relational model.
Data mart: This area is identified in the physical model by the RIDW schema. Data stored in this area is derived from the operational data store. It consists of fact and dimension tables arranged in star schema form (fact tables surrounded by dimension tables). A fact table is a table with two types of columns, a measure and a dimension. A measure column represents an aggregation such as a sum or count. Dimension columns are foreign keys pointing to dimension tables, which describe the measure.
RIDW has an open schema. Any business intelligence tool can access RIDW to generate business reports. They can also use Rational Insight Data Service if the reporting server hosting the business tools supports a REST architecture.