What are multidimensional data sources?

A multidimensional data source is optimized for analyzing large amounts of data. Such data sources are sometimes called data warehouses, or online analytical processing (OLAP) data sources.

In a relational data source, data is organized in tables. A table is a data structure with columns and rows. Multidimensional data is data that is organized by one or more dimensions. These multidimensional data structures are often referred to as cubes. Figure 3-2 depicts the tables in a relational data source and the cubes in a multidimensional data source. Oracle9i (and later) databases can include both relational data structures (tables and columns) and multidimensional data structures (cubes). This combination provides fast multidimensional data access while providing summaries of the relational data.

Figure 3-2 Tables and cubes in the Oracle database

Description of the illustration table_cube.gif

For information on relational data sources, see "What is a relational data source?".