How Is Data Represented in the Data Warehouse Analytics System?
The Data Warehouse Analytics System transfers data from a Commerce Server 2009 site to a separate database, so that a business manager can analyze site data without affecting the performance of the Web site. The data is represented in multiple ways.
Logical representation
The Data Warehouse Analytics System represents data that it reads from the site databases in a logical form. The logical representation includes the following entities:
Classes: A class represents a set of logically related information.
Data members: A data member represents an attribute of a class.
Keys: A key is a unique identifier for a class.
Relations: A relation represents a connection between two classes.
Physical representation
Because the Data Warehouse Analytics System stores data in a SQL Server database, the physical representation consists of database tables, columns, and keys. A table often corresponds to a class. A column corresponds to a data member.
The Data Warehouse Analytics System uses metadata that is stored in the Data Warehouse to define how logical data is transformed into its physical representation. You can customize how data is stored in the database by modifying the metadata. You can define the transformation implicitly by using Commerce Server Manager. You can also modify the transformation explicitly by updating the SQL Server database tables that contain the metadata.
Analysis Services representation
To simplify data analysis, the Data Warehouse Analytics System converts some data from SQL Server tables to online analytical processing (OLAP) cubes in Analysis Server. The OLAP representation of data consists of cubes, measures, and dimensions.