A database is a container for related cubes and the objects they share. These objects include data sources, shared dimensions, and database roles. If these objects are to be shared among multiple cubes, the objects and cubes must be within the same database.
Databases are immediately subordinate to the Analysis server in the object hierarchy. Thus, after an Analysis server is installed, databases are the first objects to be created. Databases are created in the Database dialog box. For more information, see Creating Databases.
After databases are created, data sources are usually the next objects that are created. For more information, see Data Sources.
Databases can be archived and restored by using either Analysis Manager or the msmdarch command. For more information, see Archiving and Restoring Databases.
In Analysis Manager, a database is represented by the following icon:
If you are programming with Decision Support Objects (DSO), the class type associated with the database is clsDatabase.
In some cases, particularly when you are programming with PivotTable® Service, the database is referred to as the catalog. A noteworthy example is the connection string to an Analysis server in which a database name is specified in the Initial Catalog property.