Storage areas are the largest physical unit of a database. Storage areas consist of one or more extents that are either operating system files, or some other operating system level device that is addressed randomly. A storage area is a distinct address space, and any physical address stored inside the area is generally stored relative to the beginning of the storage area.

Storage areas give you physical control over the location of specific database objects. You can place each database object in its own storage area or place many database objects in a single storage area. Storage areas can contain database objects of one type or of many types. For example, to achieve load balancing, you can place a particularly active table in a separate storage area, then place the most active index for that table in its own storage area. Then, in a third storage area, place all the remaining tables and indexes. You cannot split a table or index across storage areas.

To take advantage of performance features available in Type II storage areas, you can designate default storage areas by object type for creating tables, indexes, and LOBs. These default areas must be Type II areas, and you can change default areas while the database is running. For more information, see PROUTIL SETOBJECTDEFAULTAREA qualifier.

You may also improve performance by moving tables and indexes to an application data storage area on a faster disk, while the database remains online. For a description of how to move tables and indexes while the database remains online, see Maintain Database Structure.