Refining the Database Design › Refinement Options

Refinement Options

CA IDMS/DB provides options for refining the database design to ensure optimal performance in individual transactions. There is no right or wrong method for refining the physical database model. Your organization's requirements will determine the best approach for you.

Options

The following database options can be used to ensure optimal performance in individual business transactions:

• Indexes—Chapter 11, "Determining How an Entity Should Be Stored" showed you how to include indexes in the database design to provide data clustering. At this point in the design process, you have the option to include additional indexes to provide generic search capabilities as well as alternate access keys.
• Collapsing relationships—A one-to-many relationship can be expressed within a single entity by making the many portion of the relationship a repeating data element. A one-to-many relationship expressed in this way can enhance processing performance by reducing DBMS overhead associated with processing multiple entity occurrences.
• Introducing redundancy—By maintaining certain data redundantly, you can sometimes enhance processing efficiency in selected applications.

Each of these options is described in detail below following a discussion of how to estimate I/Os for transactions.