Online Transaction Processing

Databases support OLTP systems. Online transaction processing systems (OLTP) are traditional automated data processing systems where the focus is on the recording of business transactions and producing information from that data. This section describes the architecture and methodology used to produce traditional, stove-pipe application databases and introduces an architecture and a methodology to integrate those stove-pipes into the enterprise data architecture. The goal is to integrate the data resources of the enterprise.

Understanding funcionally dependent data vs. functionally independent data is a key concept of the enterprise data architecture.

Functionally independent data exist in the world. They do not rely on a business or business function to exist -- data that describe Mary Smith, for example.

Functionally dependant data are created by or used by a single business function or activity. These data only have meaning in the context of a business function or activity -- data that describe Mary Smith's checking account, for example.

As the bank implements new banking systems, Mary Smith's data (all that personal account-holder information) exists in the enterprise database, so when the developers build the new mortgage system, they only need to focus on the systematic elements of the problem. The account-holder stuff exists and the hooks (relations) exist.

We've reduced redundancy and improved the quality of the data. When Mary makes a change, any change to her data, all the applications see the single change.

Enterprise and application databases are constructed from the nine components and eight patterns contained in this website. A coupled enterprise and application database are referred to as the enterprise/application database pair.