The Origin and History of Data Warehouses

How did data warehouses come about? Why were they invented? The simple answer to this question is because existing databases were being subjected to conflicting requirements. These conflicting requirements are based on operational use versus decision support use.

Operational use in Online Transaction Processing (OLTP) databases is access to the most recent data from a database on a day-to-day basis, servicing end user and data change applications. Operational use requires a breakdown of database access by functional applications, such as filling out order forms or booking airline tickets. Operational data is database activity based on the functions of a company. Generally, in an internal company, environment applications might be divided up based on different departments.

Decision support use, on the other hand, requires not only a more global rather than operationally precise picture of data, but also a division of the database based on subject matter. So as opposed to filling out order forms or booking airline tickets interactively, a decision support user would need to know what was ordered between two dates (all orders made between those dates), or where and how airline tickets were booked, say for a period of an entire year.

The result is a complete disparity between the requirements of operational applications versus decision support functions. Whenever you check out an item in a supermarket and...