The data warehouse lifecycle system involves the creation of an efficient system for storing and retrieving computer data. Time has shown that simply dumping data into vast computer storage does not work. Instead, it is best to create a storage system, test it, and alter it as needed to fit the ever-changing data needs of the company.

Data warehouse lifecycle

Design:

 the development, from both available data inventories and DSS analyst requirements and analytic needs, of robust star-schema-based dimensional data models. Practically speaking, we believe that the best data warehousing practitioners today are working simultaneously from (a) the available data inventory in the organization and (b) the often incomplete expression of needs on the part of multiple heterogeneous analytic communities, making data warehouse and data mart design a complex and intricate process more akin to architecture and diplomacy than to traditional IT systems design. Key activities in this phase typically include end-user interview cycles, source system cataloging, definition of key performance indicators and other critical business metrics, mapping of decision-making processes underlying information needs, and logical and physical schema design tasks, which feed the prototyping phase of the life-cycle model quite directly.

Prototype:

the deployment, for a select group of opinion-makers and leading practitioners in the end-user analytic communities, of a populated, working model of a data warehouse or data mart design, suitable for actual use. The purpose of prototyping shifts, as the design team moves back and forth between design and prototype. In early prototypes, the primary objective is to constrain and in some cases re-frame end-user requirements by showing opinion-leaders and heavyweight analysts in the end-user communities precisely what they had asked for in the previous iteration. As the gap between stated needs and actual needs closes over the course of 2 or more design-prototype iterations, the purpose of the prototype shifts toward diplomacy – gaining commitment to the project at hand from opinion leaders in the end-user communities to the design, and soliciting their assistance in gaining similar commitment.

Deploy:

the formalization of a user-approved prototype for actual production use, including the development of documentation, training, O&M processes and the host of activities traditionally associated with enterprise IT system deployment. Deployment typically involves at least two separate deployments – the deployment of a prototype into a production-test environment, and the deployment of a stress-tested, performance-tested production configuration into an actual production environment. It is at this phase that the single most often- neglected component of a successful production warehouse – documentation – stalls both the transition to operations and management personnel (who cannot manage what they cannot understand) and to end-user organizations (who after all have to be taught at some level about the data and metadata the warehouse or mart contains, prior to roll-out).

Operation:

the day-to-day maintenance of the data warehouse or mart, the data delivery services and client tools that provide DSS analysts with their access to warehouse and mart data, and the management of ongoing extraction, transformation and loading processes (from traditional drop-and-reload bulk updates to cutting edge near-real-time trickle-charging processes) that keep the warehouse or mart current with respect to the authoritative transaction source systems.

Enhancement:

the modification of

(a) physical technological components,

(b) operations and management processes (including ETL regimes and scheduling) and

(c) logical schema designs in response to changing business requirements.

In cases where external business conditions change discontinuously, or organizations themselves undergo discontinuous changes (as in the case of asset sales, mergers and acquisitions), enhancement moves seamlessly back into fundamental design.