The James Martin’s methodology presents a concept FOR (Plan, Activate, Control and Finish), well known by developers of Data Warehouse, in which the processes that make up the project life cycle (which is referred to as process of the road map) are iterative, so some are repeated.
DW phases of the project life cycle are: strategic vision, the company’s engineering evaluation, the flow values, business case of DW, design and review of the architecture, implementation, transition and maintenance.
The phase strategic vision, also called strategic information plan (Sip Strategic Information Plan), is an ongoing process that aligns business and technology strategies of the company within the market. This is a prerequisite for the engineering stages of the company and re-engineering the business process. Some companies have a ready SIP, and it can serve as part of the project where the data warehouse evolves.
The evaluation phase of the company’s engineering (EEA-Enterprise Engineering Assessment) develops a vision in company level of the need to change the organization and its readiness to accept it. A data warehouse is not a solution for everything. If an organization does not have data sources and resources, a data warehouse may not be effective. Before performing a data warehouse project, the organization must decide whether to address operational data issues through the commercial re-engineering, systems development and planning information systems. This assessment is usually a prerequisite for re-engineering the business process or an assessment of the flow values.
The evaluation phase flow values (VSA- Value Stream Assessment) you can solve business problems studying (s) flow (s) of values of a company from a high level for a short period (six to eight weeks), looking for ways to improve the managerial, operational, social and technological performance. The process identifies the predatory value stream – the unique ability allowing you to move faster and produce better than your competitors – and their vulnerable areas of the market share. The knowledge provided by data warehouse technology supports the VSA.
In the development phase of the business case can identify the tasks required to create the business case for data warehousing. At this point, it also sets the team that will justify, will design and implement the data warehouse enters the process. This part is used side personnel, as selected by the consultants or internal staff (interviews, focus sessions, statistical analyzes) to document: (a) a high-level work breakdown structure for the entire project, (b) a cost / benefit, including a return on investment, if possible, (c) the critical factors for success and (d) the typical success impediments.
The analysis of cost and benefit, describes the importance of working with business managers and key business users to identify and assign relative weights to the business benefits of high-level implementation of a data warehouse in order to support the value flows or strategic initiatives. The management and key business users can provide objectives, critical success factors and future development plans for the company, along with a strategy for achieving them. One designed data warehouse must effectively help an organization make strategic decisions that can not be made through transaction systems in operation.
The critical factors for success (CSFs- critical success factors) that need to be established for the project to succeed are well defined business case, sound architectural design, including potential for growth, an amount of data can be managed, an approved and available budget and an internal and external staff dedicated.
In turn, the critical impediments to success (CSIs- critical success inhibitors) that may prevent or derail the project are: a lack of commitment and awareness of executive sponsors, the impact of other strategic projects of information technology, the inability to extracting data from source systems without adversely affect the performance of the transaction system, the uncontrollable degree of organizational change during the project, lack of access to needed source data, personnel assignments part time and lack of use patterns and models for data management.
The review phase and architecture design defines the overall technology and the structure of the process. The review stage and architecture of the project evaluates which parts of the architecture already exist in the organization (a gap analysis). This phase develops the DW logical architecture – the configuration map of the required components data storage locations including a central data storage location of the company, an optional operating data storage location, one or more data marts shopping area Individual (optional) and one or more sites of metadata storage (containing two different types of catalog reference information on the major data).
The implementation phase can be characterized by the fact that all that has been done up to that point paves the way for easy deployment. At this stage is to perform some steps, such as: (a) purchase and installation of the DW components, (b) preparation of test basis, (c) test of the ETL process, (d) tests for update programs ( e) creation of access modules for users, (f) creation and testing of OLAP queries and (g) registration of user acceptance.
The transition to production can be seen as a thermometer measuring the implementation of previous phases was satisfactory. The transition should be quick and quiet. But if something was not well planned, the project team will have to carry out development and production tasks during the next iteration. The main activities of this phase are: (a) move all the components of the development environment system for the production environment, (b) train the operations staff and end users, (c) carry out the operating system documentation and (d ) offer a warehouse that is fully operational and available to users.
Finally, the last phase is the maintenance required since a DW grows and changes with time. Hardware changes may occur, software, additional source systems, additional application processes, changes in the update cycle, changes in activities and version control responsibilities.