The data warehouse project plan is a bit of a misnomer, because the data warehouse is really more a service than a project. We are constantly reviewing change requests and building out new "cubes" (on which see Using the Data Warehouse).
If you have change requests or corrections you'd like made or if you'd like a new feature or cube, put this information into our queue by calling the Helpdesk (x5999) and asking them to fill out a ticket of type Software > Data Warehouse > New Feature Request, New Cube Request, or (in the case of a problem) Data Quality Issue.
Despite the ongoing "service" nature of the data warehouse, it would be correct to say that a considerable amount of up-front work, and an initial project plan, was required to get started. This did require considerable research and planning, and it consisted mainly in:
- Identifying and training a database administrator
- Selecting technology platforms to use, and training the data warehouse technical project team in those platforms
- Deploying key supporting infrastructure, such as database servers and data-migration tools
- Identifying an oversight committee, who could assist with prioritization and publicity
- Forming a data management committee
- Creating a data management policy
- Conducting business requirements analysis
- Selecting a first subject area, i.e., a first area of the data warehouse to build
- Selecting a desktop deployment methodology and an initial set of desktop software to give to users
- Initial production of a data dictionary
These steps were completed as of March 1, 2009, although many of them will be revisited and adjusted periodically throughout the lifetime of the data warehouse service. This is particularly true, e.g., for technology selection, training, requirements analysis, and the selection of desktop deployment methodologies.