Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»

development group B

development group A


Figure 6.27 Different development groups are developing different parts of the data warehouse environment at different levels of the architecture.

development group C

The primary concern of the data architect is to coordinate the efforts of the different development teams, both in terms of the specification of content and structure and in terms of the timing of development. For example, if Group A is significantly ahead of Group B or C, there will be a problem: When Group A is ready to populate its databases at the summary level, there may be no detailed data to work with.

One of the interesting aspects of different groups building different levels of summarization of the same data warehouse is that it is the group that is building the current level of detail that uses the data warehouse data model. Figure 6.28 illustrates this relationship.

The data model for the data warehouse directly reflects the design and development effort by the group doing current-level detailed analysis and design. Of course, indirectly the data warehouse data model reflects the needs of all groups. But because other groups are summarizing from the data found at the current level of detail, they have their own interpretation of what is needed. In most cases, the groups working on the higher levels of summarization have their own data models that reflect their own specialized needs.

One of the issues of managing multiple groups building different levels of summarization is the technological platforms on which the data warehouse levels are built. Normally, different groups choose different technological platforms. In fact, for the different development groups to choose the same platform would be very unusual. There are several reasons for this, though the primary one is cost. The detailed level of data requires an industrial-strength platform because of the large volume of data that will have to be handled. The different

Скачать в pdf «Building the Data Warehouse»