Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


However the groups that work on the data warehouse are coordinated, one requirement remains clear: The group that manages the lower-level detail must form a proper foundation of data for those groups that will be summarizing the data and creating a new level. This need is depicted in Figure 6.30.


The coordination among the groups can be as simple as an agreement on a data model that satisfies the needs of all parties. Or the agreement can be much more elaborate, if circumstances warrant.

OLAP


lightly


summarized


4/


^detailed data»^


Figure 6.29 Interconnectivity between the different levels of the data warehouse is an important issue.



summary


О



ooo Soo° О о о о °о о°оо О^О 0_ О о



detail


Figure 6.30 The detailed level forms the foundation for the summary level of data.


The coordination of the development efforts is another matter. There must be a time-sequenced arrangement among the different development groups so that no one group arrives at a point of needing data that has not yet been gathered at a lower level.

Multiple Groups Building the Current Level of Detail


An infrequent set of circumstances occurs when multiple development groups attempt to build the current level of detail in a data warehouse in a nondistributed manner. Figure 6.31 illustrates this phenomenon.


As long as the groups that are developing the current level of detail are developing mutually exclusive sets of data, there is little difficulty. In this case, if the development groups are working from a common data model and the different groups’ platforms are compatible, no problems should ensue. Unfortunately, mutually exclusive data sets are the exception rather than the rule. It is much more common for the multiple development groups to be designing and populating some or all of the same data.

Скачать в pdf «Building the Data Warehouse»