Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


But there are other reasons why data warehouse meta data is important. One such reason concerns managing the mapping between the operational environment and the data warehouse environment. Figure 5.10 illustrates this point.


Data undergoes a significant transformation as it passes from the operational environment to the data warehouse environment. Conversion, filtering, summarization, and structural changes all occur. There is a need to keep careful track of the transformation, and the meta data in the data warehouse is the ideal place to do so. The importance of keeping a careful record of the transformation is highlighted by the events that occur when a manager needs to trace data from the data warehouse back to its operational source (the ultimate in the drill-down process!). In this case, the record of the transformation describes exactly how to get from the data warehouse to the operational source of data.


Yet another important reason for the careful management of meta data in the data warehouse environment is shown in Figure 5.11. As mentioned, data in a data warehouse exists for a lengthy timespan—from 5 to 10 years. Over a 5-to-10-year time span it is absolutely normal for a data warehouse to change its structure. Keeping track of the changing structure of data over time is a natural task for the meta data in the data warehouse.


□ □□□□□□


structure


content



operational    data warehouse

Figure 5.10 The mapping between the operational environment and the data warehouse environment is another major reason for the need for meta data; without the mapping, controlling the interface is extremely difficult.

Скачать в pdf «Building the Data Warehouse»