Data Warehouse

Some Natures of Fact Tables In DWH:

1.Fact tables can be factless. Fact table can be with no “measures”. The Fact table can be an intersection of “dimension”

2.There are some junk dimensions. Like “yes” or ‘no” columns. They may be needed in the source table for the OLTP application functionality. But they are NOT dimensions in DWH.

3.Some dimensions with the same meaning that appear in more than one fact table but with different PK. The meaning of the dimension in different fact tables have different attributes in source table/s. Example: week definition. Quarter identification by different Enterprises

DWH design after rolled into Production:

1.Once QA gives a thumb’s up

2.users are properly trained

3.Documentation is distributed

4.Documentation source is published to all stake holders

5.General Maintenance Schedules are announced and processis defined

6.Emergency Maintenance Process is determind

7.RCA process is defined

8.Change Management Process is defined

9.DWH application (frontend) and DWH are to be viewed differently

10.Enhancements (incremental) qualification, integration and deployment process is to be defined

11.DWH growth trends are observed and Retention policy is to be defined

12.DWH data archival Process is to be defined

PDF For More Information