IN2BI, Microsoft Business Intelligence
data, stories & insights
rss | email | twitter

DWH DECK Part 7: Data Mart

Adapted for version 1.1 on 18/04/2011

This is the seventh part of a series I'm writing on the DWH DECK. A tool that will allow you to build and maintain your data warehouse better and faster. This series will act as documentation and will give you some insides into the way the program works.

Overview of this series

The data mart area is the area that your user will use to query the data warehouse. The architecture is based on Ralph Kimball's dimensional modeling concept (star schema):
star schema - dimensional modelling 
There are two types of tables in the data mart area:

Fact tables
A fact table that contains the measurements associated with a specific business process. In the DWH Deck this translates to: Every link table in the data vault with measures in the associated satellites will be the basis for a fact table. A fact table will contain:

  • References to dimension tables
  • All related measures (in DWH Deck: check the Measure checkbox)
  • All related Time indicators (in DWH Deck: check the Time indicator checkbox)

 dwh deck - data mart 

Dimension Tables
Dimensions describe the objects of the business, such as employee, customer etc. In the DWH Deck this translates to: Every hub in the data vault that has a relation with afore mentioned fact tables and its related satellites.
Slowly changing dimensions
Managing changes over the time has always been a major issue in data warehousing. Kimball uses the concept of slowly changing dimensions. For every attribute/column you can enter the change type:
Type 1. Overwrite Old Data. (Default) Use this method if you don't care about keeping track of historical values.
Type 2. Track History. This is a powerful technique for capturing attribute values over time. In the ETL process a new row in the dimension table is added and the previous row describing the object is end dated.

» Similar Posts

  1. DWH DECK Part 6: Data Vault
  2. Four Approaches to Data Warehousing
  3. DWH DECK Part 5: Staging

» Trackbacks & Pingbacks

    No trackbacks yet.

» Comments

    There are no comments.

Comments are closed