BASICS: VDDW Architecture   >   Overview

Enlarge VDDW Architecture VDDW Components: Overview Common Dimensions Common Schemas ABC Allocation Activity Dimension Contribution & Consumption OLAP HyperCube What's Left Out

The VDDW is the integrated whole of all the pieces depicted in the diagram to the right. To begin with, VDDW presupposes a dimensional data modeling approach. Absent a dimensional approach, and by extension, associated OLAP tools, the VDDW is simply uncompelling. (Ralph Kimball is the undisputed king of dimensional design, and for those seeking to dive deeper into dimensional design fundamentals, his books and publications are the best resources available.)

• As depicted in the diagram above, the VDDW represents a significant addition (in terms of size, complexity, and value) to the conventional corporate dimension bus. It presumes the existence of common conformed dimensions and star schemas, like those depicted in the upper-left and middle-left portions of the diagram. However, it adds to these a new dimension (the activity hierarchy), representing the enterprise's core business activities, the resources the enterprise contributes to those activities and the consumption of those activities by the company's partners and products via individual business transactions. 

• The new star schemas introduced in conjunction with this new dimension obviously must somehow be populated with data, which is the role of an ABC allocation engine that enables modelers to set up allocation rules and structures that jibe with the corporate dimension bus. Fortunately, industrial strength tools for doing so are now commercially available (though it takes some expertise in both dimensional modeling and ABC allocation mechanics to get the model right for both purposes).

• Next, it involves creative use of OLAP as the primary means for staging the entire VDDW (not just one subject area) to end users via multiple physical cubes (or "measure groups") enwrapped in one virtual cube (or "hyper-cube"). This enables ad hoc analysis of all possible ratios and drivers supported by the underlying data via a single structure (the hyper-cube) that encompasses all dimensions.

• Finally, the VDDW approach does not attempt to address every possible DW / BI sub-discipline or new technical innovation (e.g. ODS, Text Integration, Hardware appliances, Predictive Analytics, etc.), because at its core VDDW is a modeling and design approach – an approach for building and integrating a single system that delivers as much bottom-line value to the enterprise as efficiently as possible. When this goal is achieved numerous peripheral tools and concerns find their proper context and priority, and are neither precluded nor presumed.

•  

• Learn More: Information Supply Chain Basics How VDDW Streamlines the ISC VDDW and Corporate Performance Management How VDDW Delivers Results VDDW Distinctives