All data warehouse implementations encounter periods of debate in addressing the links between the conceptual modeling of business ideas, the logical modeling of related data elements, and the physical modeling of database structures. While academic research on the most effective datamodeling practices exists, the reality of physical implementation requires a history of experience - seeing the strengths and shortfalls in each of the recommended approaches. The evolution of your data warehouse structures will depend on the outcome of these debates. An understanding of the use of these structures at each layer of your implementation is essential for a successful implementation.
The first step of the warehouse design effort begins in the architecture of the warehouse staging layer. It is in this layer that incoming files of various formats begin to be cleansed and rationalized. The nuances of implementation within this layer are often product specific, driven by the data acquisition tools used within your institution. Traverse Consultants have direct development experience with the market leading data acquisition tools to date – Informatica, Ab Initio, and Ascential. This experience will prove imperative to your staging layer design.
Traverse Consultants are experts in designing Inmon-based relational structures for use in the batch processing layer of your data warehouse. Many architects advocate a pure Kimball-based, dimensional design that extends immediately from the denormalized staging layer. Those that have actually deployed programs within such structures understand that executing intensive batch processing for allocations, reconciliations, or other analytic processing is next to impossible when data input exists in such a fully dimensional design. An operational layer, defined as a relational database, is imperative for the consolidated data processing that is needed within the data warehouse. Traverse Consultants can lead this design and build effort.
The final layer of the data warehouse must be rooted in Kimball-based, dimensional structures that have proven most adequate for data access. It is in this layer of the data warehouse that all efforts of the preceding layers are consolidated for presentation to the end-user. An effective design of the dimensional layer is essential for meeting the expectations of your business sponsors. Traverse Consultants can ensure your design does not fall short of these expectations.
Call us today:
Call us today to see how Traverse Consulting's blend of technical expertise and functional direction can make your EPM implementation successful.
Contact us and receive answers to all your questions!