Databases Reference
In-Depth Information
6. What type of data access will they have? Most business users will have
interactive query capabilities. Some may be recipients of more structured,
static reports. Again, this is driven by the business requirements.
The primary data storage for the bottom-up approach is here, in the presentation
servers. These are architected using conformed dimensions as a foundation for this
approach. The top-down approach may also physically store data in a similar manner,
called
data marts.
Conformed dimensions are not required with the top-down
approach, but they are highly recommended. If this part of the top-down approach
is implemented with conformed dimensions, then this layer is quite similar to the
bottom-up approach. With the top-down approach, this part of the architecture may
also be implemented with a semantic layer on the data warehouse itself, rather than
physically storing the data
.
Use the Data: Business Intelligence Applications
Technologies used to access data today often include some mechanism to
store data as part of how the tool works. The data architecture for using the
data is dictated by how the organization's data access tool(s) function. The
organization must understand each tool, and which is being used, so that
appropriate measures can be taken to ensure that the data continues to be
handled in a consistent, reliable manner.
Both the bottom-up and top-down approaches have similar data architectures here,
as it is driven by the data access technology
.
The data architecture role here is to ensure that the organization has a
clear understanding of what data is stored as part of the business intelligence
application. The goal is to have a structured, organized approach here too. For
a technology that requires the creation of a multi-dimensional cube of data, a
new cube should not be created for every new report that is needed. Rather,
more robust cubes should be created to support a variety of reporting needs.
These application-specific data stores must also have planned maintenance,
including updates, corrections, and additions to the data. The data architecture
must also recognize the need for backup and recovery of these data stores.
Now that the basic data architecture of the bottom-up approach has been
discussed, it is time to take a similar look at the top-down approach.
Top-Down Data Architecture
This approach was developed and popularized by Bill Inmon and Claudia
Imhoff. Several other key individuals have also contributed to the refinement
and success of this approach, which you can learn more about in
Corporate
Information Factory, Second Edition
,byW.H.Inmon,ClaudiaImhoff,andRyan
Sousa (Wiley, 2001) and
Mastering Data Warehouse Design
by Claudia Imhoff,
Nicholas Galemmo, and Jonathon Geiger (Wiley, 2003).
