??? Just as
efforts to define an enterprise-wide OLTP model can take years (due to crossdepartmental
politics and the scope of the effort), similar attempts in data
warehousing also took much longer than business sponsors were willing to accept.
These efforts were further hampered by the continually changing analysis requirements
necessitated by a changing marketplace. While the data elements and
Business Intelligence Basics | 225
requirements for operational systems can remain relatively stagnant over time,
understanding business trends can be like trying to catch lightning in a bottle.
Consequently, attempts at building such enterprise-wide models that would satisfy
everyone often satisfied no one.
Data Marts
When some large-scale, enterprise-only data warehouse efforts ended in dismal
failure, frustration and impatience followed. Some reacted and built departmentfocused
independent data marts by extracting data from the appropriate operational
Figure 10-1. Typical initial business intelligence topology
Warehouse Server
Data Marts
Clients
Operational Data Store
OLTP Servers
226 | Chapter 10: Oracle Data Warehousing and Business Intelligence
source systems.
Pages:
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496