-
Book Overview & Buying
-
Table Of Contents
The Data Warehouse Toolkit - Third Edition
By :
One of the biggest challenges with the incremental ETL process is identifying new, changed, and deleted rows. After you have a stream of inserts, modifications, and deletions, the ETL system can apply transformations following virtually identical business rules as for the historic data loads.
The historic load for dimensions and facts consisted largely or entirely of inserts. In incremental processing, you primarily perform inserts, but updates for dimensions and some kinds of fact tables are inevitable. Updates and deletes are expensive operations in the data warehouse environment, so we’ll describe techniques to improve the performance of these tasks.
As you might expect, the incremental ETL system development begins with the dimension tables. Dimension incremental processing is very similar to the historic processing previously described.
In many cases, there is a customer...
Change the font size
Change margin width
Change background colour