Dimensional Data Warehousing with MySQL: A Tutorial

Overview

The first step to populate a data warehouse is extracting data from the source. You do this either by taking (pulling) the data out of the source or by requesting the source to send (push) the data to the data warehouse.

An important factor when extracting data is the volume and the availability of the source data, based on which you either extract the whole source data or just its changes since the last extraction.

This chapter covers the following two data sources extraction topics:

• Which part of a data source do you need to extract and load into your data warehouse? There are two common approaches, whole source and change data capture.

• The direction of data extraction. There are two possible modes, the pull mode (pulled by your data warehouse) and the push mode (pushed by the source).

The following sections explain the various modes of data source extraction briefly mentioned above, and then use sales order extraction as an example to show how a push-by-source and change extraction works.