Operational data stores

The operational data store is a physical area in the data warehouse that contains the latest snapshot of the operational data. During the extract, transform, and load (ETL) process, the builds extract data from the operational system and map the data to the operational data store area in the data warehouse. This topic is an overview of the operational data store used in the sample catalog.

Extracting data

The source data is extracted through the XML ODBC driver from data services or XML data files. Data from data services is extracted through staging: small ETL builds extract the data, and store them into temporary tables from where other ETL builds pick up the data, transform the data, and map it to the operational data store area of the data warehouse. For products that support delta loads, extraction from data services is through delta loads. The structure of source data is specific to the data source. The attributes are extracted according to the measurement objectives. Therefore, not all attributes available in the data sources are loaded to the data warehouse.

Transforming data

The transformation models do not contain complex business rules, or aggregations and calculations. The transformation of attributes happens in the following manner:

Delivering data

Similar data from different data sources are mapped to the same table in the data warehouse. The data is stored according to the subject or business domain. For example, a defect from Rational® ClearQuest and a defect from Rational Team Concert™ is mapped to the same REQUEST table in the operational data store. Some of the mappings that are common to most of the data sources are: Specific mappings are given in the product-specific sections.

Feedback