Basics of Data Warehouse Testing

Data Warehouse testing consists of ETL Testing and BI Testing

The path to Analytics We use our experience in implementing Data Warehouses to make your projects successful


single image

What is Data Warehouse (DWH) Testing?

Data Warehouses are composed of two major components
   - ETL (or ELT) for Extracting, Transforming and Loading data from multiple Data Sources to the Data Warehouse. The data is transformed into a Star (or Snowflake) Schema format to make it easy to report.
   - BI or Business Intelligence tools for reporting on the data in the Data Warehouse

Data Warehouse testing consists of ETL testing to validate that the data has been transformed and loaded as expected and BI testing to validate that the data shown in the reports is accurate. Data Warehouse testing is very important because the business relies upon this data to make key decisions. Bad data and issues in Data Warehouse can lead to failures of the Data Warehouse projects because lack of trust in the data and consequently low usage. validate that the data has been tranformed and loaded into the target as expected.


Challenges in Data Warehouse Testing

Data Warehouse Testing is different from application testing because it requires a data centric testing approach. Some of the challenges in Data Warehouse Testing are:
   - Data Warehouse testing involves comparing of large volumes of data typically millions of records.
   - Data that needs to be compared can be in heterogeneous data sources such as databases, flat files etc.
   - Data is often transformed which might require complex SQL queries for comparing he data.
   - Data Warehouse testing is very much dependent on the availability of test data with different test scenarios.
   - BI tools such as OBIEE, Cognos, Business Objects and Tableau generate reports on the fly based on a metadata model. Testing various combinations of attributes and measures can be a huge challenge
   - The volume of the reports and the data can also make it very challenging to test these reports for regression, stress and functionality.
Interested in more information?