Are you incorporating BI and data warehouse testing into your data world? If not, it’s time to give this more thought. The importance of testing your data warehouse and BI systems cannot be overlooked. Testing early in the process can save you time and effort, as well as allow you to make better business decisions.
In our Real World Data Testing series, we talked more in depth about this topic. Once you are on the bandwagon of BI and data warehouse testing, there is a common testing framework to consider before you sit down to write and execute your test. I discuss this here with John Welch, President of Pragmatic Works Software. This strategy (known as the three As) Arrange, Act and Assert is used in many types of testing.
Here’s where you set up all your preconditions for the test. Make sure there’s test data in place, any parameters are set, and your environment is set up the way you need it to be. Be sure your expectations are met for the preconditions for starting the test.
This is the action you are going to test the results of. In a data system, for example, that may mean running an ETL process, moving data from your source to your target.
Here, you set statements that would be true if the condition was met, and false if things did not go as expected. In our ETL example, your assertion may be: I expect that if I had 20 rows in my source table, I expect 20 rows in my target table.
The key is to consider these 3 steps before you begin testing. Most common test frameworks, such as NUnit or MS Test, all support the methodology of Arrange, Act, Assert. John goes into more detail about this strategy in his whitepaper, which you can download below.
So, if you’re ready to incorporate BI and data warehouse testing into your business, or are interested in learning more about this, Pragmatic Works can help. We are passionate about data testing and we’ve got the software, training and services to get you on your way. Visit our website or contact us today.
Leave a comment