preload

On Test Data

Posted by Albert Gareev on Jan 15, 2007 | Categories: Notes

Test Data should be designed.

  • Data, that we input to application-under-test, control business logic flow. To explore every branch of the flow, we need to choose data accordingly.
  • Wrong data trigger error-handling functionalities. To try and see all of them, we need to create data in different equivalence classes.
  • Test Data from Production might contain non-disclosing information
  • Test Data from Production reflect only “happy path” scenarios

 

Test Data should be manageable.

  • Test Data must be stored in files with organized structure: XLS, XML
  • Within a project, Test Data format should be standardized 
  • Test Data should not be duplicated
  • Volumes of Test Data should be indexed
  • Large volumes of Test Data need to be stored in database
  • For versioned applications, Test Data should be version-controlled too

 

 

Test Data should be shared.

  • Test Data is an asset for the entire team
  • Test Data should be reusable

Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported
This work by Albert Gareev is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported.