• Author
  • #19466

    We’re setting up regression testing and CI with Jenkins, and creating a test framework for this.

    Should application set-up (starting the application, setting proxy ports, starting database etc) be done in the Jenkins pipeline or should this be done in the test framework? Any recommendations?


    I would like to suggest you to do it separately through Jenkins pipeline. The more generic the framework is, maximum reusability it has.

    Why to keep the set up separate from the framework?

    • Makes it easier for maintenance
    • Easy to locate and figure out the reason for errors if it is caused because of the setup
    • Clear reports

    Do you have any specific reason to create a test framework rather than choosing from already available frameworks?


    At the moment we have a “set-up” stage in Jenkins pipeline (maybe there’s a way to not make it show though…). But I’m not sure if we later on will discover that we lack some flexibility, and that we’ll have to change the pipeline for every new set-up too.

    We already have a small collection of simple python modules for regression testing. The frameworks on the internet seems complicated.


    If this framework is solely made to fit only for a specific requirement then you can have this set up within the framework. In the end it should help you to resolve your problems and not by creating another. But in general, I would say a framework has to be generic for wider audience to build additional functionality from it.

    Congrats and all the best for your set up and new framework 🙂


    Yes… I think you’re right. It probably depends on how general the framework needs to be. Thanks for answering!

Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic.