0

Im using Testlink as our test management tool. We are using in agile environment

How to structure our test cases as test suite in our test management tool. Also how to maintain the tracerbility

We have a regression at the end of the sprint


At the moment we are maintaining 3 test suite. One for regression , one for functional and one for smoke.

Module-wise we have structure folder inside them and run them in execution mode

But just want to confirm whether this appraoch good in an agile way.

Also some people only care more about regression and they maintain it with that approach.

1

We use the following model:

  • We have some new functionality in scope of each Sprint
  • Having the scope of requirements we describe them in our test management tool. We group requirements by User Stories since we treat a User Story as a piece of altered functionality
  • After the previous step we have a number of requirements grouped by User Story which are grouped within Sprint. You can think of it as of a folder representing a Sprint, containing folders representing User Stories containing list of requirements
  • Having scope of requirements we start designing tests which are grouped in pretty much the same way: Sprint -> Story unless they are integration or E2E tests or other sorts of tests which do not imply one-to-one relationship to the feature (we have dedicated groups under Sprint for them).
  • For each test we store quite a lot of meta-information like: Component, Priority, Estimated Execution Time, If the test has ever detected defect, etc. All that helps us to compose a Regression set for each Sprint.

This model lets us trace the requirement to test run and vice versa.

  • 1
    With the user story grouping, how do you address user stories that modify the behavior created by previous user stories (so it is more expanding or modifying existing test cases than necessarily creating new ones)? This is an area I've really struggled to come up with a good answer for. – Daniel Aug 15 '18 at 14:09
  • We just compose a regression set for the functional area (see test meta-information) and while running the tests a tester encountering the changed logic just invalidates the test. This might not look super-effective but this saves time from continuous analyzing the tests for validity. Especially when you have thousands of tests. – Alexey R. Aug 15 '18 at 15:14
0

Test suite managing is an important part of the agile testing and is being followed by most of the Top software Testing Companies

We as a software testing company use the following model:

  • Create the test cases for all the product features already implemented and added the same in a Parent test suite.
  • The tag is already added in all test cases as per the requirement e.g regression, sanity, smoke, new functionality.
  • New features are introduced in various sprints. Before releasing the particular sprint with a new feature introduced we will create test cases for a new feature and add a tag in same as per the requirement e.g regression, new functionality.
  • When all testing is done for the sprint (all the tickets are validated), We create a new test suite for that sprint and add test cases in same as per the requirement (tag added in the test cases help to add the same in test suite as per requirement).
  • Once test suite is ready QA team start execution of the same and update the status on behalf of the same to the relevant person.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.