3

Where I work we have 3 environments for our product: Test, Mirror, Production.

Test is used for development purposes. Mirror is used for feature promotion, staging, and sometimes we share this environment with our key users so they can conduct UAT. Production is our live server.

For regular deployments to Production, we use this process:

  1. Functional testing for new features & bug fixes on the test server
  2. If all validated, promote to deploy to Mirror
  3. Regression testing for existing features in Mirror
  4. If all validated, promote to deploy to Production
  5. Smoke testing in Production

For hotfixes, is there a standard lightweight version of this process, and if so what is it?

It seems excessive to regress all features in Mirror for hotfix deployments.

2
  • Good question, but you have to explain it in a different way to avoid closing from other mafias as an opinion based question. – Bharat Mane Sep 7 '17 at 10:50
  • @BharatMane - I've edited to make the question less opinion-based. It is a good question. – Kate Paulk Sep 7 '17 at 11:52
1

It depends on the fix.

If the fix is something that hits low lying functionality used by the entire system, then the entire system needs to be regression tested.

If the fix involves the functionality of only one specific feature and has no interaction with any other feature, only that feature needs to be regression tested.

If the fix involves something trivial, then a smoke test level regression test is all that's needed.

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.