July 10th, 2014 by Adam Sandman
quality assurance release management
How many of you have been there? You have to make a major decision that could negatively impact the entire company. In your mind you are weighing, is it really a Sev1? Could we release with notes? What should we do?
Making the Call
Being that it was late on Sunday, Yes he had worked all weekend to do the cycle, he had no choice but to stop a release scheduled for Monday morning.
That was the right call.
There was no one who could rewrite the release notes. There were several Sev1 defects that affected major business processes. Sure, management would not like it, but stopping the release was better than souring the customers with an unworkable release.
Monday came, the QA director went in, and was promptly let go, yes fired for stopping the release. Unexpected, YES. The right thing to do, NO!
What is the real problem?
I would propose that the real problem here is not a stopped release, but a fundamental breakdown in the SDLC.
- There was no time to test, the QA director was in on the weekend testing a Monday release.
- Testing was too late in the cycle, with found Sev1 defects, there was no time for correction and a retest cycle.
- Expectations of quality were not equal. Management was willing to release an unworkable build to meet a deadline, the QA director was not.
The moral of the story, talk to management; let them know the real cost of a bad release. Let them know when testing should be done. Stay informed as to release cycles and by all means, participate in the planning. QA input on a release should not just be at the end, It should be everywhere from inception to release.
This is a real story, the names have been omitted to protect the innocent (and the not so innocent!).