Join 34,000+ subscribers and receive articles from our blog about software quality, testing, QA and security.

How to setup Testrail with regular monthly releases


Hi, we have monthly releases. I created a project that covers all releases. Next, I would define one test suite, again covering all releases (rather than a suite per release). Within the suite I then would define a test run for each release.
I would also create a milestone for each release. I’m not sure how the milestones and suites/runs relate. If you wouldn’t mind sharing your experiences, pls do.
Rishi Samlal


The milestone is the release. Associate a test run with that milestone and you’re good to go. Next month, new milestone, new test run. No need to change anything unless you mean to.


Thanks George, this is helpful. So the milestone goes hand in hand with test run.
I take it that the test suite is not setup from a time angle but more from a topic angle.
So I look at it as a repository with test cases that can be pulled from when defining the test runs.
Does this make sense?


That’s totally right Rishi. A test suite is just a way to organize test cases, which are generally pretty stable. A test plan is a time-dependant entity associated with a milestone, and it regroups test runs. While the test runs apply to single test suites (and you can select subsets within that), there can be several iterations over time for the same suite. If you want to be formal about it, because you can also add another layer of results to a case within the same run. And/or there can be test run variations on the same suite if you’re using conditions to allow for different OS/browsers/etc for instance.

One thing I haven’t figured out that I could use is how to associate sub-milestones and test runs, if someone can show me that I’d be grateful…


Hi, we have sprints with one or more epics per sprint. How would you setup Test Suites and Milestones?

I have now a Suite for each Sprint and a milestone against each Epic within the sprint. Next is how to setup the various test cycles (SIT, FAT, UAT, Regression test, Performance test, etc).

I would think to create test runs for each mentioned test cycle. Would anybody see any drawbacks in this approach?

regards, Rishi Samlal