-
Notifications
You must be signed in to change notification settings - Fork 99
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
testng parallel methods mode failures #491
Comments
Here is the isolated issue:
and the
If changing the above to parallel="none" then it works. Also if decreasing I am getting mostly NullPointerExeception, but saw also EmptyStackException, and ConcurrentModificationException . ...and another weird issue is that the fields in additional stages collide with these in the direct stages unless I changed their names, e.g. from |
Hi, @adrian-herscu Unfortunately, you question about thread-safe can at the moment only be answered by @janschaefer, as he was solely responsible for the design. In any case, I want to thank you for isolating the issue and providing a test against which to test. However I am right now lacking the time to look into the issue immedeately, so I have to ask you to be patient. Best, |
I want to tackle the Issue alongside #823, so I'm closing this PR. I will take care of it though in #829. Regarding the weird behavior of the scenario states when you call all of them "state": If you do that, then the fields have the same name and type. We have no chance to distingish fields there, because in that precise case we want to consider them as related. Spring does this in much the same way. |
Recently, I discovered that when running with TestNG in parallel=methods mode I am getting ConcurrentModificationException on ScenarioExecutor#stages field.
I am trying to isolate the issue and have a clean test that reproduces.
Yet another related issue is that ExpectedScenarioState fields are injected with instances from other threads causing test methods to fail due to uninitialized instances of these fields.
Is it safe to assume the JGiven was designed with thread safety in mind?
The text was updated successfully, but these errors were encountered: