-
-
Notifications
You must be signed in to change notification settings - Fork 2.6k
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
Allow tests to be logged multiple times with pytest-xdist #1193
Comments
Hmmm I'm not sure how to implement that other than making xdist explicitly know about "reruns". def slave_testreport(self, node, rep):
if rep.when == "call" or (rep.when == "setup" and not rep.passed):
self.sched.remove_item(node, rep.item_index, rep.duration) xdist assumes that a I'm not sure how to solve this without adding new hooks to Another idea would be to grow a new attribute into the report object, something like IMHO though I don't see a problem with |
Related: This should affect Box's flaky plugin for pytest. |
this issue cant be fixed unless we reassert that test node ids are unique |
@nicoddemus & @RonnyPfannschmidt do you have any new thoughts on this? I'm keen to address it, but may need some assistance from one or both of you. |
this issue is sill unfixable without changing xdist at its core |
@RonnyPfannschmidt can we move the issue via some admin interface, or would you like me to raise it under xdist and close this one? |
there is no admin ui to do that, gh is just fancy, not cappable |
Kamino closed and cloned this issue to pytest-dev/pytest-xdist#206 |
Note that this is not related to #927
In pytest-rerunfailures the test protocol is run multiple times in the event of a failure. An enhancement will log each iteration of the test with an outcome of 'rerun' until the maximum number of reruns is reached, at which point the genuine outcome is preserved. This works fine unless running the tests without pytest-xdist, which causes issues when the first rerun of a test is attempted to be removed from the scheduler:
For now I can avoid logging multiple reports if we're running tests in parallel, but that does mean that the only indication of multiple tests running will be the multiple logstart calls. Ideally there should be no difference between running tests in serial or parallel.
The text was updated successfully, but these errors were encountered: