test: rerun flaky test_shutdown_timeout

[jpnurmi]

How about automatically re-running the flaky test_shutdown_timeout test to mitigate frequent CI failures?

sentry-native/tests/test_integration_http.py

Lines 561 to 567 in c5367b4

	# the timings here are:
	# * the process waits 2s for the background thread to shut down, which fails
	# * it then dumps everything and waits another 1s before terminating the process
	# * the python runner waits for 2.4s in total to close the request, which
	# will cleanly terminate the background worker.
	# the assumption here is that 2s < 2.4s < 2s+1s. but since those timers
	# run in different processes, this has the potential of being flaky

[JoshuaMoelans]

TIL that this mark.flaky exists! Nice QOL addition 👍

[JoshuaMoelans]

@getsentry/dev-infra license compliance seems to fail because of us adding pytest-rerunfailures ; anything we can do to clear this up?

[jpnurmi]

Luckily, there are multiple ready-made pytest extensions for dealing with flaky tests. Let's try this instead: https://pypi.org/project/flaky/ 😄

[jpnurmi]

Interesting how manual re-runs often pass, but this didn't seem to help... 🙁

https://github.com/getsentry/sentry-native/actions/runs/21359409357/job/61474580850