Open enkore opened 7 years ago
this does sound like an integration issue with pytest-timeout, its unlikely i can investigate this myself timely soon
can anyone confirm if the timeout plugin is the source of the issue?
The upper trace is a known pytest-timeout bug: https://bitbucket.org/pytest-dev/pytest-timeout/issues/8/internalerror-valueerror-signal-only-works
But I'd say xdist should be able to deal with exceptions raised in pytest_xxx hooks; or at least should abort the test session in this case — it shouldn't return 0 and report no failures.
—
I noticed that it actually does return 1 (error), but it doesn't report the failure (== 26 passed, 1 skipped in 22.34 seconds ==
).
thanks for verifying the details
Hi,
I am running into similar issue, not sure what is causing. My Jenkins job is crashing with same log.
Need help on that.
thank you, Cauvery.
I encounter this issue with the following list of plugins:
plugins: asyncio-0.14.0, forked-1.3.0, drop-dup-tests-0.3.0, mock-3.2.0, timeout-1.4.2, xdist-1.34.0
This happens rarely when the slave receives a signal. Doesn't seem to have anything to do with a specific test, but just randomly. It's possible that this is an interaction with pytest-timeout, which causes a crash in pytest_runtest_protocol.
When this happens, the state of the session seems to become corrupted, as the master does not note a failed test and continues on, reporting no errors, and executing only a small subset of all tests.
A full log is available at: https://travis-ci.org/borgbackup/borg/jobs/260118022