[Python-Dev] Redoing failed PR checks (original) (raw)
Victor Stinner vstinner at redhat.com
Thu May 9 21:01:41 EDT 2019
- Previous message (by thread): [Python-Dev] Redoing failed PR checks
- Next message (by thread): [Python-Dev] Farewell, Python 3.4
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hi,
Le mer. 8 mai 2019 à 16:10, Mark Shannon <mark at hotpy.org> a écrit :
How do I redo a failed PR check? The appveyor failure for https://github.com/python/cpython/pull/13181 appears to be spurious, but there is no obvious way to redo it.
First, please make sure that a test failure is known in the bug tracker. Nobody here checked and no: it wasn't reported, whereas I also saw this bug once last week! (but I was too busy to report it, sorry!)
I reported the bug: "test_asyncio: test_drain_raises() fails randomly on Windows" https://bugs.python.org/issue36870
BTW, this is not the first time I've seen a PR blocked by a spurious appveyor failure.
Please check if every single test failure is known in the bug tracker, and if it's not the case: report it!
More info about Python CI: https://pythondev.readthedocs.io/ci.html
I'm trying to fix all tests which fail randomly for 5 years now (hopefully, with the help of other core devs like Pablo!). It's getting better, but sometimes new flaky tests pop up.
Victor
Night gathers, and now my watch begins. It shall not end until my death.
- Previous message (by thread): [Python-Dev] Redoing failed PR checks
- Next message (by thread): [Python-Dev] Farewell, Python 3.4
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]