Bug 2100282
| Summary: | Stopping httpd normally during ipa.service stop fails, causes 90 second delay before httpd is killed | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Adam Williamson <awilliam> |
| Component: | python3.11 | Assignee: | Tomáš Hrnčiar <thrnciar> |
| Status: | CLOSED ERRATA | QA Contact: | |
| Severity: | medium | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | rawhide | CC: | abokovoy, cstratak, ftrivino, ipa-maint, jcholast, jhrozek, mhjacks, mhroncok, pvoborni, python-maint, python-sig, rcritten, ssorce, thrnciar, torsava, twoerner, vstinner |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | openqa | ||
| Fixed In Version: | python3.11-3.11.0~b3-6.fc37 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2022-06-24 13:11:25 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
Adam Williamson
2022-06-23 00:47:28 UTC
I think this might be another sign of https://github.com/GrahamDumpleton/mod_wsgi/issues/730 THAT was it! I knew we had an existing issue similar to this, but I just couldn't find it. Thanks for the link. So...did something change in Python 3.11, or did something change in mod_wsgi? I think it must be Python 3.11, as there's no change to mod_wsgi that fits the timeframe where this bug re-appeared, which is between Fedora-Rawhide-20220619.n.0 and Fedora-Rawhide-20220621.n.3 . Ah, OK, so yeah, I think it is that again. What we did back in F34 / Python 3.9 timeframe was to revert the upstream Python commit that we identified as fixing the problem, in 00371-revert-bpo-1596321-fix-threading-_shutdown-for-the-main-thread-gh-28549-gh-28589.patch . That patch is still on our current python3.9 and python3.10 specs, but it was dropped from python3.11 when it was first created: https://src.fedoraproject.org/rpms/python3.11/c/cf1cfe95f840cfade40df7c32513237a00c876ae?branch=rawhide So it looks like this has never been fixed and has been lurking all along... Kicking to Python 3.11 for now. Python maintainers, can we have the reversion on the 3.11 branch too? The proximate cause of this bug showing up again is the reversion being dropped between our Python 3.10 and Python 3.11 packages. Yeah, I decided to drop this hoping it would be fixed in the meantime somehow better. We will add that back. FEDORA-2022-ca859a1deb has been submitted as an update to Fedora 37. https://bodhi.fedoraproject.org/updates/FEDORA-2022-ca859a1deb FEDORA-2022-ca859a1deb has been pushed to the Fedora 37 stable repository. If problem still persists, please make note of it in this bug report. |