| opendevreview | Jianjian Huo proposed openstack/swift master: swift_proxy: add memcache skip success/error stats for shard range. https://review.opendev.org/c/openstack/swift/+/858942 | 03:45 |
|---|---|---|
| clayg | timburke: yet another case where it'd be nice to signal eventlet to close the socket https://review.opendev.org/c/openstack/swift/+/870541 | 14:53 |
| opendevreview | Jeremy Stanley proposed openstack/swift master: s3api: Prevent XXE injections https://review.opendev.org/c/openstack/swift/+/870823 | 15:03 |
| fungi | timburke: clayg: acoles: mattoliver: ^ | 15:04 |
| opendevreview | Jeremy Stanley proposed openstack/swift stable/zed: s3api: Prevent XXE injections https://review.opendev.org/c/openstack/swift/+/870825 | 15:05 |
| opendevreview | Jeremy Stanley proposed openstack/swift stable/yoga: s3api: Prevent XXE injections https://review.opendev.org/c/openstack/swift/+/870826 | 15:06 |
| opendevreview | Jeremy Stanley proposed openstack/swift stable/xena: s3api: Prevent XXE injections https://review.opendev.org/c/openstack/swift/+/870827 | 15:06 |
| fungi | hopefully i got the cherry-picks right, i don't do this very often ;P | 15:07 |
| opendevreview | Jeremy Stanley proposed openstack/swift stable/wallaby: s3api: Prevent XXE injections https://review.opendev.org/c/openstack/swift/+/870828 | 15:13 |
| fungi | looks like some bitrot in the docs job maybe? the sphinx autodoc extension is failing to import auditor from swift.account because of AttributeError: module 'dns.rdtypes' has no attribute 'ANY' | 15:47 |
| fungi | https://zuul.opendev.org/t/openstack/build/8bd962b96a2e444389884da3fda04165 | 15:47 |
| fungi | i don't see any fixes proposed for that problem yet, so presumably it's new | 15:48 |
| fungi | this appears to be https://github.com/eventlet/eventlet/issues/781 | 15:52 |
| fungi | pinning eventlet<2.3 or eventlet!=2.3.0 (depending on how much expectation there is that 2.3.1 would fix it) seems to probably be the fastest workaround | 15:53 |
| fungi | er, no it's dnspython<2.3 we'd need a pin for, the and the dnspython maintainer has already decreed that it's exposing a bug in eventlet's monkey-patching which he's disinterested in working around in dnspython itself | 16:10 |
| fungi | openstack/requirements upper-constraints.txt still sets dnspython===2.2.1 so i guess this is only an issue for unconstrained builds | 16:12 |
| fungi | interestingly, the pip install claims to have applied upper-constraints.txt | 16:17 |
| fungi | i'm able to reproduce the same locally too | 16:40 |
| fungi | even with older versions of tox | 16:43 |
| fungi | it looks like it has to do with multi-stage installation... something in one of the tox install stages other than for deps is installing something which also depends on dnspython, and adding the constraints option to testenv.install_command seems to work around it but i'm unclear on whether that's the preferred solution | 16:53 |
| fungi | i'll push up a brown-bag change for further discussion | 16:53 |
| opendevreview | Jeremy Stanley proposed openstack/swift master: Force constraints in Tox's install_command too https://review.opendev.org/c/openstack/swift/+/870844 | 16:57 |
| timburke | fungi, thanks for all that -- i'll get it sorted today | 17:10 |
| fungi | thanks! | 17:11 |
| fungi | i'm a little worried about the approach in 870844 since it may break lower constraints and python 2.7 jobs. hopefully zuul will tell us | 17:11 |
| fungi | oh, actually the zuul jobs will probably work because they set the TOX_CONSTRAINTS_FILE envvar, it's only local invocations for envs needing alternate constraints lists which would be affected | 17:15 |
| fungi | looking at other projects, i do see it's a common approach to add the -c option in tox.install_command though | 17:17 |
| clarkb | fungi: you might be able to set TOX_CONSTRAINTS_FILE as an env var with a default to the lower constraints file in the lower constraints target? | 17:24 |
| clarkb | assuming such a target exists | 17:24 |
| fungi | yeah, or to /dev/null if you want to run with no constraints file | 17:25 |
| opendevreview | Jeremy Stanley proposed openstack/swift master: Force constraints in Tox's install_command too https://review.opendev.org/c/openstack/swift/+/870844 | 17:31 |
| opendevreview | Tim Burke proposed openstack/swift master: Fix docs build https://review.opendev.org/c/openstack/swift/+/870853 | 18:33 |
| timburke | ^^^ trying a more targeted fix -- api-ref and release note jobs might still be grumpy, though, if i've understood the problem correctly... | 18:39 |
| opendevreview | Jianjian Huo proposed openstack/swift master: swift_proxy: add memcache skip success/error stats for shard range. https://review.opendev.org/c/openstack/swift/+/858942 | 18:44 |
| opendevreview | Tim Burke proposed openstack/swift master: Add LDFLAGS and CPPFLAGS to tox's passenv https://review.opendev.org/c/openstack/swift/+/870858 | 19:05 |
| opendevreview | Tim Burke proposed openstack/swift master: Clean up project URLs for PyPI https://review.opendev.org/c/openstack/swift/+/870862 | 19:57 |
| opendevreview | Tim Burke proposed openstack/swift master: Don't run reno as part of building an sdist https://review.opendev.org/c/openstack/swift/+/870863 | 19:58 |
| opendevreview | Tim Burke proposed openstack/swift master: Skip coverage reports when running pytest directly https://review.opendev.org/c/openstack/swift/+/870865 | 20:06 |
| timburke | as for the eventlet issue, i submitted https://github.com/eventlet/eventlet/pull/782 -- we'll see how it's received | 20:43 |
| opendevreview | Merged openstack/swift master: Fix docs build https://review.opendev.org/c/openstack/swift/+/870853 | 22:06 |
| opendevreview | Tim Burke proposed openstack/swift stable/zed: Fix docs build https://review.opendev.org/c/openstack/swift/+/870780 | 22:12 |
| opendevreview | Tim Burke proposed openstack/swift stable/yoga: Fix docs build https://review.opendev.org/c/openstack/swift/+/870781 | 22:12 |
| opendevreview | Tim Burke proposed openstack/swift stable/xena: Fix docs build https://review.opendev.org/c/openstack/swift/+/870782 | 22:12 |
| opendevreview | Tim Burke proposed openstack/swift stable/wallaby: Fix docs build https://review.opendev.org/c/openstack/swift/+/870783 | 22:12 |
| opendevreview | Jianjian Huo proposed openstack/swift master: swift_proxy: add memcache skip success/error stats for shard range. https://review.opendev.org/c/openstack/swift/+/858942 | 22:28 |
| timburke | eventlet fix is merged & tagged :-) shouldn't be a problem in 0.33.3 | 23:12 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!