| opendevreview | Pavlo Shchelokovskyy proposed openstack/tempest master: Use upper constraints in pep8 test run https://review.opendev.org/c/openstack/tempest/+/895627 | 09:34 |
|---|---|---|
| sean-k-mooney | well thats fun | 11:23 |
| sean-k-mooney | [12:22:38]➜ .tox/pep8/bin/stestr run | 11:23 |
| sean-k-mooney | stestr: 'run' is not a stestr command. See 'stestr --help'. | 11:23 |
| sean-k-mooney | Did you mean one of these? | 11:23 |
| sean-k-mooney | help | 11:23 |
| sean-k-mooney | init | 11:23 |
| sean-k-mooney | last | 11:23 |
| sean-k-mooney | we might have a broken stestr release to deal with | 11:24 |
| sean-k-mooney | https://zuul.opendev.org/t/openstack/build/a8265f5eef02422298ddab5ed7a1734e/log/job-output.txt#917 | 11:24 |
| sean-k-mooney | reported this here https://github.com/mtreinish/stestr/issues/350 | 11:35 |
| sean-k-mooney | however this is going to break the gate and i dont think its actully a bug in stestr directly but rather one of there deps | 11:35 |
| frickler | hmm, compare pip freeze with last passing run? | 11:41 |
| frickler | meh, seems this is no longer logged in the tox dir, likely a regression in tox4? | 11:42 |
| frickler | need to extract the info from job log | 11:43 |
| sean-k-mooney | i figured it out its a dupe of https://github.com/mtreinish/stestr/issues/347 | 11:43 |
| sean-k-mooney | python-subunit 1.4.3 released yesterday which drop the testtools extras depency | 11:43 |
| frickler | ah, nice, thx | 11:44 |
| sean-k-mooney | that means that the transitive use of testtools extra was lost | 11:44 |
| sean-k-mooney | so for right now dwongrading etestools to 2.5.0 works and i think the same woudl be trun for python-subunit 1.4.2 | 11:44 |
| sean-k-mooney | yep that also works | 11:45 |
| sean-k-mooney | so we can clamp either or both of them in uc | 11:45 |
| sean-k-mooney | until this gets fixed in stestr | 11:46 |
| sean-k-mooney | https://github.com/mtreinish/stestr/issues/347 shoudl resolve that but | 11:46 |
| sean-k-mooney | there has not been a release in a year | 11:46 |
| sean-k-mooney | so i think we might eb better off clamping for bobcat | 11:46 |
| sean-k-mooney | then we can unpin on master once there is a new stestr release | 11:47 |
| sean-k-mooney | https://review.opendev.org/c/openstack/requirements/+/895650 will resolve this i think | 12:12 |
| sean-k-mooney | actully shoudli clamp that in global-requiremnts instead | 12:13 |
| sean-k-mooney | uc is generated form global right? | 12:13 |
| frickler | testtools==2.6.0 has been in u-c for half a year, I don't think we should revert that. also use != instead of < in order to just denylist known broken versions | 12:28 |
| frickler | u-c is generated from global, but you still need to amend both in the same patch | 12:29 |
| frickler | sean-k-mooney: also see https://review.opendev.org/c/zuul/zuul-jobs/+/895606 , maybe we can just wait a moment for the new release? not sure which timezone mtreinish is in | 13:22 |
| frickler | oh, that was already done https://github.com/mtreinish/stestr/releases/tag/4.1.0 | 13:24 |
| sean-k-mooney | 46 minuts ago | 13:24 |
| sean-k-mooney | ok so maybe we can just skip the block | 13:24 |
| sean-k-mooney | and bump it to 4.1.0 | 13:24 |
| sean-k-mooney | frickler: actully thre is a spereate problem in nova-specs | 13:25 |
| sean-k-mooney | its not using uc | 13:25 |
| sean-k-mooney | so we clamp python-subunit in uc | 13:25 |
| sean-k-mooney | to a version that pulls in the transitive dep | 13:26 |
| sean-k-mooney | so if i fix that in nova-specs it might be enough to fix it | 13:26 |
| frickler | o.k., so this doesn't affect u-c using projects at all? | 13:26 |
| sean-k-mooney | and then we can just proceed to use the new release when it gets pulled in | 13:26 |
| sean-k-mooney | in theory the pinned version of python-subunit should mean uc projects are fine | 13:27 |
| frickler | it would be good if we can wait for the reqs repo to branch and then do the bump on master after that | 13:27 |
| sean-k-mooney | ill confirm that when i enable the use of uc | 13:27 |
| sean-k-mooney | ya ok using uc also fixes it due to the python-subunit pin | 13:30 |
| frickler | that explains why there has been no larger outburst of broken job complaints so far ;) | 13:32 |
| sean-k-mooney | yep we just started lookign at spec for the next release | 13:32 |
| sean-k-mooney | or i did today | 13:32 |
| sean-k-mooney | and saw it failing | 13:32 |
| frickler | but without pinning it should now be working again using stestr 4.1.0, right? | 13:33 |
| sean-k-mooney | yep | 13:36 |
| sean-k-mooney | i just confirmed that locally | 13:36 |
| sean-k-mooney | https://github.com/mtreinish/stestr/issues/350#issuecomment-1723425173 | 13:36 |
| sean-k-mooney | "stestr 4.1.0 also works with python-subunit 1.4.3 and testtools 2.6.0" | 13:36 |
| sean-k-mooney | im going to push my nova-specs fix and ill abandon the requirements patch | 13:37 |
| *** ykarel is now known as ykarel|away | 15:04 | |
| opendevreview | Merged openstack/devstack-plugin-ceph master: Remote Ceph with cephadm https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/876747 | 18:42 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!