| *** dviroel|afk is now known as dviroel | 11:24 | |
| Mounika | Hi reviewers, In Zuul , Openstack-tox-pep8,py38,py310 got passed in 1st run for this patch https://review.opendev.org/c/openstack/cinder/+/869292. In the next runs, these are failing with error "tox.tox_env.python.api.NoInterpreter: could not find python interpreter matching any of the specs functional-py38". Any idea on this? | 12:23 |
|---|---|---|
| Mounika | whoami-rajat,rosmaita Is it related to tox ? | 12:23 |
| rosmaita | Mounika: yes, it is, was introduced in tox 4.1.2, and is still open: https://github.com/tox-dev/tox/issues/2811 | 12:56 |
| rosmaita | i thought there was a workaround, but it seems to introduce other problems: https://review.opendev.org/c/openstack/python-cinderclient/+/869226 | 12:57 |
| rosmaita | not sure what to do at this point | 12:57 |
| rosmaita | also, it may not be issue #2811 exactly | 12:58 |
| Mounika | rosmaita ok.Thanks for the information Brian. Then, do we have to wait until #2811 fixes? | 13:03 |
| rosmaita | #2811 isn't a priority in upstream tox, because it was reported as only applying when 'usedevelop=true' | 13:04 |
| rosmaita | that cinderclient patch removes usedevelop=true but then crashes and burns elsewhere | 13:05 |
| rosmaita | i assume that will happen with the cinder patches, too, but i have not tried it yet | 13:05 |
| rosmaita | we'll have to see what whoami-rajat thinks, but i am thinking we may need to go back to tox3 in master until upstream tox 4 is stable | 13:06 |
| rosmaita | unless there's a simple fix that i'm not seeing yet | 13:07 |
| whoami-rajat | rosmaita, with all the issues we're facing, I'm not a fan of moving to tox4 either, also there isn't anything we're missing out | 13:07 |
| rosmaita | yeah, the only downside is that eventually we'll have to make the move, so we're taking on technical debt | 13:08 |
| rosmaita | but the current situation seriously sucks | 13:08 |
| Mounika | rosmaita yes, that will be better thing to move to tox3 until tox4 is stable | 13:08 |
| rosmaita | i wonder if nova is hitting these problems, and if not, how they're getting around them | 13:10 |
| rosmaita | i'll look into that first | 13:10 |
| Mounika | rosmaita ok. Thank you | 13:16 |
| happystacker | hi team, anyone having issues with tox and python 3.8 , 3.10 checks? | 13:18 |
| happystacker | is it related to https://github.com/tox-dev/tox/issues/2811 ? | 13:19 |
| rosmaita | whoami-rajat: it doesn't seem to be hitting nova unit tests (not sure why, though they don't require a mintox = 4.0.0, so maybe they're just lucky); it's hitting the nova docs job, though | 13:58 |
| whoami-rajat | rosmaita, oh, ok, are the issues intermittent or they always occur? like wrong python interpreter being used | 14:21 |
| rosmaita | the wrong interpreter got fixed in 4.2.4 (by stephenfin!) | 14:21 |
| whoami-rajat | oh nice | 14:22 |
| rosmaita | but as far as i can tell, if you use tox4 the way openstack does, you'll hit #2811 | 14:23 |
| rosmaita | it was introduced in 4.1.2 | 14:23 |
| whoami-rajat | do we set --skip-missing-interpreters false in our jobs? | 14:31 |
| rosmaita | we do that in tox.ini, but zuul does that when it calls tox when it runs the job | 14:36 |
| whoami-rajat | ok, looks like an issue then | 15:25 |
| opendevreview | Eric Harney proposed openstack/cinder master: Tests: Fix double mock of Popen (test_backup_ceph) https://review.opendev.org/c/openstack/cinder/+/867824 | 15:37 |
| opendevreview | Walt proposed openstack/cinder master: Fix tracking allocated_capacity_gb https://review.opendev.org/c/openstack/cinder/+/836083 | 15:53 |
| rosmaita | tox 4.2.5 was just released, believe it fixes the problems we were seeing (thanks to stephenfin!) | 16:10 |
| rosmaita | go ahead and recheck patches, but please use this comment: recheck tox4 issues | 16:11 |
| *** dviroel is now known as dviroel|lunch | 16:39 | |
| *** dviroel|lunch is now known as dviroel | 17:50 | |
| happystacker | rechecking | 18:08 |
| opendevreview | Ghanshyam Mann proposed openstack/python-cinderclient master: DNM: testing devstack run-both.yaml playbook tox<4 pin https://review.opendev.org/c/openstack/python-cinderclient/+/869494 | 18:28 |
| happystacker | seems that py38 and py310 are running fine | 19:10 |
| happystacker | I still have issues with tox pep8 | 19:10 |
| happystacker | but looks ok now | 19:12 |
| happystacker | issues with tox py38 and py310 are gone, but still happening for openstack-tox-functional-py38 and openstack-tox-functional-py310 | 19:28 |
| happystacker | openstack-tox-functional-py39 sorry | 19:28 |
| opendevreview | Jean Pierre Roquesalane proposed openstack/cinder master: Dell PowerStore: Volume caching exception support https://review.opendev.org/c/openstack/cinder/+/867753 | 19:29 |
| *** dviroel is now known as dviroel|pto | 21:12 | |
| opendevreview | Brian Rosmaita proposed openstack/python-cinderclient master: DNM: check functional jobs https://review.opendev.org/c/openstack/python-cinderclient/+/869226 | 22:07 |
| opendevreview | Ghanshyam Mann proposed openstack/python-cinderclient master: DNM: testing devstack run-both.yaml playbook tox<4 pin https://review.opendev.org/c/openstack/python-cinderclient/+/869494 | 23:03 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!