| *** Guest318 is now known as diablo_rojo_phone | 04:06 | |
| *** diablo_rojo_phone is now known as diablo_rojo | 04:10 | |
| *** ysandeep|out is now known as ysandeep | 05:09 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Add wait_time param for logscraper and logsender https://review.opendev.org/c/openstack/ci-log-processing/+/836119 | 06:28 |
|---|---|---|
| *** akahat is now known as akahat|rover | 06:35 | |
| *** jpena|off is now known as jpena | 07:11 | |
| opendevreview | Merged openstack/openstack-zuul-jobs master: Move lower constrants job to ubuntu focal https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/834179 | 07:46 |
| *** ysandeep is now known as ysandeep|lunch | 08:31 | |
| *** ysandeep|lunch is now known as ysandeep | 09:15 | |
| *** prometheanfire is now known as Guest928 | 10:40 | |
| *** dviroel|afk is now known as dviroel | 11:22 | |
| *** ysandeep is now known as ysandeep|afk | 11:51 | |
| *** soniya29 is now known as soniya29|afk | 12:04 | |
| rpittau | hello everyone! Since roughly 3 days we're seeing an issue with our bifrost CI where the libvirt-python library is downloaded with pip as source instead of wheel, this breaks our upgrade job in current master as well as the jobs in stable/yoga | 12:13 |
| rpittau | this is the error that we're seeing, we're aware that we can't build libvirt-python from source there, just please focus on the tar.gz archive being downloaded instead of whl as it should be https://paste.openstack.org/show/blOh0kXqDJX2aHPRd36R/ | 12:15 |
| rpittau | this is instead from a successful job https://paste.openstack.org/show/bwgYHNGpqBhdqyKxE7U6/ | 12:17 |
| dmellado | rpittau: that's weird, may be a pip bug? it should prefer wheels whenever available | 12:24 |
| rpittau | maybe, but it doesn't look like pip version has changed recently | 12:25 |
| rpittau | I mean, the one we're using in CI | 12:25 |
| fungi | is it failing consistently or only sometimes? | 12:26 |
| rpittau | fungi: consistently | 12:26 |
| fungi | which distro/node-label? | 12:28 |
| fungi | and which python version? | 12:28 |
| rpittau | fungi: python 3.6 on centos-8-stream | 12:29 |
| fungi | okay, details i can see so far... libvirt-python does not provide prebuilt generic (manylinux) wheels on pypi, so in your working paste a centos-8 specific wheel was fetched from our wheel builder cache | 12:35 |
| fungi | the wheel fetched from our cache was for libvirt-python 7.6.0 | 12:35 |
| rpittau | yep | 12:35 |
| fungi | in the broken example it looked in our wheel builder cache but decided to grab the sdist of 7.6.0 from pypi (via our proxy) instead | 12:36 |
| rpittau | that is happening all the time now | 12:37 |
| fungi | rpittau: i see the problem | 12:39 |
| rpittau | \o/ | 12:39 |
| fungi | we don't have a wheel/centos-8-x86_64 cache any longer since we removed our centos-8 images | 12:40 |
| fungi | is should instead look in wheel/centos-8-stream-x86_64 | 12:40 |
| rpittau | mmmm ok, we're using centos-8-stream though | 12:41 |
| fungi | so we need to figure out what's still setting the wheel cache pypi consults to centos-8 instead of centos-8-stream on the centos-9-stream images | 12:41 |
| fungi | s/images/nodes/ | 12:41 |
| fungi | centos-8-stream nodes i mean | 12:41 |
| rpittau | ok, clear | 12:42 |
| fungi | rpittau: dmellado: i think the problem is here: https://opendev.org/opendev/base-jobs/src/branch/master/roles/mirror-info/templates/mirror_info.sh.j2#L61-L65 | 12:44 |
| fungi | we're still special-casing centos-9 for stream and not 8 | 12:44 |
| fungi | we should be able to drop the conditional at this point and just go with the first form | 12:45 |
| fungi | i'll push up a fix | 12:45 |
| rpittau | fungi: thanks a lot for the help! :) | 12:47 |
| fungi | rpittau: dmellado: https://review.opendev.org/c/opendev/base-jobs/+/836181 No longer special-case CentOS Stream | 12:48 |
| rpittau | +1'ed thanks again! | 12:49 |
| *** ysandeep|afk is now known as ysandeep | 13:28 | |
| *** dasm|off is now known as dasm|ruck | 13:38 | |
| *** ysandeep is now known as ysandeep|out | 14:53 | |
| *** dviroel is now known as dviroel|lunch | 15:09 | |
| *** dviroel|lunch is now known as dviroel | 16:00 | |
| opendevreview | Jeremy Stanley proposed openstack/project-config master: Match configure-mirrors for CentOS wheel URLs https://review.opendev.org/c/openstack/project-config/+/836200 | 16:14 |
| *** rlandy is now known as rlandy|biab | 16:55 | |
| *** jpena is now known as jpena|off | 17:04 | |
| *** rlandy|biab is now known as rlandy | 17:45 | |
| *** dviroel is now known as dviroel|brb | 20:05 | |
| *** dviroel|brb is now known as dviroel | 21:16 | |
| melwitt | has anyone else seen stable/xena lower-constraints jobs fail with "ERROR: InterpreterNotFound: python3.6"? https://zuul.openstack.org/build/142d9b1c4180480f8fdaffb144fb88d7/log/job-output.txt#10495 | 21:58 |
| melwitt | I'm guessing it's related to https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/834179 but I don't understand why it's trying to use 3.6 | 21:59 |
| clarkb | looking | 22:09 |
| melwitt | hm, maybe it's bc of our tox.ini | 22:10 |
| melwitt | I see things in there that want to use python3.6 | 22:10 |
| clarkb | ya I think the job is running on ubuntu focal and python3.6 is not available there so anything hardcoded to use python3.6 will fail | 22:11 |
| clarkb | that job runs tox -e lower-constraints which if pinned to python3.6 will fail that way | 22:11 |
| melwitt | ok I see | 22:11 |
| clarkb | I suspect that there are two options here. Drop the job from older branches or update openstack-zuul-jobs to have a job for older branches on older ubuntu with older python | 22:12 |
| melwitt | hm, looks like it's just one of the envs, functional-py36. we also have functional-py37 and functional-py38 too | 22:12 |
| melwitt | oh wait | 22:13 |
| fungi | worth noting, there was consensus on the openstack-discuss ml that lower-constraints jobs are mostly a waste of effort, but also that they're entirely unmaintainable on stable branches and should only be run on master if at all | 22:13 |
| melwitt | yeah ok I found it | 22:13 |
| clarkb | https://opendev.org/openstack/nova/src/branch/stable/xena/tox.ini#L346-L347 is where python3.6 comes from | 22:13 |
| clarkb | fungi: ya so drop the job from stable branch is liekly the best option | 22:14 |
| melwitt | thanks | 22:14 |
| fungi | of the projects which haven't dropped their lower-constraints jobs entirely, most have at least dropped them from all stable branches | 22:14 |
| melwitt | cc gmann ^ | 22:14 |
| gmann | yeah, it was updated to run on focal with this https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/834179 | 22:17 |
| gmann | agree on dropping it from stable branch but if we want to run then we can keep one version of it for stable branch running on bionic with py3.6 | 22:18 |
| gmann | if it is failing many other projects also like nova | 22:18 |
| gmann | fungi: clarkb melwitt ^^ you want me add stable branch version of l-c job with this ^^? | 22:19 |
| melwitt | thanks for the input gmann | 22:19 |
| clarkb | gmann: it sonds like three isn't much value in that according to fungi | 22:19 |
| melwitt | gmann: I am ok with dropping it | 22:20 |
| gmann | cool, that is even better | 22:20 |
| gmann | clarkb: melwitt fungi also on same note of l-c requirements and its testing, we will discuss it in TC PTG on Friday 13 UTC https://etherpad.opendev.org/p/tc-zed-ptg#L131 | 22:21 |
| clarkb | oh I should look at that schedule | 22:21 |
| melwitt | ah cool | 22:22 |
| gmann | as it is not well maintained and not tested as expected let's see if we can get agreement on this. | 22:22 |
| fungi | i finally found a chunk of time today to work out what ptg sessions i'll be in and when | 22:22 |
| gmann | great, I will add you in ping list just in case | 22:22 |
| clarkb | Looks like I'll have to try and have an early start next friday | 22:27 |
| fungi | 6am for you, yeah | 22:29 |
| clarkb | fungi: outside of the TC topics that overlap with us are there sessions in other ptg groups that should get our attention that you've seen? | 22:29 |
| fungi | for opendev collaboratory and tact sig stuff, not really. i basically put down all the qa, tc and release sessions | 22:29 |
| clarkb | ya 9am thursday which isn't too bda but 6am firday will be fun :) | 22:30 |
| fungi | for me, i'm also going to be in the security sig sessions, diversity & inclusion, barbican when they're talking about the consumers api (for context on image encryption), and i put a topic in for horizon to revisit the xstatic wrapper model and consider potential alternative approaches | 22:31 |
| fungi | other than those, i expect to just float when i'm free | 22:32 |
| fungi | probably put a session on in the background and multi-task unless something catches my interest | 22:32 |
| fungi | and try to keep on top of meetpad and ptgbot issues, if any | 22:32 |
| clarkb | ya I planned to basically try and keep on top of tooling issues at least | 22:33 |
| clarkb | looks like the qa schedule isn't as set in stone as the TC one. I'll just keep an ear out for things when awake | 22:34 |
| *** dasm|ruck is now known as dasm | 22:47 | |
| *** dasm is now known as dasm|ruck|off | 22:47 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!