| *** rlandy|bbl is now known as rlandy|out | 00:45 | |
| *** ykarel_ is now known as ykarel | 04:51 | |
| *** ysandeep|out is now known as ysandeep|rover | 04:52 | |
| *** ysandeep|rover is now known as ysandeep|rover|brb | 05:51 | |
| *** ysandeep|rover|brb is now known as ysandeep|rover | 05:56 | |
| *** akekane_ is now known as abhishekk | 06:46 | |
| *** jpena|off is now known as jpena | 07:36 | |
| *** ysandeep|rover is now known as ysandeep|rover|lunch | 09:31 | |
| *** ysandeep|rover|lunch is now known as ysandeep|rover | 10:09 | |
| *** rlandy|out is now known as rlandy | 10:27 | |
| *** dviroel|out is now known as dviroel | 11:26 | |
| *** rlandy is now known as rlandy|mtg | 11:26 | |
| *** ysandeep|rover is now known as ysandeep|rover|afk | 11:33 | |
| *** rlandy|mtg is now known as rlandy | 12:02 | |
| *** ysandeep|rover|afk is now known as ysandeep|rover | 12:26 | |
| *** dasm|off is now known as dasm | 13:21 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Add workaround for custom builds to be send https://review.opendev.org/c/openstack/ci-log-processing/+/834119 | 14:58 |
|---|---|---|
| *** dviroel is now known as dviroel|lunch | 15:26 | |
| *** ysandeep|rover is now known as ysandeep|out | 15:29 | |
| *** dviroel|lunch is now known as dviroel | 16:25 | |
| *** jpena is now known as jpena|off | 16:31 | |
| *** timburke__ is now known as timburke | 17:28 | |
| opendevreview | Ghanshyam proposed openstack/openstack-zuul-jobs master: Add openstack-tox-functional-py39-fips job https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/842593 | 18:04 |
| gmann | clarkb: fungi: can you check this fips py39 job, needed for glance for fips testing https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/842593 | 18:25 |
| fungi | clarkb: "no Zuul jobs will run against that project branch" isn't entirely true. per discussion in matrix, jobs applied from the config repo would likely still run, leading to changes merging with minimal testing | 19:13 |
| fungi | which is probably much worse than just not being able to merge anything on the branch | 19:14 |
| clarkb | oh I missed that in the zuul conversation | 19:25 |
| clarkb | feel free to respond to clarify | 19:25 |
| prometheanfire | looks like we are getting more errors due to not generating constraints for py39/310 https://review.opendev.org/842610 | 20:01 |
| clarkb | prometheanfire: reading your comments implise there is something the infra team needs to do? THis is news to me | 20:08 |
| clarkb | can you point out whatever that is? | 20:08 |
| clarkb | as a side note contraints should not be used to control which versions are valid | 20:11 |
| clarkb | requirements do that. Constraints is a pre calculated set of valid versions based on requirements to ensure consistent behaviors | 20:11 |
| prometheanfire | well, there have been a couple of reviews to add more python versions to generate-constraints, the problem is that there wasn't an image that had py36/38/39/310 available iirc (or just 38,39,310) | 20:12 |
| prometheanfire | right, that review was editing the wrong file, but we don't have constraints for versions of python that master is said to support | 20:12 |
| prometheanfire | https://review.opendev.org/c/openstack/project-config/+/828909 | 20:14 |
| prometheanfire | https://review.opendev.org/c/openstack/project-config/+/837815 | 20:14 |
| clarkb | right but I don't understand why that would be an infra problem. That distro doesn't exist so we can't provide it | 20:14 |
| prometheanfire | ok, question 1: can we drop py36 (I think the answer is yes here) | 20:15 |
| clarkb | thats up to the TC and I think they said yes | 20:15 |
| prometheanfire | right | 20:16 |
| prometheanfire | question 2 is if there's a distro that supports 38/39/310 or 38/39 or 38/310 (I think that's order of preference?) | 20:16 |
| clarkb | no, yes, no | 20:17 |
| clarkb | ubuntu focal has packages for both 38 and 39 | 20:17 |
| prometheanfire | ok, so that's what I think we can change to, at least until when/if something better pops up | 20:18 |
| prometheanfire | updated https://review.opendev.org/c/openstack/project-config/+/837815 and abandoned the earlier review | 20:21 |
| fungi | last time this came up, i suggested separate jobs for each python version, just like how we use those python versions in tests | 20:24 |
| fungi | trying to have one job on one system generate constraints lists for multiple python versions isn't going to be the easy path | 20:25 |
| fungi | it worked when all we were trying to do was build constraints for python 2.7 and 3.5 at the same time | 20:26 |
| clarkb | gmann: ^ if you want to ack that change I think we can land it | 20:27 |
| prometheanfire | fungi: that's probably a better way forward in the long run | 20:27 |
| clarkb | for Zed 3.10 support isn't planned is it? Maybe it is via centos 9 stream? | 20:27 |
| prometheanfire | even totally separate constraints for each python target could be nice | 20:27 |
| clarkb | I know that it isn't planned via jammy at least | 20:27 |
| fungi | clarkb: i believe that was the intent (using jammy once available to run non-voting py310 jobs) | 20:31 |
| clarkb | well I suggested that and got a bit of pushback. It seemed they didn't want to bother until the next cycle. I mean I htink that is a good idea btu we'll see | 20:33 |
| prometheanfire | imo, as long as we keep one of the 'old' pythons around each cycle it's good, (going from 36,38 before to 38,39 now, and 39,310 next cycle) | 20:34 |
| *** dviroel is now known as dviroel|out | 20:45 | |
| *** dasm is now known as dasm|off | 21:26 | |
| gmann | clarkb: checking | 22:18 |
| gmann | clarkb: prometheanfire +1 | 22:19 |
| opendevreview | Merged openstack/project-config master: update generate constraints to py38,39 https://review.opendev.org/c/openstack/project-config/+/837815 | 22:35 |
| prometheanfire | woot | 23:12 |
| opendevreview | Ghanshyam proposed openstack/openstack-zuul-jobs master: Add openstack-tox-functional-py38-fips job https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/841368 | 23:41 |
| opendevreview | Ghanshyam proposed openstack/openstack-zuul-jobs master: Move master template from py36 to py39 https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/842646 | 23:50 |
| gmann | fungi: clarkb : py39 fips did not work so adding py38 fips job which worked fine https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/841368 | 23:51 |
| clarkb | gmann: which distro has py38 and fips? | 23:52 |
| clarkb | oh using old centos | 23:52 |
| gmann | clarkb: I tested on c8s with py38 https://zuul.opendev.org/t/openstack/build/3e19855c81ff447ea1760c949c9e0c52 | 23:53 |
| gmann | yeah, just moving py36 to py38 but some time we should move them to py39 or ubuntu based if fips image available | 23:53 |
| clarkb | ubuntu based fips isn't an image issue. It is a licensing issue | 23:54 |
| clarkb | just to be clear | 23:54 |
| fungi | similar to why we don't have rhel nodes | 23:55 |
| gmann | yeah that one | 23:55 |
| gmann | bcz c9s also going more unstable now a days | 23:56 |
| clarkb | yes I even wrote an email about it today :) do we know why it cs9 fails? | 23:56 |
| clarkb | I'm really beginning to think we need to treat stream more like fedora and rely on stable distros for long term testing that we intend to be reliable | 23:57 |
| gmann | libvirt one and detach volume, ssh timeout are main issues happening now | 23:57 |
| clarkb | basically its good early signal. But its bad long term stable test system | 23:57 |
| gmann | yeah, that was my intent o make it voting so that we notice its stability and decide if we want to test such unstable distro as testing runtime | 23:57 |
| gmann | and if none help to fix these issue and it stay unstable then drop it from next testing runtime | 23:58 |
| clarkb | I'm looking at our centos 8 stream mirror and don't see python3.8 there. Where is python3.8 coming from? | 23:59 |
| * clarkb looks at the nova log | 23:59 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!