| *** gibi_off is now known as gibi | 07:08 | |
| opendevreview | Maxim Sava proposed openstack/tempest master: Add test cases for image caching and fix image base class https://review.opendev.org/c/openstack/tempest/+/957612 | 08:38 |
|---|---|---|
| opendevreview | Merged openstack/devstack stable/2026.1: Fix noble OVN source Zuul parent https://review.opendev.org/c/openstack/devstack/+/989553 | 10:11 |
| opendevreview | Merged openstack/devstack master: Remove leftover of glance-search https://review.opendev.org/c/openstack/devstack/+/989867 | 10:25 |
| opendevreview | Elod Illes proposed openstack/grenade stable/2025.2: Fix skip level grenade for EOL'd branch https://review.opendev.org/c/openstack/grenade/+/989786 | 10:42 |
| *** haleyb|out is now known as haleyb | 12:53 | |
| opendevreview | Elod Illes proposed openstack/tempest master: WIP: Lower constrain pbr to fix grenade jobs https://review.opendev.org/c/openstack/tempest/+/990123 | 14:44 |
| opendevreview | Elod Illes proposed openstack/grenade stable/2025.2: Fix skip level grenade for EOL'd branch https://review.opendev.org/c/openstack/grenade/+/989786 | 14:45 |
| elodilles | gmaan stephenfin : what do you think about this workaround patch? is it appropriate to merge (at least being there for the time while we have grenade jobs that has ubuntu-jammy as base nodes)? https://review.opendev.org/c/openstack/tempest/+/990123 | 16:57 |
| elodilles | this is needed to fix the gate of stable/2025.2 (skip-level grenade job) and stable/2025.1 (grenade job) now that the mentioned change has landed on tempest | 16:58 |
| elodilles | btw, it can be seen working with this patch: https://review.opendev.org/c/openstack/grenade/+/989786 | 16:58 |
| gmaan | elodilles: so skip level grenade job on 2025.2 and grenade job on 2025.1 doing upgrade from 2024.2 right? | 17:01 |
| gmaan | and 2024.2 is EOL | 17:01 |
| elodilles | exactly | 17:01 |
| gmaan | should not we just remove the jobs | 17:01 |
| elodilles | yes, that is another option, though i'd give some months for grenade to work, until all those who chose (badly) bobcat as their deployment, to have a working upgrade path | 17:02 |
| gmaan | yeah but upgrade testing from unmaintained releases are ok but I am not sure testing upgrade form EOL release is something we can do or continue to keep it green | 17:03 |
| elodilles | definitely it won't work for ages o:) i'd say let's keep it green as long as it completely breaks OR just a couple of months (less than 3-4, let's say) | 17:04 |
| elodilles | (currently it works by checking out 2024.2-eol tags. so we surely reach a point when things will break) | 17:05 |
| gmaan | yeah, it is difficult to keep effort on those or have workaround especially to keep them green | 17:06 |
| gmaan | my vote is to remove them but we can wait for feedback from frickler or dansmith (this one https://review.opendev.org/c/openstack/grenade/+/989786) | 17:07 |
| elodilles | yepp. but at least we see now for a couple of months if we backport something that breaks grenade | 17:08 |
| dansmith | dropping EOL makes sense to me because they are E...OL | 17:08 |
| elodilles | gmaan: ACK, thanks | 17:08 |
| frickler | I'm also strongly in favor of dropping testing involving eol branches | 17:08 |
| elodilles | don't get me wrong: me too :) | 17:09 |
| elodilles | i'm just saying maybe we are too early for that, as 2024.2 bobcat went EOL some weeks ago only o:) | 17:09 |
| frickler | so we are some weeks too late IMHO | 17:09 |
| gmaan | i would have delayed the EOL in that case if needed but if it is EOL means EOL :) | 17:10 |
| frickler | also wasn't 2023.2 bobcat? | 17:10 |
| elodilles | frickler: things that now break are because the branches were deleted o:) | 17:10 |
| opendevreview | Takashi Kajinami proposed openstack/devstack master: Drop unused old utility for Python 3 transition https://review.opendev.org/c/openstack/devstack/+/990145 | 17:10 |
| elodilles | frickler: yes, you are right. 2024.2 dalmatian went to EOL recently o:) | 17:10 |
| elodilles | my bad :) | 17:11 |
| elodilles | anyway, now nova, cinder, grenade, and a couple of stable gates are blocked by this | 17:11 |
| sean-k-mooney | so 2024.2 never got the fixzes in teh requriemet repo | 17:11 |
| sean-k-mooney | to make it work with newer setup tools | 17:11 |
| sean-k-mooney | and it never got the pbr bump either | 17:12 |
| elodilles | (stable/2025.2 by skip-level, and stable/2025.1 by simple grenade jobs) | 17:12 |
| sean-k-mooney | so installing it to then upgrade form it is not trivial | 17:12 |
| elodilles | sean-k-mooney: it got | 17:12 |
| gmaan | elodilles: true, IMO, correct fix for gate is to remove those jobs | 17:12 |
| elodilles | sean-k-mooney: this is a new issue, introduced by tempest patch yesterday | 17:13 |
| sean-k-mooney | https://github.com/openstack/requirements/blob/2024.2-eol/openstack_requirements/cmds/check_py2.py#L17 | 17:13 |
| sean-k-mooney | the requiremtns repo on 2024.2 is sitll driectly using pkg_resouces | 17:14 |
| sean-k-mooney | we stop backportin gthe remvoe of that in 2025.1 | 17:14 |
| sean-k-mooney | well that might be a bad example | 17:15 |
| elodilles | sean-k-mooney: yes, but all the other necessary workarounds were backported | 17:15 |
| sean-k-mooney | but it is true fo some of the other fileees | 17:15 |
| sean-k-mooney | https://github.com/openstack/requirements/commit/7b62011174e030d696d7ca42eb11ad82f2b19a3d | 17:16 |
| sean-k-mooney | is in 2025.1 | 17:16 |
| sean-k-mooney | but not in the older release so the requireem check jobs fail on older branches | 17:16 |
| elodilles | anyway, if everyone says remove the jobs, i'm not fighting it to keep for some time, just wanted to get your opinions, and was thinking to help folks who use 2024.2 dalmatian still as their deployments | 17:16 |
| sean-k-mooney | well the .2 release are not eligable for exetned maintance | 17:17 |
| elodilles | (they definitely chose badly, and hopefully they'll learn to either upgrade quickly from non-SLURP releases, or ignore them at all o:)) | 17:17 |
| sean-k-mooney | you coudl have the josb test form 2024.1 but i dont think we shoudl have any greande jobs that upgrade form somethin that is not stable/* | 17:17 |
| elodilles | sean-k-mooney: i disagree with that o:) of course, it's a best effort, and in documentation we don't maintain it, but still it gives value. if it blocks a gate and no-one is stepping up fixing it, then sure thing: remove it. | 17:21 |
| elodilles | s/in documentation/in documentation we say/ | 17:21 |
| sean-k-mooney | it also has a cost | 17:21 |
| sean-k-mooney | your energy in maintianin an unmainined branch | 17:21 |
| sean-k-mooney | ci cost in maintinin hte mirros and images for longer | 17:22 |
| sean-k-mooney | and in the job execution | 17:22 |
| elodilles | it has. but i don't think it has noticable cost. | 17:22 |
| sean-k-mooney | its deffintly ofsett by the retirement of tripleo :) | 17:22 |
| elodilles | tripleo was a beast :) | 17:23 |
| sean-k-mooney | but if we were cacisty constraitn tha the choce was betwen reducitg coverage on teh matser/stable branches | 17:23 |
| sean-k-mooney | or nuking the unmainteed jobs then the choce would be clear | 17:23 |
| sean-k-mooney | for example the [infra] AARCH64/ARM64 OpenDev CI capacity thread | 17:24 |
| sean-k-mooney | if we have not turned off the arm tox jobs on unmaintained | 17:24 |
| sean-k-mooney | we proably shoudl | 17:24 |
| elodilles | sean-k-mooney: with that i agree | 17:24 |
| elodilles | it's definitely less relevant than a grenade job, imo | 17:25 |
| elodilles | gmaan: when you have time, can you take a quick look at this patch? this would also prevent some blocking issues on stable gates in the future if it gets merged: https://review.opendev.org/c/openstack/devstack/+/989546 | 17:30 |
| gmaan | elodilles: thanks, +2. | 17:32 |
| elodilles | thanks too o/ | 17:32 |
| opendevreview | Elod Illes proposed openstack/tempest master: Remove grenade-skip-level from stable/2025.2 https://review.opendev.org/c/openstack/tempest/+/990152 | 18:03 |
| opendevreview | Elod Illes proposed openstack/tempest master: Remove grenade-skip-level from stable/2025.2 https://review.opendev.org/c/openstack/tempest/+/990152 | 18:04 |
| opendevreview | Fernando Ferraz proposed openstack/devstack-plugin-ceph master: Fix cephadm fail to find lvm device when creating OSDs in Ubuntu Noble https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/990214 | 22:25 |
| opendevreview | Fernando Ferraz proposed openstack/devstack-plugin-ceph master: Fix cephadm fail to find lvm device when creating OSDs in Ubuntu Noble https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/990214 | 22:27 |
| opendevreview | Fernando Ferraz proposed openstack/devstack-plugin-ceph master: Fix cephadm fail to find lvm device when creating OSDs in Ubuntu Noble https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/990214 | 22:36 |
| opendevreview | Fernando Ferraz proposed openstack/devstack-plugin-ceph master: Fix cephadm fail to find lvm device when creating OSDs in Ubuntu Noble https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/990214 | 22:36 |
| opendevreview | Fernando Ferraz proposed openstack/devstack-plugin-ceph master: Fix cephadm fail to find lvm device when creating OSDs in Ubuntu Noble https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/990214 | 22:39 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!