14:00:32 <ttx> #startmeeting releaseteam
14:00:32 <opendevmeet> Meeting started Fri Sep 12 14:00:32 2025 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:32 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:32 <opendevmeet> The meeting name has been set to 'releaseteam'
14:00:41 <ttx> Ping list: release-team elod
14:00:49 <ttx> Our agenda for today is:
14:01:02 <ttx> #link https://etherpad.opendev.org/p/flamingo-relmgt-tracking#L369
14:01:08 <elodilles> o/
14:01:36 <ttx> #topic Review task completion
14:01:53 <ttx> - Process any remaining library branching exception. (all)
14:02:06 <ttx> #link https://review.opendev.org/q/topic:flamingo-stable-branches+is:open
14:02:36 <ttx> https://review.opendev.org/c/openstack/releases/+/959165 is still open
14:02:42 <frickler> \m/
14:03:16 <elodilles> yes, i've pinged them on the other patch:
14:03:35 <elodilles> https://review.opendev.org/c/openstack/releases/+/957695
14:03:46 <elodilles> as this is the one that blocks the other one ^^^
14:04:13 <ttx> At this point we can drop it from release, or give it a bit more time
14:04:30 <ttx> I'd be fine giving it a bit more time before pulling it out
14:04:37 <elodilles> i think next week we have to do something :/
14:04:39 <ttx> since not much depends on it
14:04:51 <ttx> yeah, by next week we'll have to pull it
14:05:09 <elodilles> i think we might need to release it broken, as is
14:05:19 <ttx> or release broken yes
14:05:27 <elodilles> if it won't be fixed by then
14:05:37 <ttx> ok let's push back to next week for final call
14:06:02 <elodilles> +1
14:06:33 <ttx> - generate release requests for all deliverables that do not have a suitable candidate yet: (elod)
14:07:36 <elodilles> some were proposed already, i've copied them in the list; and i've generated release patches for the rest
14:08:13 <elodilles> bifrost: https://review.opendev.org/c/openstack/releases/+/960047
14:08:13 <elodilles> ironic-inspector: https://review.opendev.org/c/openstack/releases/+/959565
14:08:17 <elodilles> ironic-prometheus-exporter: https://review.opendev.org/c/openstack/releases/+/960048
14:08:20 <elodilles> ironic-ui: https://review.opendev.org/c/openstack/releases/+/960049
14:08:22 <elodilles> networking-generic-switch: https://review.opendev.org/c/openstack/releases/+/959567
14:08:25 <elodilles> swift: https://review.opendev.org/c/openstack/releases/+/959567
14:08:29 <ttx> Just approved bifrost
14:08:57 <elodilles> note: there are some further ironic release patch
14:09:36 <elodilles> i had a couple of -1's there for asking the teams to double check the version bumps
14:09:48 <elodilles> but no answer yet
14:10:07 <ttx> hmm
14:11:01 <ttx> maybe you can +2/PTL+1 them with a note that they will be merged asap unless we get a -1... and I will approve them on Monday if no response
14:11:12 <ttx> together wit those RC1
14:11:38 <elodilles> ACK, I'll do that after the meeting then
14:12:42 <ttx> I'll add a task for that
14:12:53 <ttx> good activity for my Monday train ride
14:13:44 <elodilles> :]
14:14:07 <ttx> - investigate https://bugs.launchpad.net/octavia/+bug/2121578 (elod)
14:14:33 <elodilles> yeah, i had other things to investigate so had no time to deal with this :/
14:14:54 <ttx> Let's push it forward to next week
14:14:55 <elodilles> on the other hand i haven't heard complains about this issue
14:15:22 <elodilles> so i don't know how serious this is
14:15:35 <ttx> - heat-translator release patch - broken gate (elod)
14:15:47 <elodilles> (let's hope it won't blow up in our face at release day :S fingers crossed :X)
14:15:52 <ttx> I think we covered it
14:15:55 <elodilles> yes
14:16:17 <ttx> - Send weekly email. (ttx)
14:16:20 <ttx> will do in a minute
14:16:35 <ttx> #topic Assign R-2 week tasks
14:16:56 <ttx> All set! Thanks elodilles
14:17:07 <elodilles> +1
14:17:12 <ttx> #topic Review weekly countdown email
14:17:21 <ttx> #link https://etherpad.opendev.org/p/relmgmt-weekly-emails
14:18:04 * elodilles clicks
14:19:47 <opendevreview> Merged openstack/releases master: Release bifrost for 2025.2 Flamingo  https://review.opendev.org/c/openstack/releases/+/960047
14:20:02 <elodilles> mail LGTM
14:20:29 <elodilles> ttx: sorry, i think i missed to mention some thing at one task
14:21:05 <elodilles> i mean two points:
14:21:12 <elodilles> 1) Branch devstack-plugin-* deliverables:
14:21:39 <elodilles> #link https://review.opendev.org/q/topic:devstack-plugin-flamingo
14:21:59 <ttx> ok will send in a few
14:22:08 <elodilles> ttx: ++
14:22:12 <ttx> this one needs attention from gmaan
14:22:16 <elodilles> so, Ghanshyam is on PTO
14:22:27 <ttx> (https://review.opendev.org/c/openstack/releases/+/960039)
14:22:45 <elodilles> he'll be back after 2025.2 Flamingo release :)
14:23:07 <elodilles> so probably we should review and process this patch without his approve
14:23:12 <ttx> looks like we'll have to make a call then
14:23:21 <ttx> will do
14:23:51 <elodilles> yes, frickler ttx , please do review and approve if it looks OK
14:24:13 <ttx> done
14:24:19 <ttx> a bit of a no-brainer
14:24:26 <elodilles> 2) cycle-with-rc that are not trailing deliverables and that have not done a RC1 yet
14:24:39 <elodilles> #link https://review.opendev.org/q/topic:flamingo-rc1-deadline
14:24:58 <elodilles> i've gone through the patches without response from team
14:25:06 <ttx> yeah for those I will approve them on Monday unless they get -1ed
14:25:12 <elodilles> and updated the hashes where it was needed
14:25:17 <elodilles> ttx: thanks o/
14:25:28 <elodilles> sounds good o/
14:25:46 <ttx> #topic Open Discussion
14:25:51 <fungi> summarizing the discussion with zigo yesterday: temporary disappearance of notes for the upcoming release between stable branch creation at and merging the auto-proposed change adding it to the index confuses some downstream consumers (e.g. distro package maintainers)
14:25:52 <elodilles> probably majority were approved and/or updated by teams :)
14:25:55 <fungi> yesterday it first got raised to the opendev sysadmins as a potential content hosting problem or publication job failure, before we tracked it back to openstack's release process working as designed
14:25:59 <fungi> i can't help but wonder if there's a way to shorten or eliminate that gap, since even if maintainers are around and approve those changes quickly, there is still some period of time where the upcoming changes have vanished from the site entirely
14:26:08 <fungi> food for thought (maybe a good ptg topic)
14:26:30 <ttx> +1
14:26:34 <ttx> I'll add it
14:26:45 <fungi> thanks!
14:27:00 <fungi> probably needs some brainstorming of solutions
14:27:20 <elodilles> +1
14:28:02 <ttx> elodilles: want to talk pyroute2?
14:28:15 <elodilles> ttx: ah, yes,
14:28:26 <elodilles> fortunately it got solved meanwhile :)
14:28:28 <elodilles> so
14:28:36 <ttx> ok so maybe no need :)
14:28:41 <elodilles> pyroute2 had a bug
14:28:46 <fungi> well, the immediate emergency got solved
14:28:57 <elodilles> and needed some capping
14:29:25 <elodilles> (elod) pyroute2 bug: https://meetings.opendev.org/irclogs/%23openstack-release/%23openstack-release.2025-09-08.log.html#openstack-release.2025-09-08.log.html#t2025-09-08T12:52:46
14:29:31 <elodilles> requirements capping to pyroute2==0.8.1 was accepted - https://review.opendev.org/c/openstack/requirements/+/960366 / https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/SQLZ6ZOPUF5EHJ722ZQGPT7PSPV4PCNE/
14:29:44 <elodilles> (just to have it here, for ship's log o:))
14:30:28 <elodilles> fungi: indeed, we might need to bump pyroute2 version on stable/2025.2 later on, if that is possible
14:30:30 <fungi> also based on discussion, there's not yet a fix upstream in the library
14:30:35 <elodilles> as far as i understood
14:30:44 <elodilles> fungi: exactly
14:31:00 <elodilles> no the current fixes proved to be not enough
14:31:05 <fungi> increasing constraints for non-openstack-controlled projects in a stable branch is generally something we don't do
14:31:07 <elodilles> no, the current fixes proved to be not enough
14:31:24 <elodilles> so yes, teams are waiting for further fixes
14:31:36 <elodilles> fungi: exactly
14:31:36 <ttx> ok, anything else?
14:32:07 <fungi> probably we should stick with pyroute2 0.8.1 in our constraints list for the lifetime of the stable/2025.2 branches
14:32:08 <elodilles> fungi: so i don't know whether there could be given some exception for pyroute2 later, or we have to stick to the capping
14:32:40 <elodilles> fungi: i see. i did not remember the requirements team's policy on this
14:32:43 <fungi> i'm not aware of having previously made exceptions for cases like that
14:32:56 <elodilles> fungi: i thought some exceptional cases we could bump versions
14:33:09 <frickler> usually exceptions are made for CVEs
14:33:26 <frickler> but I'm not sure how strict we need to be on this
14:33:35 <fungi> we freeze constraints (except for stable point releases of our own software) in order to stabilize testing, and we don't make any security guarantees about versions of software in the frozen constraints list
14:33:38 <elodilles> fungi: i think some rare cases when CI had issues we did that
14:33:54 <elodilles> and what frickler says, when there were CVEs
14:34:17 <fungi> if we have made exceptions for vulnerabilities solely in a dependency, that seems like a poor choice and sets incorrect expectations for users of those files
14:34:46 <elodilles> but i'm not fully aware of requirments processes, so i was just wondering
14:35:02 <fungi> we cannot, as a project, be expected to track vulnerabilities for all dependencies in our maintained stable branches, and changing versions of dependencies destabilizes testing
14:35:21 <elodilles> hmmm. i see
14:35:46 <elodilles> so probably those things needs to be handled downstream by vendors
14:35:59 <fungi> and we've repeatedly discouraged downstream consumers from using the frozen stable constraints lists exactly because we don't track vulnerabilities in those dependencies
14:36:22 <frickler> except for kolla and osa there are many deployments not involving downstream vendors, so they plainly use u-c
14:36:42 <elodilles> O.o
14:36:53 <fungi> and it's their responsibility to alter the versions of dependencies or backport patches to them in those cases
14:37:27 <fungi> i really wish we had a better way to let them know, because i suspect in many cases they don't even realize that's the case
14:38:31 <frickler> we (kolla) also in vain try to convince people not to use the images we produce in production
14:38:57 <frickler> (which also are produced using upstream stable u-c)
14:39:21 <ttx> anything more we can do at this stage?
14:39:33 <fungi> not that i know of
14:40:08 <elodilles> thanks for the discussion! good to know this details.
14:40:23 <ttx> alright, anything else to discuss?
14:40:24 <opendevreview> sean mooney proposed openstack/releases master: Release watcher RC1 for 2025.2 Flamingo  https://review.opendev.org/c/openstack/releases/+/960123
14:40:31 <elodilles> ttx: one more thing
14:40:44 <elodilles> #link https://review.opendev.org/q/topic:fix-release-requirements-update-constraints
14:41:24 <elodilles> some upper constraints bumps were missed because people started to use python_version>=3.10 in upper-constraints.txt
14:41:36 <frickler> ah, right, I need to review these, thx for the reminder
14:41:56 <elodilles> so i've created the above fixes to not miss them in the future ^^^
14:42:05 <frickler> though also with 3.9 gone, we could probably drop most of the >=3.10 stanzas
14:42:06 <ttx> yeah we need project-config core to approve those
14:42:34 <elodilles> well, one fix is for cases when teams replaces setup.py with pyproject.toml
14:42:57 <frickler> yes, that one is needed anyway indeed
14:43:00 <elodilles> frickler: i didn't even know we have py3.9 lines there
14:43:21 <elodilles> frickler: i was surprised, as flamingo only supports py310+
14:43:51 <frickler> they were re-introduced when people still wanted to run centos9 jobs for some special cases
14:44:21 <elodilles> but centos9 is also not supported runtime, am i right? o:)
14:44:26 <ttx> I need to run, so I'll close the meeting, but feel free to continue the open discussion in-channel
14:44:32 <fungi> looking
14:44:47 <elodilles> https://governance.openstack.org/tc/reference/runtimes/2025.2.html
14:45:01 <elodilles> ttx: okay, thanks o/
14:45:22 <frickler> it is not, but some projects were slow in updating stuff, in particular since centos10 wasn't available for quite some time during this cycle
14:45:35 <ttx> #endmeeting