16:00:24 <johnsom> #startmeeting Octavia
16:00:25 <opendevmeet> Meeting started Wed Jul  9 16:00:24 2025 UTC and is due to finish in 60 minutes.  The chair is johnsom. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:25 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:25 <opendevmeet> The meeting name has been set to 'octavia'
16:00:32 <johnsom> Hello everyone!
16:00:41 <gthiemonge> o/
16:01:35 <johnsom> #topic Announcements
16:01:52 <johnsom> We are fully in milestone 2 now.
16:02:10 <johnsom> Feature freeze is coming up the week of August 25th
16:03:24 <johnsom> Some release patches were created last night in preparation for that. I will go through those today
16:04:05 <gthiemonge> thanks johnsom
16:04:33 <johnsom> Any other announcements
16:04:57 <gthiemonge> no
16:05:00 <gthiemonge> yeah
16:05:05 <gthiemonge> one thing
16:05:26 <gthiemonge> Action Required: Renew Your OpenInfra Individual Membership by July 19
16:05:32 <gthiemonge> https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/7CITYVVCQDVJSCK65LOKTN25HWKAM237/
16:06:04 <gthiemonge> "Following OpenInfra's integration with the Linux Foundation, all individual members must complete a brief renewal process"
16:06:16 <johnsom> Ah, I have not seen that yet this morning, but I did get that email
16:08:59 <johnsom> So, yeah, I guess we should read that. If we have questions we can ask in the opendev channel
16:09:35 <gthiemonge> +1
16:09:36 <johnsom> Foundation folks hang out there. There is also the openinfra-board channel
16:10:06 <fungi> not opendev, this isn't an opendev thing, it's an openinfra foundation thing
16:10:34 <fungi> there is a #openinfra channel, though i've also been ferrying questions from the #openstack-tc channel to the staff who are in charge of this bit
16:11:02 <gthiemonge> ack
16:11:21 <fungi> you can also feel free to just ask me questions directly and i'll try to get answers
16:11:22 <johnsom> +1 thanks for clarifying. I didn't highlight openinfra as there seems to be no traffic there, so I wasn't sure it was being used anymore
16:11:39 <fungi> it doesn't get much traffic, no, but i keep an eye on it
16:12:19 <johnsom> +1
16:12:32 <johnsom> #topic Brief progress reports / bugs needing review
16:12:47 <johnsom> Moving on, I am mostly in review mode right now.
16:13:50 <johnsom> The UEFI patch adds ~500MB to the amphora image which is over the flavor size. I have commented on that patch we need to decide how to proceed. Currently our nova flavor is 2GB, but this change may push it to 3GB minimum.
16:14:05 <gthiemonge> no updates from me, I think I'll focus on reviews in the next weeks
16:14:23 <johnsom> Mostly I am concerned with the upgrade impact of the change. We need to add more to the release notes at a minimum.
16:15:14 <johnsom> Otherwise, I think our gates are stable now, so it's working through that lovely review backlog. grin
16:15:44 <johnsom> #topic Open Discussion
16:16:00 <johnsom> Ok, any other topics today? I'm guessing fungi has a topic.
16:16:05 <fungi> yeah
16:16:18 <fungi> i'll try to be quick, but there's a lot we dug into... for some background on openstack-wide metrics analysis see ildikov's most recent ml post from a few weeks ago:
16:16:25 <fungi> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/NTBNI7YIDCWBR6BTPEKVZIODWTVUIOXO/ BtG metrics analysis
16:16:33 <fungi> also could be worthwhile to revisit her previous post in that thread going over the contributor and maintainer survey results (and anyone who hasn't filled those out for epoxy, please see if you can find a few minutes to do that!)
16:16:41 <fungi> as a follow-up activity, we've started doing some team-specific analyses, focusing on teams that had multiple contributor and maintainer survey responses (already did nova, cinder and ironic in recent weeks, neutron and kolla up next)
16:16:47 <fungi> we're early in the process of analyzing these stats with a focus on improving the experience for maintainers and contributors, so for now this is probably a lot of stuff you already know, or at least confirming what you expected
16:16:56 <fungi> a big part of this is establishing a baseline so that we can better gauge whether future attempts at improving have any observable impact, but we aren't at the point where we have much in the way of guidance or recommendations yet
16:17:11 <fungi> we have plans to continue with this sort of surveying and metrics analysis over coming release cycles; as for the initial results...
16:17:18 <fungi> the contributor survey had 2 responses for octavia and both respondents had contributed for at least two years and were contributors to at least two other open source projects
16:17:29 <fungi> feedback varied widely (averages of 1.5-4.5 out of 5) with the highest score on "You receive actionable feedback from other reviewers" and lowest for "Changes you propose are reviewed in a timely manner"
16:17:40 <fungi> the top challenges reported were trouble with review attention and change of expectations and scope
16:17:49 <fungi> the maintainer survey also had 2 responses with slightly less spread between scores than the contributor survey (averaging 2-4) with contributor documentation receiving high marks while timely review of changes was still at the bottom
16:18:00 <fungi> contributing challenges reported by maintainers were similar to those from the contributor survey (review attention)
16:18:11 <fungi> top challenges with reviewing were lots of low-priority changes proposed, lack of reviewer familiarity with parts of the code a change is modifying, and lack of timely response from change owners to review comments and ci failures
16:18:18 <fungi> looking at metrics we gathered from gerrit for the past 5 development cycles, time to first review was comparatively fast suggesting changes are getting initial attention promptly
16:18:26 <fungi> however we saw a 3x difference between the mean (average) and median time to merge changes indicating a significant long tail of changes not getting movement, which backs up the survey responses; probably insufficient maintainer bandwidth?
16:18:38 <fungi> key takeaways were communication from maintainers on priorities/bandwidth could set clearer contributor expectations, and organizing sessions for maintainers to increase their familiarity with more of the codebase might be useful as well
16:18:43 <fungi> sorry, i know that's a pretty big info dump (i tried to pare it down as much as possible), and i'm happy to answer questions or take feedback either here in the meeting or any time after
16:19:33 <johnsom> "insufficient maintainer bandwidth" is a huge factor IMO
16:20:15 <johnsom> If you look at the review priority list linked in the channel description, it's huge.
16:20:44 <johnsom> Many of us have much more "downstream" distractions that are causing issues.
16:21:21 <fungi> also, again probably not surprising, but this seems to be a pretty common situation across openstack projects, octavia isn't alone in this
16:21:39 <johnsom> Yes, fully agree
16:22:07 <fungi> i'm hoping that as we continue to look into the team-specific results and get more brainstorming happening in teams, we'll be able to share solutions for improving the situation between them
16:22:40 <fungi> ideally ways to make things easier for maintainers, as well as improving the experience for contributors
16:23:01 <johnsom> On think I wish we had more of is +1/-1 reviews. There are many things that can be caught early and updated that end up being in core review comments.
16:23:49 <fungi> one idea we'd been kicking around is cut-n-adjust review response templates for common situations like low priority or needing the change owner to revisit ci job results
16:24:27 <fungi> i know i often find myself re-typing the same messages over and over, and even copying from some of my old responses to save myself some time
16:24:45 <johnsom> Yeah, including release notes in a change is a common comment.
16:25:15 <fungi> but yeah, encouraging non-maintainers to pitch in with more reviewing to help polish changes has come up in a few discussions already as helpful
16:25:17 <johnsom> They are super useful, but get overlooked
16:26:23 <fungi> the main thing to be mindful with that is it's a double-edged sword, infrequent contributors can't easily distinguish between a -1 from a maintainer and other reviewers, so may get a lot of early suggestions to fix up a change that ultimately isn't going to merge because it's not a good fit for the project
16:27:05 <fungi> so early feedback from maintainers when a change isn't worth investing more time improving is pretty critical in situations that could otherwise leave a bad taste in a contributor's mouth
16:27:24 <johnsom> We don't get a lot of those. I would say the vast majority of patches are valuable to the project and will eventually merge, even if another contributor needs to adopt the patch
16:27:55 <fungi> yeah, then it definitely helps to prioritize non-maintainer reviews in that case
16:28:44 <johnsom> We don't do "auto-abandons" in the octavia projects because for the most part, they are useful patches.
16:29:40 <fungi> but also be mindful that, as with a lot of things, an occasional bad experience tends to undo/outweigh lots of good experiences, unfortunately. we've observed a lot of that in getting feedback from contributors to various projects, one bad exchange often sours them on wanting to be involved in the future
16:30:28 <johnsom> Yeah, review/merge velocity is a problem. No denying that.
16:31:24 <fungi> even (and usually in fact) unintentional cases where they feel like they're being ignored but it's really just that there's nobody around to give them feedback or their change got accidentally overlooked and fell through the cracks, so figuring out how to better communicate and increase visibility is something we need to figure out better solutions for i guess
16:32:13 <fungi> anyway, i didn't have anything else to present on this, but open to feedback and happy to answer questions, either now or later as they come up
16:32:27 <johnsom> Yeah, things like recently we had to shift focus for a few weeks to fix gate failures (setuptools, etc.) had an impact
16:32:56 <johnsom> Thank you fungi for looking into this and helping us build a stronger community.
16:33:04 <johnsom> Any other questions/comments?
16:33:40 <gthiemonge> nop, thanks fungi for bringing this topic
16:33:46 <fungi> thanks for listening!
16:34:15 <johnsom> Ok, any other open agenda items today?
16:34:25 <gthiemonge> not from me
16:34:52 <johnsom> Ok, thanks for joining and have a good week!
16:35:03 <johnsom> #endmeeting