15:00:08 <iurygregory> #startmeeting ironic 15:00:08 <opendevmeet> Meeting started Mon Feb 2 15:00:08 2026 UTC and is due to finish in 60 minutes. The chair is iurygregory. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:08 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:08 <opendevmeet> The meeting name has been set to 'ironic' 15:00:15 <alegacy> o/ 15:00:16 <dtantsur> o/ 15:00:20 <iurygregory> o/ 15:00:31 <iurygregory> lets see if we have quorum =) 15:00:45 <TheJulia> o/ 15:01:13 <TheJulia> We clearly need a fancy coffee machine for the IRC channel 15:01:26 <iurygregory> fair enough =D 15:03:36 <iurygregory> lets wait till 12:05 to see if more people will join 15:03:37 * TheJulia hears crickets this morning 15:03:45 * fungi is lurking as well 15:03:54 * TheJulia lurks to the coffee machine 15:04:02 <cid> o/ 15:04:19 <iurygregory> ok, we have about 6 ppl since we have fungi :D hehe 15:04:45 <iurygregory> Welcome to our first meeting of Feb \o/ (January is over, finally)! 15:05:00 <iurygregory> #topic Announcements / Reminders 15:05:15 <iurygregory> #info Standing reminder to review patches tagged ironic-week-prio and to hashtag any patches ready for review with ironic-week-prio: https://tinyurl.com/ironic-weekly-prio-dash 15:05:19 <opendevreview> Jad Haj Yahya proposed openstack/ironic master: Update heartbeat on receiving inspection data https://review.opendev.org/c/openstack/ironic/+/975397 15:06:29 <iurygregory> #info we are at the R-8 of Gazpacho Release Schedule in case you have something for oslo or openstackclient their FF is next week 15:07:39 <opendevreview> Jad Haj Yahya proposed openstack/ironic master: Update heartbeat on receiving inspection data https://review.opendev.org/c/openstack/ironic/+/975397 15:07:47 <iurygregory> #info Ironic will be participating in the next PTG, we just created the etherpad https://etherpad.opendev.org/p/ironic-ptg-2026.2 and filled the form 15:08:07 <iurygregory> Anyone has something to add to announcements/reminders? 15:08:53 <dtantsur> A small FYI: I'll be out the next 2 weeks 15:08:57 <TheJulia> I will be out of country next week, so don't expect me to attend the weekly meeting next week or be available for reviews 15:09:30 <iurygregory> ack, tks! 15:09:44 <iurygregory> moving on 15:09:46 <iurygregory> #topic Working Group Updates 15:10:03 <iurygregory> #info Standalone networking 15:10:11 <iurygregory> #link https://etherpad.opendev.org/p/ironic-standalone-networking 15:10:13 <alegacy> Had a flurry of comments on my patches last week. 15:10:23 <alegacy> I've updated them accordingly 15:10:35 <alegacy> waiting for approvals to move them forward! 15:10:56 <TheJulia> I'll try to review them in the next couple of days, effectively my week is over around wednesday due to travel 15:11:19 <alegacy> k, thanks 15:11:47 <cid> I've been lacking a little in terms of reviews. I expect to do more this week, with particular focus on TBN and standalone networking. 15:12:00 <iurygregory> awesome, tks everyone! 15:12:16 <iurygregory> next one 15:12:28 <iurygregory> #info Async IO 15:12:31 <iurygregory> #link https://etherpad.opendev.org/p/ironic-asyncio 15:12:46 <TheJulia> I didn't have time to review that last week, been heads down on vxlan 15:12:51 <iurygregory> Anything you would like to share dtantsur ? 15:13:14 <iurygregory> We have a spec up for review for it 15:13:17 <iurygregory> #link https://review.opendev.org/c/openstack/ironic-specs/+/972754 15:15:22 * TheJulia is guessing unstable connection 15:15:27 <opendevreview> Julia Kreger proposed openstack/networking-baremetal master: l2vni baremetal mech driver https://review.opendev.org/c/openstack/networking-baremetal/+/973889 15:15:27 <opendevreview> Julia Kreger proposed openstack/networking-baremetal master: Trunk port reconciliation for L2VNI attachments https://review.opendev.org/c/openstack/networking-baremetal/+/974619 15:15:42 <iurygregory> please review if you have interest in the spec :D 15:15:44 <iurygregory> moving on 15:15:57 <iurygregory> #info VXLAN Networking 15:15:59 <iurygregory> #link https://etherpad.opendev.org/p/ironic-vxlan 15:16:16 <iurygregory> any updates for this wg? 15:16:36 <dtantsur> yeah, sorry, the status for async/await is like Iury said: please review the spec 15:17:16 <TheJulia> I was heads down on this last week. The tl;dr is given where we are at in the cycle, my hope is to begin landing some of the patches since we're approaching a "good" state. It might not be perfect, but that is the enemy of good. Reviews for networking-baremetal would be appreciated. I'll be revisiting networking-generic-switch patches this week, if folks who care about cisco hardware and networks could chime in, that 15:17:16 <TheJulia> would also be helpful. 15:18:14 <TheJulia> Once we have building blocks merged, I'll propose some docs against ironic to try and bring clarity to the topic from the operator point of view 15:18:35 <TheJulia> Any vxlan-ey questions? 15:19:53 * TheJulia takes silence as a no and lets iurygregory resume 15:19:59 <iurygregory> ack 15:20:17 <iurygregory> we don't have discussion topics so I will skip 15:20:29 <iurygregory> #topic Bug Deputy Updates 15:20:38 <iurygregory> #info we got 4 new bugs 15:20:48 <iurygregory> #link https://bugs.launchpad.net/sushy-tools/+bug/2137387 15:20:59 <cid> A little quiet. There were 3 newly filed bugs 15:21:20 <cid> And the dnsmaq one got new activity 15:21:26 <cid> https://bugs.launchpad.net/ironic/+bug/2026757 15:22:33 <iurygregory> tks cid ! 15:22:57 <kubajj> o/ 15:22:58 <iurygregory> https://bugs.launchpad.net/ironic/+bug/2139557 is triaged 15:23:31 <iurygregory> but doesn't have a person assigned, so in case someone is interested feel free to pick it up 15:23:59 <iurygregory> it's related to PySNMP, which I recall we had some problems with the newer version for sure 15:24:13 <cid> Yeah. It looks sane to me. 15:24:24 <iurygregory> not sure if ppl using SNMP will try to fix or not... 15:24:29 <cid> Nobody is assigned because the fix is not in progress yet. At least to my knowledge 15:24:44 <iurygregory> yeah 15:24:52 <iurygregory> Who is the next bug deputy? 15:25:14 <cid> I would. 15:25:55 <iurygregory> I can take next week 15:26:08 <cid> okay. tks! 15:26:19 <iurygregory> #info cid to be the bug deputy this week and iurygregory next week 15:26:36 <iurygregory> we don't have RFE to review, skipping 15:26:42 <iurygregory> #topic Open discussion 15:26:54 <iurygregory> we have one topic on it, not sure who added 15:27:02 <fungi> following up on last week's "Bridging the Gap Flamingo Cycle Retrospective" topic... 15:27:11 <fungi> this effort started with representatives from member organizations sharing frustrations with, primarily, their employees' experiences trying to contribute patches in various openstack projects 15:27:18 <fungi> investigating these reports on a case-by-case basis, foundation staff concluded most misunderstandings were due to mismatched expectations arising from incomplete communication or silence (and not necessarily on the part of the maintainers) 15:27:27 <fungi> what we've observed from successful exchanges in some projects is that increasing communication effectiveness ultimately leads to improved efficiency and time savings for all parties involved; some examples include: 15:27:33 <fungi> documenting review priorities and the prioritization process, as well as publicizing it more (e.g. with pointers in review comments), so that change owners know the priority for their own work and how to get involved in that decision 15:27:39 <fungi> proactively communicating reviewer availability, changes in availability, and explicit handoff to other reviewers, so that change owners know how long they might be waiting (and encouraging them to communicate their own availability) 15:27:48 <fungi> overall clearer communications with change owners, for example avoiding heavy reliance on acronyms and team-specific jargon, since english is not the primary language for a majority of our community and their familiarity with it varies 15:27:55 <fungi> to this end, i and other community managers on the foundation staff are working on putting together some materials that we hope teams will find useful, and will collaborate with us to help improve and expand further 15:28:02 <fungi> one thing we want to try is cut-and-paste review response templates for common situations; the vulnerability management team has used similar techniques for bug triage and assembling advisories, and they've found it saves a lot of time 15:28:10 <fungi> another is assembling contributor and reviewer checklists to help avoid common pitfalls and anti-patterns, based on all of the earlier focus group feedback and brainstorming sessions we held with the community in 2024 15:28:17 <fungi> something else we want to try is collaborating to produce case studies with companies who have been investing in project maintenance work and whose employees have a track record of successful contribution patterns 15:28:24 <fungi> we also plan to continue identifying additional tactics and strategies that teams have positive experiences with, to see if there's a way they can be generalized for adoption by other teams facing similar challenges 15:28:32 <fungi> if anybody has follow-up thoughts or questions on any of the above, as well as for the survey and metrics analysis from last week now that you've had time to mull it over, i'm happy to answer questions and entertain feedback here 15:28:41 <fungi> (or here in irc after the meeting, or on the openstack-discuss mailing list, or even directly/in private for that matter) 15:29:24 * iurygregory reading... 15:31:04 <dtantsur> I have a nagging suspicious that overloaded reviewers are a big part of the problem, but we cannot really solve that (esp. in a contributor-friendly way) 15:32:20 <fungi> it is, but the underlying satisfaction problem is less the lack of reviewer availability and more that casual contributors don't know there's a shortage, so in a lot of projects (not necessarily ironic) they never get a response on their uploads 15:33:17 <fungi> in last cycle's brainstorming discussions, the nova team committed to scaling back their spec approvals, for example, and being clearer about how much they'd realistically be able to review in the cycle 15:33:39 <TheJulia> I mean, we see that with other projects where we also know to nag them and we know the disconnect. I suspect its much more a "there is too many things" and not enough reviewers problem (mix in burnout of course) 15:33:54 <fungi> acknowledging that approving specs that they don't have time to review the changes implementing is setting incorrect expectations for the spec proposers 15:35:38 <dtantsur> I have mixed feelings about throttling contributions through specs, to be honest 15:35:42 <TheJulia> I think setting/adjusting expectations is critical 15:35:45 <TheJulia> I think its bonkers 15:36:04 <fungi> yeah, communicating what changes are unlikely to get reviewed and why sets clearer expectations 15:36:06 <TheJulia> Specs were intended to slow the process to a crawl 15:36:24 <TheJulia> cycle time/engagement are going to always vary though 15:37:36 <TheJulia> Bottom line, its a dialog, teams need to not just blindly promise to review/engage and not follow-up. Overloading wise, its easy, we're all well intending 15:37:38 <fungi> in a lot of the cases we investigated, contributors assumed things that weren't correct, in the absence of communication from maintainers, and this is what ultimately led to dissatisfaction with the process 15:38:15 <fungi> also things like not knowing who to reach out to and how to find out whether their chjange was likely to get reviewed 15:38:32 <dtantsur> We need a queueing system with numbers :D pick a number for your PR, wait for it to show up on the screen :D 15:38:43 <TheJulia> heh 15:39:18 <dtantsur> Seriously though, it's very good food for thought, thank you fungi! 15:39:35 <fungi> one classic case was someone trying to get a change reviewed for nova, had some questions, finally figured out to ask in irc, but the one time they did so on december 26 when there was basically nobody around 15:39:53 <iurygregory> ohhhh 15:39:55 <iurygregory> =( 15:40:07 <fungi> (maybe that should have been obvious to them, but they were from part of the world where that's not holiday season) 15:40:12 <dtantsur> I wonder if office hours could help 15:40:25 <dtantsur> with a schedule published somewhere, obviously, including any major shutdowns 15:40:38 <iurygregory> and mentioning in the review could help also 15:40:41 <TheJulia> But would it be findable? 15:40:48 <TheJulia> Would it be easy to discover/become aware of 15:40:51 <iurygregory> sometimes ppl are shy 15:41:04 <dtantsur> I recall at some point any new contributors would get a special message on their first patch? 15:41:17 <iurygregory> on the first they do 15:41:35 <TheJulia> Which is *often* against the test project 15:41:41 <iurygregory> but maybe the problem happened on the second etc 15:41:44 <dtantsur> Well, if we can customize these per project, or even if we cannot, that's where to mention the office hours and IRC 15:41:48 <iurygregory> yeah, normally is the sandbox 15:42:16 <dtantsur> I guess it could be technically possible to show the message both on the 1st sandbox commit and on the 1st real one 15:42:23 <TheJulia> I mean, I kind of pushed the idea of a ai driven initial response, but maybe a "hey, you are not a regular, weekly meeting is at this time" comment might not be a bad idea 15:42:28 <iurygregory> maybe Zuul could add a message to the patch? in case CI failed (hey, in case you are unsure about the failure, reach out to the tem on this channel on irc) 15:42:51 <iurygregory> not sure if would help, just a random idea =) 15:43:03 <dtantsur> that is separately a good idea 15:43:44 <fungi> worth noting, this isn't so much "new" contributors we were hearing from, but (mostly operator) contributors at openinfra foundation member companies who had limited time per week to devote to their patches 15:43:50 <opendevreview> Merged openstack/ironic master: Allow node lookup for in-band servicing https://review.opendev.org/c/openstack/ironic/+/971459 15:44:18 <TheJulia> So its a class of folks who may only have an hour or two a month ? 15:44:30 <TheJulia> which may not even align with anything the team has available? 15:44:37 <fungi> or maybe more, but don't know how to use that time effectively/efficiently to engage with the maintainers 15:45:36 <fungi> basically in lots of these cases, upstreaming patches is not their primary job, it's just something they get tasked with in hopes of not having to carry patches for bug fixes or small features forever 15:46:12 <TheJulia> its a convenience thing for them, they will carry that debt not recognizing it becomes a crushing debt 15:46:52 <fungi> and some of them might be untapped candidates for spending more time upstream and helping with maintenance, but the experience they have struggling to get changes in front of the right people leaves a sour taste in their mouths 15:47:45 <TheJulia> Has anyone done an analysis of these sorts of changes? Are they fundimentally missing key required aspects? 15:48:16 <TheJulia> or is it just entirely silence and they become frustrated and don't bother to respond/revisit ? 15:48:32 <fungi> yes, the foundation staff dug into a bunch of the ones that member company representatives forwarded to us, and tried to categorize them and find common themes 15:48:48 <opendevreview> Dmitry Tantsur proposed openstack/ironic master: Allow aborting deployments in "wait call-back" state https://review.opendev.org/c/openstack/ironic/+/973279 15:49:17 <fungi> in basically all cases, as i said, it was ultimately a communication breakdown, and i would say in most cases on the part of the people trying to contribute the change 15:50:09 <fungi> reviewers would leave questions or ask for revisions, the contirbutors would disappear for too long and then not be able to figure out how to address those comments or not be able to get reviewer attention again once they did 15:50:35 <fungi> they would ask for help with their changes in the wrong places, at the wrong times, in the wrong ways 15:50:57 <iurygregory> =( 15:51:17 <fungi> but terse communication or missing follow-up from maintainers helped stoke their misunderstandings 15:52:12 <iurygregory> not sure if a bot could help? (no updates in 2 week, bot adds a comment with some custom message?) 15:52:13 <TheJulia> That likely goes right back to overloaded reviewers 15:52:17 <fungi> that was the initial research which led to us putting together this whole effort, essentially trying to find ways to fix the communicaiton gap between these operators and other employees at companies in our ecosystem and the upstream maintainers 15:53:16 <TheJulia> iurygregory: I would be good with that, as long as we don't do anything bonkers like auto-abandon after some number of weeks of no response 15:53:17 <dtantsur> If the foundation of the infra team wants to invest into something, a review dashboard could be helpful 15:53:27 <TheJulia> We see bug fixes from operators die on the vine like that elsewhere :\ 15:53:28 <dtantsur> We tried maintaining our own, but it's always a bit of a striggle 15:53:31 <fungi> and yheah, i agree, we concluded that one of the best things that could be done to help is to find ways to relieve some of the maintainer/reviewer burden. we can't magically find more people, but maybe we can help the ones we have find ways to make their tasks easier 15:53:53 <dtantsur> Such a dashboard could, for instance, highlight patches without feedback for 2 weeks or whatever 15:54:06 <dtantsur> prioritizing those without any -1 15:54:17 <iurygregory> TheJulia, agree 15:54:50 <fungi> i was talking with the manila team last week about maybe putting together a review intake bot that can leave comments with initial pointers to team-specific contributor docs and review priorities 15:55:17 <fungi> basically help to set expectations at the time the change's first revision is uploaded 15:55:31 <fungi> and make sure these casual contributors know where to find the right resources and information 15:56:44 <janders> before we wrap, can I bring up a quick topic? 15:56:53 <janders> (sorry for missing most of the meeting) 15:57:00 <fungi> yeah, sorry, i didn't want to eat up the remainder of the meeting 15:57:08 <janders> no worries, this is important! 15:57:16 <iurygregory> fungi, totally fine, don't worry 15:57:24 <TheJulia> janders: whats up? 15:57:24 <janders> I am trying towards being a reviewer, those ideas sound very helpful 15:57:32 <janders> TheJulia monitoring 15:57:40 <janders> 1) https://review.opendev.org/c/openstack/ironic/+/974985 - I would appreciate reviews on doco 15:57:40 <iurygregory> I think we have things we can try to improve ++ 15:57:43 <janders> when you folks have time 15:57:46 * TheJulia makes beep boop whirling sounds 15:58:06 <janders> and 2) I wanted to ask around to see if anyone in the community would be keen to test the monitoring feature on mid-scale real-hardware 15:58:15 <janders> so in other words - calling for testers! 15:58:27 <janders> would be good to see how this code handles real failures at a bit more scale 15:59:01 <janders> I think kubajj may be looking at this but if anyone else is interested in trying too, please let me know 15:59:18 <janders> would be great to get some real world feedback before advertising this feature a bit more 15:59:24 <janders> thanks! :) 15:59:28 <TheJulia> Given how risk adverse operators with existing infra tend to be, we might not see that until after the end of cycle, which is fine, it happens. 15:59:45 <TheJulia> That whole, their primary thing is not testing, its operating the metal 15:59:49 <kubajj> janders: I will have a look into it once I get rid of inspector 😉 15:59:58 <janders> thank you! 16:00:03 <janders> that's all from me 16:00:55 <TheJulia> So one of the things with the vxlan work is that I had to kind of realize "we're not going to have perfect testing of it this cycle", so on some level we just need to be prepared to accept reports and backport fixes 16:01:26 <iurygregory> #chair TheJulia 16:01:26 <opendevmeet> Current chairs: TheJulia iurygregory 16:01:31 <TheJulia> Is there anything else? Or can we close the meeting up? 16:03:16 * TheJulia hears crickets 16:03:20 <TheJulia> Thanks everyone! 16:03:25 <janders> o/ 16:03:32 <TheJulia> #endmeeting