Monday, 2025-06-30

rpittaugood morning ironic! o/07:04
rpittaufungi: usually on the day of the meeting, so today07:04
opendevreviewMerged openstack/ironic master: trivial: follow-up on I07172e48207e09c0858298e34eea038c776d3c74  https://review.opendev.org/c/openstack/ironic/+/95368208:16
abongalegood morning ironic! 08:42
abongaleHey guys, the gating pipeline for https://review.opendev.org/c/openstack/ironic/+/953323 failed on the metal3-integration job due to a connection issue while collecting logs: "rsync: connection unexpectedly closed". Job link: https://zuul.opendev.org/t/openstack/build/f8d73a4e59f3423d9031afbe2c1a0384 — is it possible to get the pipeline re-run?09:03
rpittauabongale: not the single job, you'll have to do rerun all of them09:04
abongalerpittau: do we have docs that can guide me how to do it, please? and do I need a specific permission to do it?09:11
rpittauabongale: just type recheck in a comment and the reason09:11
queensly[m]Good morning 09:18
abongalerpittau: thank you09:46
opendevreviewVerification of a change to openstack/ironic master failed: fix: handle unexpected 'loop' in actions field  https://review.opendev.org/c/openstack/ironic/+/95332310:05
fungirpittau: should i update the date on the agenda myself then when adding a topic? it still says it's "for June 23, 2025"12:50
TheJuliagood morning13:00
TheJuliafungi: if your going to propose an item, yeah, it helps13:01
fungigot it. i didn't know if the meeting chair did that in advance before soliciting agenda additions13:08
fungiand i was worried if i put something on there now it would be an addition to last week's agenda and dropped when the agenda was switched to this week's13:09
TheJulianah, team practice is generally whoever runs the meeting, which may be anyone really at this point, does the needful13:12
TheJuliaand just transforms, and sometimes means we revisit stuff, but that is often when we need to like if a bunch of folks are away13:12
fungicool, i updated the date on it to reflect today's meeting in that case13:13
fungithanks!13:13
iurygregorygood morning ironic13:18
TheJuliafungi: moved to "Discussion" since it is now a planned topic13:18
TheJuliafungi: open discussion is much more for freeform items like grabbing a purple or green sash or the discussion of an unladen european swallow13:19
TheJuliaand it's maximum speed. ;)13:19
fungino coconuts, got it13:20
* TheJulia wonders when we suddenly got stuck on a tropical island after our three hour tour13:21
* TheJulia wonders who is the professor13:21
opendevreviewTakashi Kajinami proposed openstack/ironic-inspector master: sqlalchemy: Use built-in declarative  https://review.opendev.org/c/openstack/ironic-inspector/+/95376413:21
opendevreviewMerged openstack/ironic master: fix: handle unexpected 'loop' in actions field  https://review.opendev.org/c/openstack/ironic/+/95332313:31
opendevreviewAbhishek Bongale proposed openstack/ironic-tempest-plugin master: WIP Add Tempest tests for inspection rules in Ironic  https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/95176113:46
cardoegood morning ironic14:19
TheJuliagood morning14:22
priteauHello Ironic! Is there a way to clear the last_error node field without making a state change? One of our customers have BMCs that sometimes become unresponsive. BMC reset fixes it, resolving the problem for Ironic, but the last_error entry remains and is displayed as an error pop-up in ironic-ui.14:55
TheJuliapriteau: Greetings, as far as I'm aware, there is no way to clear it14:56
TheJuliaI guess it would make sense to auto-clear upon power sync recovery if node history is enabled14:56
TheJuliao/15:01
* TheJulia wonders if we have quorum to hold a meeting15:01
iurygregoryo/15:01
fungiisn't most of europe on vacation for the next 3 months? ;)15:01
rpittauo/15:01
cido/15:01
rpittaufungi: if only! :D15:01
TheJulia... not *all* of the EU15:02
queensly[m]o/15:02
TheJulia#startmeeting ironic15:02
opendevmeetMeeting started Mon Jun 30 15:02:06 2025 UTC and is due to finish in 60 minutes.  The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot.15:02
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:02
opendevmeetThe meeting name has been set to 'ironic'15:02
TheJuliao/15:02
rpittauo/15:02
TheJuliaGood morning everyone!15:02
cido/15:02
TheJulia#topic Announcements / Reminders15:02
TheJulia#undo15:02
opendevmeetRemoving item from minutes: #topic Announcements / Reminders15:02
TheJuliaOur meeting agenda can be found on the wiki.15:02
TheJulia#link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_June_30.2C_202515:02
TheJulia#topic Announcements / Reminders15:02
TheJuliaAs a general reminder, please spend time and review ironic-week-prio items.15:03
TheJulia#link https://tinyurl.com/ironic-weekly-prio-dash15:03
TheJuliaAs a reminder, according to the release schedule, we're in week R-13.15:03
TheJulia#link https://releases.openstack.org/flamingo/schedule.html15:03
TheJuliaDoes anyone have anything to announce or remind us of?15:04
rpittautime flies?15:04
TheJuliaIt does!15:04
fungiforum session proposal deadline?15:05
TheJuliafungi: no idea on the date, if you have it and a link, please remind folks15:05
iurygregoryJuly 0915:05
iurygregoryops15:06
iurygregory8 July: The Forum CFP closes 15:06
fungiaha, thanks, i was hunting for the announcement link15:06
iurygregory#info Forum CFP closes on July 8 https://summit2025.openinfra.org/15:06
TheJuliaExcellent, Thanks!15:07
iurygregorynp15:07
TheJuliaOh, one last item before working group updates. Next monday I'm going to be stuck in line to enter a service bay for some repair work, so.... I doub't I'll make the meeting.15:07
TheJuliaIf someone else can be sure to run the meeting, that would be awesome15:08
TheJulia#topic Working Group Updates15:08
TheJuliaFirst up, the Standalone networking group15:08
TheJuliaLooks like alegcy is not around today, I think they will be back next week.15:08
TheJuliaPlease take a little time to review the spec15:09
TheJulia#link https://review.opendev.org/c/openstack/ironic-specs/+/95212615:09
TheJuliaAs for Eventlet Removal!15:09
TheJuliaDo we have anything to really discuss, aside from what I noted in the IRC log on Friday?15:09
TheJulia(We'll need to retool signal handling/stopping the conductor, and explicitly opt out of eventlet entirely with additional code15:10
TheJulia)15:10
cidNothing from my end.15:10
TheJuliaOkay, then, on to Discussion topics!15:11
TheJulia#topic Discussion15:11
TheJuliafungi brings us a topic this week, "initial analysis of ironic surveys and metrics".15:11
TheJuliafungi: you have the floor15:11
fungii'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 2 weeks ago:15:11
fungi#link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/NTBNI7YIDCWBR6BTPEKVZIODWTVUIOXO/ BtG metrics analysis15:11
fungialso 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!)15:11
fungias a follow-up activity, we've started doing some team-specific analyses, focusing on teams that had multiple contributor and maintainer survey responses (did nova and cinder a couple weeks ago, octavia coming up later this week)15:12
fungiwe'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 expected15:12
fungia 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 yet15:12
fungiwe have plans to continue with this sort of surveying and metrics analysis over coming release cycles; as for the initial results...15:12
fungithe contributor survey had 2 responses for ironic and both respondents had contributed for at least a year and were contributors to at least two other open source projects15:12
fungimost feedback was relatively positive (averaging 3-5 out of 5) with the highest score on "You receive actionable feedback from other reviewers" and lowest "[contributor documentation] It is easy to navigate"15:12
fungithe top challenge reported was trouble with review attention when not already working closely with a maintainer, while additional feedback mentioned it was hard to get involved outside of internship programs15:13
fungithe maintainer survey also had 2 responses with slightly lower scores than the cntributor survey (averaging 3-4)15:13
fungicontributing challenges reported were similar to those from the other survey (review attention) but also disagreements between maintainers delaying changes as well as delays due to testing problems15:13
fungithe top challenges with reviewing were insufficient familiarity with some parts of the code, contributors not fixing their changes or responding to feedback in a timely manner, and ai-generated changes with insufficient explanation15:13
fungilooking at metrics we gathered from gerrit for the past 5 development cycles, we saw review activity was quite good with relatively fast turn-around compared to openstack as a whole (even if not meeting expectations from contributors)15:13
fungikey takeaways seem to be that improving or overhauling the contributor docs could help, as well as diverting more available time to shepherding newer and non-maintainer contributors, perhaps also some attention to test stability15:13
fungithe maintainer disagreements over reviews seem like they might be related to the feeling of lack of familiarity with parts of the codebase, i don't know whether doing some deep-dive presentations on the code could help there15:13
fungiwith ironic doing comparatively well both in terms of survey responses and review metrics, i'd like to set aside time at some point soon to get a better understanding of your team-specific workflows to see if any might be portable15:14
fungisorry, 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 after15:14
* iurygregory reading15:14
TheJuliaSo, regarding codebase layout and knowledge, we tend to treat that as 1-on-1 sessions related to folks interests/challenges, and then try to work with them, but I think it would help us know if people really want/need that level of help. Its one of those things which does sink a bit of time into, and I don't think anyone who is a maintainer would really object, we would just need to find the time and actively be aware of 15:15
TheJuliaa desire.15:15
TheJuliaI suspect, folks are just too afraid to ask?!15:15
iurygregoryI think some times people are afraid to ask15:16
fungiyeah, note it was on the maintainer survey that the lack of familiarity with some parts of the codebase was highlighted as something that made it harder for them to review contributions15:16
iurygregory"hard to get involved outside of internship programs"? I'm surprised by this one .-.15:17
TheJuliaI guess I'm not really if they are afriad to jump in because it is complex15:17
fungikeep in mind that we had a whopping total of 2 responses to each survey, this is far from statistically significant response, but rather individual experiences reportedc15:17
TheJuliaThere is likely no one single fix, but some smaller fix we likely need to ponder15:17
clifmaybe some encouragement to do reviews even if you're not totally familiar with that part of the codebase? view it as an opportunity to learn about different parts of the code while also hopefully improving incoming code quality15:17
TheJuliahmm15:17
TheJuliaso perhaps a lense needs to be applied then15:18
fungiyeah, on other projects i've worked on, reviewing changes to parts i'm unfamiliar with has improved my familiarity15:18
clifalso, reviewing doesn't have to result in actionable feedback to the submitter... sometimes there's nothing you have to add15:18
TheJuliaThe sample size is sort of a skewing aspect. We should likely discuss and ponder how we could be more helpful on the maintainer side15:19
rpittauI guess that's where we all started, building familiarity with reviews, and not viceversa?15:19
iurygregoryone thing to note, sometimes people doesn't even know how to reach out via irc .-.15:19
TheJuliaWell, we're in this weird shift point in the lifecycle of an OSS proejct15:19
TheJulianow it is much more about tactical contribution15:19
fungiand as i said at the beginning, the best way to get a more statistically significant overview is for more people to take those surveys15:19
iurygregoryI received this feedback during OIS Asia last year .-.15:19
TheJuliaPerfecte example, alegacy working on networking15:20
* TheJulia wonders where that extra e came from15:20
TheJuliaiurygregory: what feedback did you receive at OIS Asia ?15:20
dtantsuron topic of code organization, following the deployment logic through all the modules and obscure functions can deprive anyone of sanity15:21
iurygregorythat they would prefer other way to contact us (other than irc..)15:21
TheJuliaYeah, the whole module model is definitely an area folks struggle15:21
rpittaumm I agree with dtantsur 15:21
iurygregorynewer people probably have a hard time15:21
TheJuliaYeah, IRC15:21
iurygregory++ agree with dtantsur also15:22
TheJuliaI think fungi has given us a bit to think about, would it make sense for us to have a group discussion via a higher bandwidth medium at some point soon to try and figure out some action items, above and beyond "hey, go take this survey" messaging15:22
fungiwell, also i don't expect what we've seen so far is terribly enlightning, but it gives us a starting point for future comparisons15:23
cidIndeed15:23
cidI can't relate to the "hard to get involved" without onboarding, which is why I think internships does help.15:23
cid*can :D15:23
iurygregory++15:24
fungianyway, i didn't have anything else to present, so happy to yield the floor if there are no further questions/comments for now15:24
iurygregoryI think mentorship (even without internship helps)15:25
iurygregoryat least that is how winiciusallan[m] has been contributing from time to time =)15:25
cid++.15:25
fungiyeah, one of the comments implied that working directly with a maintainer made it easier to get changes merged15:26
TheJuliaSo, let try to circle back next week and see if we have any idea ?15:26
TheJulias/idea/ideas/15:26
rpittausounds good15:26
fungithanks for listening!15:27
cidtks, fungi!15:27
TheJuliaThanks fungi!15:27
iurygregorytks fungi !15:27
TheJuliaOnward!15:27
TheJulia#topic Bug Deputy Updates15:27
TheJuliacid: I think you were our deputy last week15:27
cidThere two bugs and an RFE15:27
cidAt least, I triaged it as an RFE15:27
cidIt was from cardoe, I doubt he's around today15:27
TheJuliaOne of those open bugs I suspect I could just close. Its... likely more a doc issue or a troubleshooting guide issue15:28
cardoesorry I'm here I just have 3 meetings wide right now.15:28
cidRight...15:28
cidThis is it by the way, https://bugs.launchpad.net/ironic/+bug/2115471 - Consistent object shape of inventory inspection data15:28
TheJuliaI think, at a high level it makes sense15:29
TheJuliaI guess the question ends up also being is any of that data sourced via in-band inspection15:29
TheJuliaWho volunteers to be the deputy for this week?15:31
cidcardoe, ^^. I would think from in-band inspection15:31
cidTheJulia, I don't mind!15:31
TheJuliacid: okay15:31
TheJuliacardoe: cid: so I mean out of band inspection, data not coming from IPA15:31
cardoecid: so no the issue is with all data.15:32
cardoeI'm trying to get the in-band (redfish) inspection working.15:32
cardoeBut that's generating different data.15:32
cardoeSo I'm trying to create some objects to provide some validation (which is what IPA has in tree)15:32
cardoeBut it turns out the checks in IPA are bypassed today and it uploads whatever data objects its gotten.15:33
TheJuliaYeah, its always going to create different data unfortunately.15:33
cardoeIf you use a different hardware manager, the shape of the data is different15:33
TheJuliayup15:33
cardoeSo what I'm aiming to do is define inside of Ironic fields that are "required"15:33
cardoeAnd if the required fields are missing, treat like the inspection data is missing (which is what by suggestion is on that bug)15:34
TheJuliacardoe: fwiw, I commented on the rfe, indicating general +1, just feels like we need a little bit more clarity/consensus and maybe that is through patches, dunno15:34
cardoee.g. https://review.opendev.org/c/openstack/ironic/+/953414 is something that Haseeb is trying to fix.15:34
cardoeBecause it turns out that the name field is required in a few different paths of Ironic15:35
TheJuliacardoe: might be worth while to then make it a bit more verbose since thats not how I'm really reading it, but on some level it may also be just an implementation detail15:35
cardoeYou mean write more details on the bug?15:35
TheJuliacardoe: yes15:35
cardoeCan do.15:36
TheJuliaThanks!15:36
TheJuliaOnward to open discusion?15:36
cardoehttps://bugs.launchpad.net/ironic/+bug/2115471 which is what cid linked is just my initial WIP at a model that we can use for validation in the future.15:36
TheJuliaalegacy: o/15:37
TheJulia#topic Open Discussion15:37
TheJuliaDoes anyone have anything they would like to discuss further? Plot? Scheme? etc.15:38
cardoeEssentially I'm using dataclasses so that you can use something like pyright / mypy to confirm we're messing with fields and you'd just get a warning in your IDE with pyright if you're accessing a field that's not explicitly defined so that should give someone pause to make sure it's good.15:39
cardoesorry for my poor communication. Trying to listen in on the Neutron BGP call and also on a call providing an update to my leadership chain about my team's work.15:40
TheJuliacardoe: no worries! That is a lot!15:41
TheJuliaWell, if there is nothing else, I'll give everyone back at least 15 minutes to plot/scheme and try to take over the world.15:42
cid\o/. 15:43
* TheJulia wonders if we forgot to make coffee this morning15:43
TheJuliaThanks everyone!15:43
TheJulia#stopmeeting15:43
TheJulia#endmeeting15:43
opendevmeetMeeting ended Mon Jun 30 15:43:55 2025 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:43
opendevmeetMinutes:        https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-06-30-15.02.html15:43
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-06-30-15.02.txt15:43
opendevmeetLog:            https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-06-30-15.02.log.html15:43
opendevreviewVerification of a change to openstack/ironic master failed: Advanced vmedia deployment test ops  https://review.opendev.org/c/openstack/ironic/+/89801016:40
*** iurygregory__ is now known as iurygregory16:58
iurygregorydtantsur, "I wonder if we want >=500 in case someone gets creative", you mean we should also add the >=500 case (since we have conflict 409, bad request 400)? or just consider >= 500 ?16:59
priteauAnyone familiar with this "Ready to submit" state? https://review.opendev.org/c/openstack/bifrost/+/94824517:09
iurygregorythis is new to me17:09
iurygregoryO.o17:09
iurygregorymaybe is waiting for the promote pipeline?17:10
opendevreviewMerged openstack/ironic master: Mark SNMP driver unsupported for removal  https://review.opendev.org/c/openstack/ironic/+/94684317:43
TheJuliadid it merge?17:47
TheJuliaor is it still pending17:47
TheJuliawe've seen it once or twice, a rebase typically fies17:47
opendevreviewIury Gregory Melo Ferreira proposed openstack/sushy master: Don't fail to eject vmedia in case there is no vmedia  https://review.opendev.org/c/openstack/sushy/+/95273017:50
iurygregory_TheJulia, "Ironic has responded with a conflict" do you have some details in what scenarios this can happen?17:58
TheJuliaI'm not sure what topic youre referring to17:59
iurygregory_TheJulia, https://opendev.org/openstack/ironic-python-agent/src/branch/master/ironic_python_agent/ironic_api_client.py#L27918:06
iurygregory_being honest i never saw this happening in any case, and now we have a customer case D:18:07
*** iurygregory_ is now known as iurygregory18:07
opendevreviewIury Gregory Melo Ferreira proposed openstack/ironic master: Handle unresponsive BMC during Firmware Updates  https://review.opendev.org/c/openstack/ironic/+/93810818:22
TheJuliaoh, heh19:09
TheJuliaokay, so if memory serves it is a conclict when attempts to lock the node in the conductor and it is already locked by another task, either heartbeat or other action. I think the idea is that if its a conflict we just backoff on the next retry a little bit and things should resolve themselves19:10
fricklerpriteau: for https://review.opendev.org/c/openstack/bifrost/+/948245 the "ready to submit" is indeed misleading, it needs a rebase since a newer version of the parent was merged19:12
opendevreviewJulia Kreger proposed openstack/ironic master: WIP: Set a threading backend  https://review.opendev.org/c/openstack/ironic/+/95368321:15
opendevreviewJulia Kreger proposed openstack/ironic master: WIP: Set a threading backend  https://review.opendev.org/c/openstack/ironic/+/95368323:18
TheJuliastevebaker[m]: ^^^ 23:18
TheJuliaits process -> multiprocessing -> Threads, lots of threads23:19
TheJuliaso it *seems* to work, at least the other copy on my devstack machine which can't talk to gerrit23:19
TheJuliaI'll give it more of a spin tomorrow, but it will be interesting to see23:21
TheJuliamost interesting thing is the teardown of the workers, that seemed to hang which makes no sense in that context23:24
stevebaker[m]cool, I'm taking a look23:25
TheJuliaI *suspect* we should similarly have a "we're shutting down" event flag to begin to prevent inbound work23:34
TheJuliabut the shape of the graceful/drain logic... dunno23:35

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!