Monday, 2025-09-08

opendevreviewBaptiste Jonglez proposed openstack/networking-generic-switch master: Fix support for Aruba AOS CX devices  https://review.opendev.org/c/openstack/networking-generic-switch/+/95823407:38
opendevreviewBaptiste Jonglez proposed openstack/networking-generic-switch master: Aruba: Add error message patterns to detect errors  https://review.opendev.org/c/openstack/networking-generic-switch/+/95823107:39
opendevreviewBaptiste Jonglez proposed openstack/networking-generic-switch master: Detect device configuration errors on Dell Force10 and OS10  https://review.opendev.org/c/openstack/networking-generic-switch/+/86831807:40
opendevreviewBaptiste Jonglez proposed openstack/networking-generic-switch master: Cast boolean Netmiko kwargs to native types  https://review.opendev.org/c/openstack/networking-generic-switch/+/95807907:40
abongalegood morning ironic o/08:07
opendevreviewcid proposed openstack/ironic-python-agent master: Add type annotations and checking  https://review.opendev.org/c/openstack/ironic-python-agent/+/95833309:32
opendevreviewDmitry Tantsur proposed openstack/ironic master: WIP run the ironic-standalone-operator tests as non-voting job  https://review.opendev.org/c/openstack/ironic/+/95969014:04
opendevreviewDmitry Tantsur proposed openstack/ironic master: WIP run the ironic-standalone-operator tests as non-voting job  https://review.opendev.org/c/openstack/ironic/+/95969014:18
* dtantsur is curious if we even have quorum for the meeting14:51
JayF\o14:52
dtantsurmaybe people are just silent :D (so was I)14:54
kubajjo/14:59
clifo/15:00
masgharo/15:00
dtantsurWho is chairing today btw? I'm waiting for a delivery within 15 minutes15:00
opendevreviewDmitry Tantsur proposed openstack/ironic master: Run the ironic-standalone-operator tests as non-voting job  https://review.opendev.org/c/openstack/ironic/+/95969015:01
TheJuliaI'm taking today off, so not me :)15:01
JayF#startmeeting ironic15:02
opendevmeetMeeting started Mon Sep  8 15:02:39 2025 UTC and is due to finish in 60 minutes.  The chair is JayF. 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
TheJuliaOTOH, I saw an awesome post from someone on LinkedIn this past weekend which spoke very highly of the Ironic project. The comments were fairly positive, one that redfish is basically awful and doesn't work which I encouraged them to file bugs. The other comments were mostly around our docs.15:02
dtantsurThanks Jay!15:02
JayFHello everyone, welcome to the ironic meeting! Going to wait a few minutes for folks to get ready to meet; totally not using it to open the agenda :D 15:03
dtantsurTheJulia: nice, thanks for sharing! Docs are .. docs.15:03
JayFdtantsur: link me that please15:03
dtantsurJayF: you meant to ask TheJulia, right?15:03
TheJuliadtantsur: yeah, I think the perception is largely rooted in the layout overall and the fact we call back to prior behavior a lot15:03
JayFSo current agenda in the wiki is 8/2515:03
JayFyes, I did15:03
JayFI'm going to wing it15:03
JayF#notes Reminder to review things in our ironic-week-prio dashboard: https://tinyurl.com/ironic-weekly-prio-dash15:03
JayFAll flamingo releases have been cut now, right?15:04
JayFI didn't check to see if the patches merged, but I reviewed them15:04
dtantsurI feel like Ironic is pending still15:04
dtantsurat least I don't see the branch15:04
JayF#note Ironic releases are pending review from release team: https://review.opendev.org/q/topic:%22ironic-releases-2025.2%2215:04
TheJuliaJayF: https://www.linkedin.com/posts/nilesh-chandekar-304b4b57_opensource-osa-openstack-activity-7365744037502345218-Nx3A should work15:05
TheJuliaYeah, I am not a release manager so there is no "trust" of me putting the patch in to release15:05
JayFWe generally ask members of the community, core reviewers, contributors, operators, to take some time before the official release late this month/early October to help test.15:05
JayF#topic Working Group Updates15:05
JayFDo we have any updates for Standalone networking?15:06
JayFI assume it's probably been a little quiet with release activities?15:06
TheJuliaAlan is out this week as well, he should be back next week15:06
TheJuliaor, maybe he is back this week but just getting back today15:06
TheJuliaso I wouldn't expect anything until sometime next week.15:06
JayFFor eventlet removal team, https://review.opendev.org/c/openstack/ironic/+/959870 was posted last week and needs to land, both in master and be backported to the (eventual) 2025.215:06
JayFWe thought we were done but surprise! 15:06
JayFI suspect we'll find a few more surprises as time continues.15:07
JayFThere are no discussion topics on the agenda. Skipping that section. New business can go in Open Discussion.15:07
JayF#topic Bug Deputy Updates15:08
JayFclif was bug deputy for the last 2 weeks, since the meeting was cancelled15:08
JayFclif: do you have anything of note for the meeting?15:08
JayFAlso; who wants to be the next deputy?15:08
clifPretty quiet two weeks. I think there was just one new ironic bug filed in that time.15:08
cido/15:08
JayFcid: is that a volunteer to be deputy next week?15:09
clifPerhaps two? I self-assigned this https://bugs.launchpad.net/ironic/+bug/212170215:09
clifWill take a look at it soon15:09
JayF#link https://bugs.launchpad.net/ironic/+bug/2121702 15:09
JayF>  When triggering a automated cleaning the ports are not created in the cleaning network Edit 15:09
cidJayF, I guess so. It also doubles as attendance :D15:09
JayFhaha be careful when you raise your hand :D thank you for your volunteering :D 15:10
cid:)15:10
JayF#note Next bug deputy: cid15:10
JayFNo RFEs for review.15:10
JayF#topic Open Discussion15:10
JayFI have one thing for this15:10
JayFWe have multiple repos in Ironic/Bare Metal program which are coming up on EOL.15:11
JayFMetalsmith. *inspector*15:11
JayFI will likely be taking action (starting with a formal ml discussion) around retiring these in October.15:11
JayFPlease speak now if you have interest in keeping these alive15:11
JayFor if you have other repos under our governance which are hiding under a rock that need to be removed from our governance15:12
TheJulia++15:12
kubajjI am just excited to try using Ironic without inspector as a separate repo ;P15:12
dtantsur++15:13
JayF#agreed Metalsmith and Inspector-related repos being considered for retirement early in the *G* cycle. Please speak quickly if you have good reason for them to stay alive, and/or want to volunteer to maintain them. Note that inspection will still exist and has been migrated to primary Ironic repos.15:13
JayFAnything else for open discussion15:13
JayFOne other interesting question: Anyone else going to be at Summit in Paris?15:14
dtantsurYep15:14
JayFEnough for an Ironic dinner, perhaps?15:14
TheJuliaI might be, awaiting approval.15:14
JayFI'll be there +2 folks from the downstream GR team15:14
dtantsurFrom us, at least rpittau, masghar and myself, I think, and I don't remember about janders and iurygregory15:15
JayFoh, nice! So yeah, we can pencil something in I hope.15:15
dtantsurLet's do it rather early than late if possible please since we're also trying to have a team dinner :)15:15
JayFI know little about the area and don't speak French so that may be a limitation, but I doubt we'll have trouble finding a nice place for a few folks to eat in *Paris*15:16
dtantsurit's not quite *in* Paris15:16
JayFHere is where someone volunteers to organize it who isn't an ocean away and only speaking english15:16
dtantsurwe can wait for rpittauto return next week and make puppy eyes15:16
TheJuliadtantsur: that sounds like an excellent plan15:16
masghar++15:16
JayF#voluntold rpittau to help arrange team dinner (please?)15:17
kubajjJayF: I will be there15:17
TheJulia#prettypleaseultiplier15:17
JayF#agreed 10+ Ironic contributors/operators will be at France summit. We're going to try to do a team meal.15:17
JayFAnything else before I close the meeting?15:18
dtantsuryep, 1 sec15:18
dtantsurI want to highlight https://review.opendev.org/c/openstack/ironic/+/95969015:18
dtantsur(I hope the delivery does not interrupt me right now lol)15:18
dtantsurThis is an attempt to bring a more granular Metal3 job to Ironic15:18
dtantsurTogether with a potential addition of a baremetal-operator functional tests job, this should eventually replace the cryptic metal3-integration job which nobody really knows how to debug :)15:19
kubajjI would also want to highlight https://review.opendev.org/c/openstack/ironic-python-agent/+/937342 as it might spark some discussion15:19
JayFDoes this have external dependencies which will break at release barriers?15:19
dtantsurJayF: fewer. Neither depend on CAPI, which is a big source of these breakages.15:20
dtantsurOf course, we need to observe how the newly added jobs behave on OpenInfra nodes15:20
JayFack; sounds good. My only real question is if experimental makes more sense; but it's a weak question as experimental jobs are easily forgotten :D 15:20
dtantsurI want some statistics15:20
* dtantsur picks up delivery, sorry15:20
JayFWe can pickup the discussion in gerrit15:21
JayFsimilarly kubajj I've added your patch to my review queue15:21
kubajjJayF: thanks, it is the one where you said you will not like to review 🫠15:22
JayF#note please review 959690 (ironic standalone operator testing) and 937342 (IPA raid-device-cleaning-skip fixes)15:22
JayFkubajj: I like patches sitting with no feedback for weeks much, much less :( 15:22
JayFLast call (chapter 2) for anything new for open discussion15:24
TheJuliaptg?15:24
JayFHas anyone done serious planning for the PTG other than creating the etherpad?15:25
dtantsurWe already have an etherpad? :D15:25
JayF#note Ironic already has a PTG etherpad: https://etherpad.opendev.org/p/ironic-ptg-2026.115:25
JayFit's been in the whiteboard for a couple of weeks 15:25
kubajj(I had to look up when it is - October 27 to 31)15:27
TheJuliaI've already added a few topics to the etherpad15:28
TheJuliaLargely I'm getting tons of network/storage asks these days15:29
TheJuliaStrong enough I'm tempted to just publish a prioritized list15:29
JayFthat sounds really useful15:29
JayF"prioritized list" = operator menu, and I can see if my downstream wants a #3 (proverbially)15:30
TheJuliaWell, where folks can collaborate around most value15:30
JayFAbsolutely. Or maybe find a common thread that can unite some of the asks15:31
TheJulialike, my networking stuff is drastically different than yours, but... there is overlap and opportunity collaborate regardless15:31
TheJuliaexactly15:31
JayFSounds good.15:32
JayFA couple reminders along these lines: 15:32
JayFwe need to get on scheduling the ptg in the next week or two15:32
JayFwe need to ensure release liasons get +1s on the pending release PRs that were linked at the top of the meeting15:32
JayFMost of the rest is stuff that can be coordinated in etherpad15:32
TheJulia++15:34
JayFI think that's it for the meeting? 15:34
JayFLast call v3 :)15:34
JayF#endmeeting15:37
opendevmeetMeeting ended Mon Sep  8 15:37:17 2025 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:37
opendevmeetMinutes:        https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-09-08-15.02.html15:37
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-09-08-15.02.txt15:37
opendevmeetLog:            https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-09-08-15.02.log.html15:37
dtantsurAn example IrSO job run: https://zuul.opendev.org/t/openstack/build/b641d9b173f740cfb58a7bd0f2a8acc1/artifacts. And its report https://e0d088a46d3a62d38611-0fe503db5fc22ef43aac4f060a09645b.ssl.cf2.rackcdn.com/openstack/b641d9b173f740cfb58a7bd0f2a8acc1/controller/report.html15:47
JayFDoes that operator do any network fun yet?16:04
JayFand/or keystone? 16:05
JayFmainly wondering if it's 'standalone' (basic auth, no network) or 'standalone' (no nova)16:05
kubajjJayF: sadly not yet16:12
JayFkubajj: so from my reading of your change16:15
JayFkubajj: a sufficiently motivated root user on a machine going to cleaning can kill the logical name of a raid device and prevent cleaning16:15
JayFkubajj: but I can't quite wrap my mind around if this is a new security risk, or an existing one16:15
* JayF puts it on gerrit patch with another piece of feedback16:16
kubajjJayF: can they remove the logical name?16:16
JayFlogical name is a piece of metadata in the software raid 16:17
JayFright?16:17
kubajjyes16:17
JayFso  a root user could remove that and cause a cleanfail16:17
JayFmaybe that's OK, but that's what my brain is thinking16:17
kubajjI mean, technically they could, but I would advise against it :D16:18
kubajjWe aim to deploy this downstream for our hypervisors16:19
kubajjJayF: maybe I am wrong, but I think it would not cause a clean fail, but rather just cleaning of the array16:21
JayFkubajj: https://review.opendev.org/c/openstack/ironic-python-agent/+/937342/6/ironic_python_agent/hardware.py#3563 ?17:14
kubajjJayF: yes, but we only raise that error if the skip list contains a volume name - i.e. the user is using RAID safeguards - which do not work now, without the change17:16
JayFso what I'm saying is17:16
JayFyou have a node with skip list configured17:16
JayFI am a nefarious evil, evil admin who wants to use this as step 1 in a chain of vulns to access cleaning network17:16
JayFso I remove the configured logical name on my raid array and circumvent cleaning, so my payload keeps running for longer17:17
JayFI suspect a sufficiently determined root user can likely already cause cleanfails, but it's the sorta thing I look out for 17:17
kubajjJayF: sorry, I was thinking about the other comment you left on the change17:17
kubajjmy bad17:17
JayFoh, backwards compat, got it17:17
JayFthat makes sense then17:17
JayFcan you make sure to reflect that in a release note?17:17
kubajjwill do17:18
kubajjfor the first comment, as I said, I think it would not fail, because we check in cleaning the existing raids and try to match them with the property, since the name will not match, it will just clean the disks - which is the same as if the root user removed the property from Ironic17:19
JayFI haven't forgotten you, am looking still 17:26
* JayF is trying to get claude to help17:27
* JayF is going to make an IRC bot called "claude" which just responds occasionally with "That's right!"17:34
JayFkubajj: I found a concrete issue; I'll put it on the gerrit change17:34
JayFkubajj: more or less verbatim quoted the claude stuff in the review after filtering out some invalid stuff, I think we found a couple of minor things we can cleanup 17:38
JayFkubajj: if you push an update to this and I don't look organically in a couple of business days, feel free to ping me in here17:39
TheJuliaSo, just putting it out there, I forgot to update the release mappings. worst comes to worst I can do it and backport it tomorrow, but I have stuff I need to do today which is still a day off for me and tomorrow I have time during a meeting where I can do it.17:43
JayFSo update ironic release mappings; update the releases PR for ironic to HEAD() once it's landed?17:45
JayFIf that's what's needed I can do it today17:45
TheJuliaor just let it land and backport the change17:47
TheJuliaeither, its all good17:47
TheJuliait just has to bed done early on post-release17:47
TheJulia(since your not going to pin to the release itself, you pin to latest or unset it.17:48
TheJulia)17:48
TheJulia(or, your going to pin to an even older version)17:48
JayFack I'm on it17:51
TheJuliamuch appreciated!17:52
TheJuliaOkay, I'm going to go annoy some neighbors before we disappear for a few weeks17:52
frickleris anyone aware of ironic sdk jobs failing for almost a month? I'm assuming based on what lutimura wrote in the sdk channel that the issue is caused by changes in the sdk, but maybe some still might want to help fixing? :) https://zuul.opendev.org/t/openstack/builds?job_name=openstacksdk-functional-devstack-ironic&project=openstack%2Fopenstacksdk&skip=018:09
JayFlooks like allocations API schema issues18:10
JayFwhich was changed this cycle on our end18:10
JayFstephenfin and adamcarthur5 have been working on those changes18:10
JayFhopefully one of them can take a look if not I'll try to get around to it18:10
lutimurai think it's related to this one: https://review.opendev.org/c/openstack/ironic/+/94521818:10
JayFagreed18:11
frickleryes, I was just wondering how the regression could come from the sdk itself when that ironic job is voting18:11
JayFhttps://review.opendev.org/c/openstack/ironic/+/945218/3/ironic/api/schemas/v1/allocation.py#78 it should work if on the correct microversion18:11
JayF"should" being the most evil word in software18:12
JayFlol18:12
lutimurafrickler: i'm pretty new to openstacksdk/ironic, so i wasn't really sure where this should be fixed (my bad!)18:12
JayFwe're all figuring it out together lutimura :) 18:12
JayFat least you have the excuse of being new :D18:12
fricklerlutimura: no problem, it was a good step to find the issue and mention it already18:13
lutimuraJayF: i'm not sure if that's what you suggested, but i tried to bump the microversion of these failing tests (specifically): https://review.opendev.org/c/openstack/openstacksdk/+/959332 -- unfortunately, it didn't work :(18:17
JayFHonestly this makes me wanna cross gate with these jobs18:19
kubajjJayF: thanks, will have a look tomorrow 😉18:19
JayFbecause it seems like it's giving us bonus api coverage18:19
opendevreviewJay Faulkner proposed openstack/ironic master: Update release mappings for 32.0.x  https://review.opendev.org/c/openstack/ironic/+/96017318:21
JayFcid: cardoe: dtantsur: other cores ^ can I get a quick set of reviews on this? 18:21
cardoeJayF: cid and I +2’d19:18
JayFI'll wait for V+1 before I self-approve, I don't love self-approving but for something like this it seems correct19:19
JayFAdam just closed the loop with me, he's going to be looking at the SDK CI failure tomorrow.19:53
opendevreviewMerged openstack/ironic master: Update release mappings for 32.0.x  https://review.opendev.org/c/openstack/ironic/+/96017321:33
opendevreviewArnaud Morin proposed openstack/ironic master: Avoid the wsgi application to start with eventlet  https://review.opendev.org/c/openstack/ironic/+/95987022:31
opendevreviewArnaud Morin proposed openstack/ironic master: Deprecate usage of ironic.api.wsgi:initialize_wsgi_app  https://review.opendev.org/c/openstack/ironic/+/96020922:31

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