15:02:24 <JayF> #startmeeting Ironic
15:02:24 <opendevmeet> Meeting started Mon Sep 15 15:02:24 2025 UTC and is due to finish in 60 minutes.  The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:02:24 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:02:24 <opendevmeet> The meeting name has been set to 'ironic'
15:02:31 <alegacy> o/
15:02:42 <JayF> Welcome to the Ironic team meeting. We're operating under the OpenInfra Code of Conduct.
15:02:46 <JayF> #topic Announcements/Reminders
15:02:55 <JayF> #link https://tinyurl.com/ironic-weekly-prio-dash
15:03:07 <JayF> As always please specifically take time to review hashtag: ironic-week-prio patches
15:03:17 <dtantsur> o/
15:03:39 <JayF> It's September 15. R-2. The Flamingo is looming!
15:03:49 <JayF> Release week is Sept 29-Oct 3
15:04:34 <JayF> Going to give a few minutes before moving on to the less-rote portions of the meeting for us to potentially gain quorum.
15:06:32 <cid> o/
15:07:55 <JayF> #topic Working Group Updates
15:08:06 <JayF> Is there an update for Standalone Networking? alegacy?
15:09:03 <JayF> I don't think there's anything for Eventlet removal. I'll note Julia and I were both interviewed for https://removal.eventlet.org/guide/testimonials/ with our (written) interviews up there now.
15:09:16 <alegacy> Yes... I've started to open some reviews for some of the pre-requisite type changes required, including a change to the spec to discuss some relatively minor changes to state machine required
15:09:34 <JayF> Please make sure they are hashtag ironic-week-prio and I'll have a look post-meeting
15:09:44 <alegacy> given that the release is imminent I don't expect these to get accepted right away.
15:09:47 <JayF> Thanks for your work in this area
15:10:09 <JayF> https://github.com/openstack/ironic/tree/stable/2025.2 has already been cut
15:10:09 <alegacy> currently working on trying to finish getting OVS included in the bifrost setup for testing of this feature.
15:10:16 <JayF> so really there's no reason we can't start merging stuff for next()
15:10:20 <alegacy> perfect
15:10:34 <JayF> although I'd expect contributors to mostly be focusing on things other than reviews, I think looking at code that's going to be a focus at PTG is a wise move
15:11:30 <JayF> Giving another minute for any additional standalone networking or eventlet update.
15:11:50 <alegacy> No further update from me on standalone networking
15:13:11 * dtantsur pretends he hasn't just read "OVS in bifrost" :D
15:13:14 <JayF> No discussion topics, skipping that agenda item.
15:13:27 <JayF> #topic Bug Deputy Updates
15:13:32 <JayF> cid: anything?
15:14:06 <cid> I triaged a few bugs, but none from this week
15:14:16 <cid> So, 0 new bugs and 0 new RFEs
15:14:19 <JayF> Thanks. Anyone want to volunteer for next week?
15:14:29 <JayF> It's been quiet. Good opportunity for a new person to try, perhaps?
15:14:31 <cid> I'd be happy to
15:14:36 <cid> Indeed
15:15:01 <JayF> #note CID to serve again as bug deputy
15:15:12 <JayF> There are no RFEs to review
15:15:17 <JayF> #topic Open Discussion
15:15:20 <JayF> I have an item for this
15:15:38 <JayF> There are several things in Ironic that are pending retirement:
15:15:43 <JayF> 1) Metalsmith (next cycle)
15:15:47 <JayF> 2) Inspector (next cycle)
15:15:51 <JayF> 3) SNMP (next cycle)
15:16:10 <JayF> This is a last call for 1) anyone interested in saving these at the last minute and 2) anyone who wants to point out more stuff we need to get rid of
15:16:25 <JayF> I'm going to spend some time early this cycle, just post release process, running the paperwork for all this to try and lighten our load.
15:17:00 <dtantsur> When you say "next", is it 2026.1 or .2?
15:17:17 <JayF> I meant 2026.1, although looking at it, I think SNMP won't be eligible yet: https://docs.openstack.org/releasenotes/ironic/2025.2.html#relnotes-31-0-0-stable-2025-2-deprecation-notes
15:17:32 <JayF> although given the technical burden, I'd suggest we may want to break the rule around slurp-yness a little
15:17:42 <JayF> unless someone really wants to get in the weeds with upgrading pysnmp now that eventlet is gone
15:18:18 <dtantsur> Well, we may dip our toes into asyncio soon(ish), so who knows
15:18:40 <JayF> I will say this: I don't think GR-OSS will be investing more time in it.
15:18:47 <dtantsur> very fair
15:18:55 <JayF> If RH won't either, then any "maybe" is sorta disingenuous
15:19:02 <dtantsur> I remember.. /someone/ said they would fix it
15:19:07 <JayF> not saying that to be mean, just want us to be honest about the likelihood it gets some <3 or not
15:19:23 <dtantsur> Well. RH has interest in asyncio'ing some parts of drivers for the sake of high parallelism in the sensor collection.
15:19:34 <dtantsur> We don't care much about snmp, metal3 does not even support it.
15:19:55 <dtantsur> But once some asyncio is there, maybe the migration to new pysnmp will be much easier? Purely speculating.
15:20:13 <JayF> AIUI NobodyCam and friends at nvidia are the only active community members who use SNMP
15:20:25 <dtantsur> Yep, I'm politely hinting at them ;)
15:20:32 <dtantsur> I thought StackHPC also used it?
15:20:41 <JayF> dtantsur: I think likely migrating Ironic will be easy; migrating VirtualPDU will be painful as heck
15:21:23 <dtantsur> We could leave the SNMP driver without CI. It's terrible, but we have a lot of code that is not really covered by the CI.
15:21:24 <JayF> dtantsur: if there's any StackHPC folks in here, we should ping them. Isn't p riteau StackHPC?
15:21:42 <JayF> In the past we've seen, specifically PySNMP library break in quiet ways without ci
15:21:53 <dtantsur> cc priteau
15:22:24 <dtantsur> Yeah, true. I'm not going to defend it, to be honest. My life is all Redfish nowadays.
15:22:30 <JayF> maybe I should just unleash claude on the problem
15:22:43 <JayF> dtantsur++ similarly, and redfish itself is getting stranger by the minute
15:22:51 <dtantsur> Oh it surely is
15:23:14 <dtantsur> I'm still recovering from your "System is not necessarily a computer" on the PTG etherpad :D
15:23:19 <dtantsur> (or was it someone else?)
15:23:22 <JayF> System may be "N" computers :)
15:23:28 <JayF> or more like
15:23:29 <JayF> 1/N computers
15:23:50 <JayF> multiple system objects, one chassis
15:23:53 <dtantsur> It kinda doubt it was the initial intention. I'm really curious what DMTF people think about such an interpretation.
15:24:37 <JayF> I mean, I'm in a world where I have the gear, we have to provision it, and I did maximum-leaning on the hardware providers
15:24:51 <JayF> so DMTF might help long term, but it won't help folks escape the problem
15:25:31 <dtantsur> Fair. Anyway, let's table it until the PTG.
15:25:51 <JayF> ++
15:25:59 <JayF> So back to my original statement
15:26:09 <JayF> Metalsmith/Inspector gone as soon as Flamingo release process is done
15:26:14 <JayF> SNMP maybe gets a short reprieve
15:26:30 <JayF> I might take a swing at getting ironic unit tests happy and then kill functional test CI
15:26:43 <JayF> to at least get us unblocking requirements project even if it still dies in a cycle
15:27:39 <dtantsur> ++
15:27:39 <TheJulia> Ugh
15:27:48 <TheJulia> Sorry, been on a call which ran way over
15:28:35 <JayF> Anything to add Julia or generally anyone have anything else for open discussion?
15:28:59 <priteau> dtantsur: were you pinging me about SNMP? I don't think we use it anywhere.
15:29:12 <JayF> good to know thanks priteau
15:29:18 <dtantsur> thx!
15:31:10 <JayF> Last call for Ironic meeting.
15:31:38 <TheJulia> I haven't even had a chance to read the meeting
15:31:43 <TheJulia> so... I guess just close it out
15:31:48 <JayF> TheJulia: I mean, we can hang out for 5-10 :)
15:31:55 <dtantsur> TheJulia: we just assigned all bugs to you
15:31:58 <dtantsur> nothing spectacular
15:32:47 <TheJulia> dtantsur: cool cool, thanks <3
15:32:57 <dtantsur> you're always welcome <3
15:33:19 <TheJulia> it seems like our release need to be re-sorted since we're blocked on version number nitpicking
15:33:43 <TheJulia> Is everyone looking at me to deal with that?
15:33:55 <JayF> I can if needed, but even if you or I do it release managers will have to re-approve
15:34:01 <JayF> ideally the release manager JFDI so it doesn't take two people
15:34:20 <TheJulia> they have already blocked it because of it so they have to re-approve regardless
15:34:21 <JayF> and as far as I care, bump the version number to the moon if we need to lol
15:34:36 <dtantsur> the other way around: reduce the number :)
15:34:38 <JayF> TheJulia: if a release liason DPL pushes the change, it auto PTL+1
15:34:43 <TheJulia> Yeah, on the list of importance its like... so low down.
15:34:47 <JayF> TheJulia: otherwise release liason has to +1 it to get PTL+1
15:34:51 <JayF> so yes we can skip a step
15:34:52 <TheJulia> JayF: yeah, that would be ideal
15:34:58 <dtantsur> who's that, iurygregory?
15:35:14 <TheJulia> I think so.... but I could be wrong?!
15:35:41 <dtantsur> the downside of the DPL model: who the hell has agreed to do what exactly? :D
15:35:53 <TheJulia> Yeah, I have a ptg topic regarding that concern
15:35:54 <TheJulia> ;)
15:36:02 <TheJulia> but thats a whole other stream of discussion
15:36:26 <TheJulia> (totally get it, we're all absurdly busy, its easy to miss stuff too)
15:36:28 <JayF> rpittau: is release liason
15:36:37 <TheJulia> TIL!
15:36:42 <dtantsur> Riccardo back tomorrow, I believe
15:36:43 <JayF> iurygregory: is events so theoretically should be planning PTG
15:36:51 <JayF> I am security, Doug is TC-liason, Dmitry is CI
15:36:54 <JayF> (tact-sid)
15:37:01 <TheJulia> ... (we need backups for each role)
15:37:03 <dtantsur> yep, I remember mine
15:37:17 <dtantsur> by the way, yes. backups are actually needed, even if it's another person from the same circlwe
15:37:22 <TheJulia> ((anyway, we can discuss that at the ptg))
15:37:31 * JayF is curious if the model-itself supports that
15:37:35 <JayF> I might ask in advance of PTG
15:37:41 <TheJulia> JayF: I *think* it does, but that is a good question
15:38:09 <dtantsur> Realistically, the liaison model was to a certain extent a backup model for PTL's
15:38:14 <dtantsur> if we regressed on that, we need to fix it
15:38:45 <JayF> OK well if this is a topic for PTG
15:38:51 <JayF> and rpittau isn't around for the release fix r/n
15:38:56 <JayF> I will look at those patches
15:39:09 <JayF> Link 'em here when they are updated
15:39:16 <JayF> Any other topics we wanna hit while the meeting is open?
15:39:33 <TheJulia> I don't think so
15:41:05 <JayF> Thanks all
15:41:06 <JayF> #endmeeting