15:02:27 #startmeeting ironic 15:02:27 Meeting started Mon Sep 16 15:02:27 2024 UTC and is due to finish in 60 minutes. The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:27 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:27 The meeting name has been set to 'ironic' 15:02:30 I ran it last week :) 15:02:32 #chair TheJulia 15:02:32 Current chairs: JayF TheJulia 15:02:49 Hi all, welcome to the Ironic meeting, we hold our meetings under the Openinfra CoC 15:02:57 #topic Announcements/Reminders 15:03:01 o/ 15:03:10 #note 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 15:03:38 o/ 15:03:44 #note This is R-2 week. Next week final releases must be cut and the integrated OpenStack release is early october. 15:03:51 #link https://releases.openstack.org/dalmatian/schedule.html 15:04:04 Which means we need to cut *this* week 15:04:07 o/ 15:04:28 TheJulia: the linked thing seems to indicate we could do it next week 15:04:34 TheJulia: but I am very +1 to getting it done in advance 15:04:45 JayF: they will force us for this week most likely, it has happened nearly every time 15:04:50 anxiety overrules all 15:05:01 That's my motto 15:05:03 lol 15:05:17 So lets cut one this week then, sgtm 15:05:37 #note the next OpenInfra PTG which will take place October 21-25, 2024 virtually! Registration is now open! 15:05:37 #link https://ptg.openinfra.dev/ 15:05:48 #link https://etherpad.opendev.org/p/ironic-ptg-october-2024 15:06:07 Please add your topics to the etherpad, and review them async so we can keep the PTG live sessions to the point and on topic as much as possible. 15:06:42 That's all on the agenda for announcements 15:06:49 #topic Review Ironic CI Status 15:07:01 Re releases: rpittau comes back on ?Wednesday? 15:07:13 Re CI: we put out SOOO many fires last week 15:07:16 yeah 15:07:26 Our CI has bitrotted significantly, in a handful of different ways 15:07:34 Broken inspector should be fixed by https://review.opendev.org/c/openstack/ironic-inspector/+/928783 15:07:40 then... hopefully.... 15:07:40 we wanted to fix some of it with a replacement for ipa; we didn't do that this cycle 15:07:56 there's a lot of random failures like Jens alludded to earlier- including downloading cirros from the internet 15:08:25 I feel like CI can consume infinite time, but between releases I think it'd pay off to try and tackle as many of these intermittant issues we can 15:08:51 I'm wondering if we need to instead of looking at CI during hte ptg through the lens of fix/reduction but to radical revisioning, or determination of what we can do to make it easier for ourselves to have more stable/reliable CI 15:09:00 note, that is a big picture thought, we don't have to discuss now() 15:09:08 I agree a lot 15:09:15 including documenting our specific CI matrix, what jobs cover what 15:09:25 * dtantsur has deja vu 15:09:46 dtantsur: I said similar in the existing ptg notes 15:10:00 and in the past.. and in the past.. and in the past 15:10:07 We have a few topic on PTG for this: replacing dnsmasq to help remove that periodic CI failure, training up folks on how to troubleshoot CI so it's less of a strain on the longer term folks 15:10:12 all this has happened before, all this will happen again 15:10:20 * dtantsur nods wisely 15:10:22 dtantsur: we've been doing this for ten years, I'd be amazed if we don't do this every two years 15:10:30 and I'm not sure it's a bad thing :D 15:10:31 * TheJulia channels https://en.battlestarwikiclone.org/wiki/Pythia 15:11:48 Looks like we'll have a handful of things to chat about CI during PTG 15:11:57 Discussion topics in the meeting agenda are old; am not going to cover those 15:12:00 #topic Bug Deputy updates 15:12:15 cid and masghar had volunteered to be deputies over a couple of weeks, at least when the agenda was updated 15:12:22 is there anything to note? 15:12:33 I forgot to update the agenda, apologies 15:12:34 * JayF notes the agenda was last updated 8/26/2024 so that promise is beyond expired 15:12:39 No new bugs for September 15:12:43 ++ Thanks 15:12:57 Do we have a volunteer to be the next bug deputy? 15:13:03 No new bugs 15:13:16 I can probably take it if nobody else is signing up? 15:13:20 I haven't in a month+ 15:13:39 masghar, we can both keep an eye until after release (?) 15:13:50 I wouldnt mind it at all :) 15:13:56 ++ 15:13:58 I'll be around if you have any questions 15:14:05 Perfect, thanks 15:14:06 #note CID and masghar agree to bug deputy thru release 15:14:06 JayF: ty 15:14:24 #topic RFE Review 15:14:31 there are no RFEs posted for review 15:14:35 #topic Open Discussion 15:14:52 braaaaains 15:15:11 cardoe has some items requiring attention 15:15:20 I can take a look after the meeting 15:15:21 I have a small thing for here, honestly probably a plug more than anything else -- I made a short video about OSSA-2024-003, https://www.youtube.com/watch?v=cz71C4tW3Pw -- feel free to point anyone at it who has questions 15:15:31 TheJulia: I landed most of them already I think 15:15:54 JayF: what I linked? That was me scrapping about 20 minutes ago. 15:16:04 Yeah, I've been busy 15:16:23 Anything else for open discussion? 15:16:32 uhhh 15:16:38 Availability in weeks leading up to the PTG> 15:16:39 ? 15:16:43 I'll note it's unlikely I'll be here a week from today 15:16:48 as it's my birthday and I usually try not to work 15:17:22 JayF: Happy birthday in advance 15:17:38 ++ ^ 15:17:49 As an FYI, the first whole week in october I'll be in the EU time zone for meetings. The week after that I'll be in Indiana for OpenInfra days 15:18:00 The week after that i the PTG 15:18:02 I might be out at right around that time, too 15:18:09 * JayF checks calendar 15:18:33 Yeah like 10/3 - 10/9 or so 15:18:39 of further note, the week after the ptg, I have jury duty too 15:18:43 My brother is getting married :) 15:19:01 ** Oct 3 - Oct 9 (I know that was a us-ian way to write that) 15:19:33 I'll be out for a few days around Oct 1 15:19:51 #note Many cores out first two weeks of October; please set expectations appropriately for getting code landed 15:20:10 ++ 15:20:15 Last call for items for Open Discussion? 15:20:28 Uhh 15:20:47 OpenInfra Days NA - Indiana, about a month from now() 15:21:00 I will have a project update/feedback session, if anyone else will be there 15:21:15 o/ 15:21:25 o/ kubajj 15:21:27 o/ 15:21:45 TheJulia: nice; if you want I can try to attend remotely but will not be on site 15:22:09 I don't know if they are recording forum sessions 15:22:26 since I'm sort of splitting the time into high level what is going on and then requirements gathering 15:22:56 Yeah, that's what I figured, just tossing out the willingness out there 15:23:11 I know having you and dtantsur on the line when rpittau and I did BM SIG @ CERN was super helpful 15:23:37 (and I think we've incorporated much of that actionable feedback into this release \o/) 15:24:03 I'll see what I can do, but won't know likely until the day of 15:24:15 Merged openstack/ironic master: Remove default override for RBAC config options https://review.opendev.org/c/openstack/ironic/+/928283 15:24:18 You know how to get ahold of me if you need :D 15:24:29 \o/ 15:25:19 Last call (part deux) for open discussion items? 15:25:37 I can haz a review request? 15:26:04 the nc-si / no-power-off spec is sad and lonely: https://review.opendev.org/c/openstack/ironic-specs/+/926654 15:26:05 dtantsur: we can haz cheezeburger? 15:26:21 mmmm do want 15:26:38 dtantsur: I can take a look in a day or two, I need to get some other stuff sorted first 15:27:11 dtantsur: I have some concerns but I think the only way to resolve them is "don't support this hardware" which isn't reasonable 15:27:23 dtantsur: so I'll take a look 15:27:45 I just independently had a crazy idea without looking at the spec, I wonder how similar it might be?! 15:27:48 Concerns should at least be listed, mitigated and documented 15:27:52 ++ 15:28:00 Yeah, mainly I think there are cases we can leak a running ramdisk 15:28:07 The spec is kinda boring, so a crazy idea might turn out something completely different 15:28:15 which are ~unavoidable, in the same way the "we flip the power and hope the machine comes up" is unavoidable 15:28:17 JayF: for network booting? 15:28:29 like think abuot 15:28:34 what happens if the agent completely goes away 15:28:45 the "timeout" bug gets 100000x worse if you can't reboot via oob somehow 15:28:53 that goes from "a deploy fails" to "a human must touch this machine" 15:29:09 You can reboot, you cannot power off 15:29:12 oh yeah, possible, I was more thinking these cases are likely vmedi abased 15:29:17 dtantsur: you can reboot /via oob/? 15:29:34 Yes, sure. Only power off is a problem. 15:29:36 dtantsur: if so that will obviate like 90% of my concerns 15:29:37 I think we always try inband first 15:29:49 okay nice, I was thinking the inband method too 15:29:54 We do try in-band power off first, which will also need to be changed to a reboot 15:30:01 oh, yeah 15:30:04 ugh 15:30:06 so it's not like the interface to power off doesn't exist 15:30:09 or maybe smarter?! 15:30:13 it's like, power off for this hardware isn't a usual thing to do? 15:30:16 Please check the spec, it answers all these questions :) 15:30:19 changing running default across a fleet is slightly problematic 15:30:21 I've read the spec 15:30:21 ack 15:30:30 * TheJulia read a much earlier version, I think 15:30:34 probably same 15:30:37 * TheJulia needs to refresh the brain 15:30:48 Oh, have we posted our prelude? 15:31:08 Yeah it does not specifically say it's oob, I think there are some unsaid assumptions in there, I'll give it a quality review once meeting is over 15:31:13 There is a morale imperitive for a silly prelude 15:31:17 TheJulia: nope, that's another thing on my list for today if nobody else pushes it 15:31:18 Happy to elaborate wherever needed 15:31:20 I am not a silly writer 15:31:28 if you want a boring prelude I will write one 15:31:42 okay, I'll see what I can hammer out after my next meeting 15:31:45 * JayF prefers boring for docs that will be read by multiple cultures / esl folks 15:32:04 I don't trust myself to know what is funny for everyone :) 15:32:08 That might be better 15:32:15 * dtantsur cannot English well enough for a funny prelude (at least the one that most people will find funny) 15:32:33 * TheJulia thinks dtantsur underestimates his english skills 15:32:35 I was going to suggest we highlight the operator-facing enhancements: 15:32:38 same 15:32:42 or German has replaced htem 15:32:52 like 2-3 features spceifically asked for at BM SIG have landed this release, ironic-guest-metadata, runbooks 15:32:58 we have a lot of operator-happiness changes in this one 15:33:12 TheJulia: I do have a mess in my head nowadays :D 15:33:16 btw, I discovered a bug (that I caused) in the RAID skip_block_devices and am working on fixing it 15:33:29 Did docs for runbooks merge? I ask because I had someone send me a feature list and it lacked it 15:33:48 kubajj: oh, good to know. Is there a bug in launchpad? 15:34:06 cid: ^ re: runbook docs 15:34:07 I think they did? 15:34:13 I think they were alongside 15:34:41 TheJulia: I'll have a look, not sure 15:34:53 nothing on runbooks in https://docs.openstack.org/ironic/latest/admin/ 15:34:58 JayF: yes, unless something was left out, it was all part of the same patch. 15:35:01 and that is where I have folks looking for "features" 15:35:41 nor https://docs.openstack.org/ironic/latest/user/ 15:35:45 I bet it is just not linked in 15:35:48 Unless objection, going to close the meeting as I think this is rapidly flowing into normal conversation? 15:35:50 me too TheJulia 15:36:02 JayF: sgtm 15:36:07 #endmeeting