15:00:11 #startmeeting ironic 15:00:11 Meeting started Mon Apr 20 15:00:11 2020 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:12 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:13 o/ 15:00:14 o/ 15:00:15 The meeting name has been set to 'ironic' 15:00:15 hi dtantsur, wrt https://storyboard.openstack.org/#!/story/2006910 -- was it approved? (I saw a PR associated with it and was going to review it but... ? ) 15:00:21 \o 15:00:26 o/ 15:00:27 o/ 15:00:31 o/ 15:00:32 o/ 15:00:34 \o 15:00:48 Greetings everyone! 15:00:55 rloo: I think we've become pretty relaxed with approvals. that being said, this work should wait for V to beging. 15:00:56 Our meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting 15:01:02 hi 15:01:11 #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting 15:01:17 rloo: I think it was sort-of-approved-at-least-nobody-objected 15:01:30 dtantsur: ok :) 15:01:31 #topic Announcements / Reminder 15:01:59 \o 15:02:11 #info This is the end of the ussuri cycle as we will likely need to be ready to cut our stable branch this week or MAYBE early next week at the absolute latest. 15:02:21 o/ 15:02:24 o/ 15:02:30 o/ 15:02:53 #info TheJulia needs to understand who is is intending to attend the virtual PTG, we will discuss this more later in the meeeting 15:03:01 Does anyone else have anything else to announce or remind us of? 15:03:12 o/ 15:03:20 I guess we're in sort of feature freeze now? 15:03:39 At least we'll need to freeze API/RPC at some point to be able to update the release mapping 15:04:25 o/ 15:04:27 Essentially yes, so I'd like to propose we review and maybe merge any outstanding changes that would be good to merge that may impact API/RPC, so we can update the release mappings in the next couple of days 15:04:41 I'm intending to write a prelude later today and revise any confusing release notes 15:04:52 I don't think we have API changes that are realistic to merge in Ussuri 15:04:59 * rpittau following from cellphone as my isp decided to die 15:05:41 looks like you blocked the one RPC change that I know of off the top of my head that made sense 15:05:51 heh :) 15:05:58 well, I see no big point in merging it without everything else 15:06:08 and delaying till V will get us a bigger margin of error 15:06:23 I'd love to see https://review.opendev.org/#/c/687910/ in, but *shrugs* 15:06:24 patch 687910 - ironic - Add `network_data` field to ironic node object - 30 patch sets 15:06:38 I'd apply same argument. We have no chances of finishing the feature in Ussuri. 15:06:41 No way to know if we would be able to get the rest of the stack merged in time 15:06:45 yeah 15:06:47 So landing API that is not backed by anything... 15:06:53 yup 15:07:02 Traditionally, we have been trying to land API changes *last* 15:07:09 so that the backend is in place by the time they're exposed 15:07:12 etingof: ^^^ 15:07:31 +1 15:07:39 Indeed, although tradition doesn't make sense in all cases 15:07:57 Anyway, we should move on 15:08:08 I should be worried that I'm becoming too conservative? :) 15:08:33 heh 15:08:45 I would say maybe =) 15:09:01 * dtantsur prepares to yell at cloud 15:09:09 dtantsur: I'm afraid you've already been doign that 15:09:15 Considering the ci status also I doubt anything new will land in time 15:09:20 we just didn't issue you a cane to shake at the cloud 15:09:23 lol 15:09:31 Anyway, moving onward 15:09:32 yeah, where's my cane kids? 15:09:39 #topic review action items from the prior week 15:09:39 Merged openstack/ironic-ui master: Stop configuring install_command in tox. https://review.opendev.org/720813 15:10:01 Well, it was more like two weeks ago 15:10:02 * iurygregory joins dtantsur to yell 15:10:18 and iurygregory I believe you sent an email regarding cancelling the meeting, so I think were good! 15:10:30 TheJulia, yup 15:10:42 #topic Review subteam status reports 15:10:48 #link https://etherpad.openstack.org/p/IronicWhiteBoard 15:11:02 Line 248 15:11:31 so deploy steps work going to next cycle correct? 15:12:17 not fully, I hope 15:12:22 I saw there was an outstanding uefi raid patch, but it seems like its done 15:12:27 at least one patch has to land now 15:12:38 #link https://review.opendev.org/#/c/721236/ is technically breaking 15:12:38 patch 721236 - ironic-python-agent - Add raid.apply_configuration deploy step - 1 patch set 15:13:51 Well, I guess it is a good time to merge that then 15:14:11 * TheJulia adds to list 15:14:38 TheJulia: this one? https://review.opendev.org/#/c/720781 15:14:39 patch 720781 - ironic - Don't break UEFI install with older IPAs - 1 patch set 15:14:42 seems reasonable 15:15:50 interesting 15:16:02 Is anyone else aware of patches that must merge? 15:16:14 tzumainn: all done w/r/t multitenancy? 15:17:25 TheJulia, yep! as far as I know :) 15:17:29 we're keeping six for sushy-tools for the time being right? 15:17:39 tzumainn: you owe us a demo video ;) 15:17:44 +++++ 15:17:58 +1,000,000 15:18:38 dtantsur, hahaha, I can do that 15:18:55 that would be amazing (and you can show it on your yearly performance review!) 15:18:57 TheJulia: re sushy-tools six, yes, for now 15:19:41 we can drop already no? =) 15:19:58 On a plus side, looks like we got a number of awesome things merged this year 15:20:00 well 15:20:01 cycle 15:20:09 we need to make sure we consume sushy-tools from releases on stable branches 15:20:15 same change I did for vbmc 15:20:15 I'd rather wait for the next cycle to be honest 15:20:17 I'm good with proceeding whenever you all are ready. 15:20:45 it's low priority either way, let's concentrate on the release for now 15:20:51 yep 15:21:14 ok, I will check the stable branches to make sure we are using sushy-tools from releases 15:21:23 * TheJulia feels like we drifted into priority setting anyway 15:21:42 iurygregory: it boils down to removing it from required-projects/$PROJECTS 15:21:49 I think. I hope. 15:22:07 dtantsur, I hope too =) let's see 15:23:19 Are we good to proceed? 15:23:52 I think so 15:24:06 #topic Priorities for the week 15:24:30 So I believe the priorities are kind of simple. The must land patches, prelude, and release mapping update. 15:25:43 Do we want to change up CI dramatically right before we release? 15:25:54 iurygregory: this is in regards to your job changes 15:26:13 TheJulia, you mean the l-c changes? 15:26:44 I guess the standalone job? 15:26:50 oh 15:26:52 iurygregory: standalone chagnes 15:27:04 well we merged the change in ironic-tempest-plugin 15:27:19 if we don't mind not testing that it's fine? 15:27:32 I guess we do release the plugin... 15:27:49 but I'm assuming that since we split we could add a job to check 15:28:23 https://review.opendev.org/720675 this adds the job in ironic 15:28:23 patch 720675 - ironic-tempest-plugin - Add standalone redfish jobs - 1 patch set 15:28:32 ops in tempest-plugin 15:28:41 and https://review.opendev.org/712042 in ironic 15:28:42 patch 712042 - ironic - Add ironic-standalone-redfish - 7 patch sets 15:28:55 check job non-voting and then make it gate/voting after the stable branch is cut? 15:29:06 ++ sure =) 15:29:09 will update 15:29:10 ++ 15:29:33 TheJulia: about CI, we should also land the last dib conversion patches 15:29:58 rpittau: links? 15:30:07 or topic link I guess :) 15:30:21 https://review.opendev.org/#/q/topic:ironic-convert-dib+(status:open+OR+status:merged) 15:30:39 the -1 are because of the recent ci issues 15:30:51 Iury Gregory Melo Ferreira proposed openstack/ironic master: Add ironic-standalone-redfish https://review.opendev.org/712042 15:31:00 okay, I guess we're just going to have to wait and see on those. 15:31:05 yep 15:31:31 I think it is if they make it awesome, if not it is still okay 15:31:35 Anything else? 15:32:06 Iury Gregory Melo Ferreira proposed openstack/ironic-tempest-plugin master: Add standalone redfish jobs https://review.opendev.org/720675 15:34:00 I guess not, Are we good to proceed to discussion 15:34:44 ++ 15:35:30 #topic Discussion 15:35:55 First topic, is PTG Attendance. I would like to get an idea of those that will be interested in attending and what time zones would be best for themselves. 15:36:15 * dtantsur Central Europe 15:36:20 or that they can cover. I suspect, after all these years, I could kind of guess at least at the later question, but wanted to put it out there for everyone. 15:36:41 Central Europe 15:36:43 * arne_wiebalck Central Europe 15:37:27 eastern time zone 15:37:30 Europe 15:37:31 * iurygregory Central Europe 15:38:13 eastern 15:38:36 I guess another question and moderately unrelated is are there any topics that people are interested with regards to the scientific sig, edge working group, etc? 15:39:31 a status check with the edge group may be useful 15:39:48 there may be some discussions with ansible SIG re our modules 15:40:00 not sure if it's interesting for everyone though 15:40:10 not sure if there will be a discussion about Goals for V... 15:40:20 I've been going back and forth in email with Ilidko and they are hoping to have some ironic discussion around ironic 15:40:36 iurygregory: what do you mean? 15:40:48 I mean, aside from https://review.opendev.org/720100 :) 15:40:49 patch 720100 - ironic-specs - WIP - Victoria Cycle Priorit(y|ies) - 1 patch set 15:40:54 iurygregory: there certainly will, but here's a thing to keep in mind 15:41:07 we no longer have to fit our internal topics into the same week 15:41:18 it's pretty expected that teams may set up discussions outside of it 15:41:21 TheJulia, just to *avoid* overlap if possible 15:41:23 at least that's the impression I've got 15:41:32 dtantsur, yeah 15:41:48 the week should be dedicated to questions requiring a wider audience 15:42:04 like cross-project interactions, reaching out to the ops community, and so on 15:42:08 same, although I have concerns about requirements gathering and the feedback loop being disjointed 15:42:24 we can have something even before AND after 15:42:51 I kind of like that idea, even if it is just an hour sync-up or call to discuss priorities again after opendev 15:42:59 I'm just pointing out that we have more flexibility and we should use it 15:43:05 Agreed 15:43:06 rather than only suffer from downsides :) 15:43:10 ++ 15:43:31 the week of the PTG will be cramped. many teams will meet in the same EU-US timeslot 15:43:33 apparently 15:43:47 Yeah :\ 15:43:53 3pm UTC or something like that 15:44:25 * dtantsur returns the mic to Julia 15:44:32 :) 15:44:41 Oh, I'm about to give it right back :) 15:44:58 oh? 15:45:24 Well, that discussion answered the questions I needed input on, I'll see what I can do! 15:45:38 ah, right. I think we chatted about http://lists.openstack.org/pipermail/openstack-discuss/2020-April/013757.html previously already? 15:45:57 we may talk about potential follow-ups though 15:46:32 Yes, and there has been some additional discussion. That being said I think the victoria cycle priorities draft is a good place to continue the discussion amongst the contributors 15:46:49 ++ 15:47:18 of immediate follow-ups I see: a potential web site, rethinking release schedule, reducing (hard) dependencies, making the docs less openstack-centric 15:47:36 but we probably should shake off Ussuri first 15:48:00 ++ There is some additional discussion occuring but I can't comment on it... yet :) 15:48:14 ++ 15:48:17 * dtantsur is intrigued :) 15:48:46 dtantsur: attend open board meetings, 'tis all I can say 15:48:52 definitely docs less openstack-centric. regardless of anything else :) 15:49:04 TheJulia: oh nope :D 15:49:11 or more docs about standalone 15:49:15 ++ 15:49:24 we have one page that is not ideally structured 15:49:42 we *may* even use 'baremetal' instead of 'openstack baremetal' 15:50:00 * dtantsur will compose a list of potential actions that don't involve drastic political changes 15:50:10 dtantsur: +1,000,000 15:50:31 seems like a reorg of this content: https://docs.openstack.org/ironic/latest/, some things will be common to both openstack & standalone, others won't. 15:50:41 yeah 15:50:47 ++ 15:50:59 ++ 15:51:25 Anyway, docs and topic presentation dives deep into perceptions and it is one of those discussions _should_ likely save for a social hour or soemthing 15:51:36 Speaking of which, we should do another SPUC! 15:51:42 totally 15:52:08 Anyway, I guess we're good to stop over at the Baremetal SIG if arne_wiebalck has anything he wants to talk about, and then move to open discussion. 15:52:22 +++ 15:52:37 II SPUC \o/ 15:52:42 nothing from me, please have a look at the white paper and contribute or review :) 15:52:48 ++ 15:52:59 More text keeps flowing into it, it is kind of awesome! 15:53:07 Moving to open discussion then! 15:53:12 #topic Open Discussion 15:54:00 I did add some philosopical/whimsical thoughts to the end of the white paper and went through the authorship history 15:54:10 hey folks, we have a patch that we think is going to be important to get in: https://review.opendev.org/693328 15:54:10 patch 693328 - ironic - DRAC : idrac-redfish inspect updates pxe port - 11 patch sets 15:54:12 at least, what was visible :( \o/ for anonymous edits 15:54:30 TheJulia: thanks 15:54:51 cdearborn: I'm a little hesitant to merge features until you guys merge the sushy-oem-idrac changes and cut a release 15:54:52 i have a quick question wrt the adopt process, does anyone know why it's validating instance_info/image_source ? 15:54:58 cdearborn: https://review.opendev.org/#/c/710327/ needs to be merged AND released first 15:54:58 patch 710327 - x/sushy-oem-idrac - Export System Configuration - 2 patch sets 15:55:08 kaifeng: it doesn't have to 15:55:15 kaifeng: to be able to rebuild 15:55:21 TheJulia, dtantsur, good point 15:55:37 kaifeng: but yes, as dtantsur just pointed out, so rebuild could be supported. If you want to create a knob for it, I think it would be easy to merge that 15:56:04 I wanted to point out that the redfish driver doesn't set pxe_enabled on the ports in ironic, which will cause major problems with nova if more than 1 port is enabled on a baremetal server 15:56:05 kaifeng: generally, the fewer special cases, the easier it is to reason about a state of a node 15:56:11 and by knob, I mean configuration parameter because I've had the same question asked elsewhere recently but in a more subtle way 15:56:29 Rebuild is also not really widely used afaik 15:56:34 TheJulia: I'm soft -1 on a new parameter. if we think it's better without the check - let's just remove it. 15:56:46 oh, we have some case that adopt existing nodes, then perform routine tasks like power control, or even monitoring 15:56:46 yeah, we overestimated the importance of rebuild back then 15:56:58 dtantsur: I think that is reasonable 15:57:17 so lets remove the requirement \o/ 15:57:18 cdearborn: good point :( I wonder if etingof has any plans about it 15:57:36 3 minute warning :) 15:57:55 * etingof can't read that quickly 15:58:14 dtantsur, unfortunately, i don't think there is a generic way to get the pxe ports via redfish, so not sure if there's anything that can be done in the redfish driver itself 15:58:17 * arne_wiebalck would love an ironic-aware openstack server rebuild 15:58:31 arne_wiebalck: it is ironic-aware Oo 15:58:37 etingof: no worries 15:58:45 I mean.. we even tested it in the CI at some point 15:58:45 * etingof has a silly question 15:58:46 dtantsur: including s/w RAID 15:58:51 * kaifeng is in the same both with etingof 15:58:56 arne_wiebalck: oh. let's discuss afterwards 15:58:59 *boat 15:58:59 is it too late to revert anything? 15:59:11 I noticed a patch lately that has quite some issues to follow up. I've worked on it the whole today, but now I think it would be way easier to revert than to follow up.... 15:59:12 etingof: never too late, depends on what and why 15:59:28 this patch https://review.opendev.org/#/c/677996/16 15:59:28 patch 677996 - ironic - Generalize ISO building for virtual media driver (MERGED) - 16 patch sets 15:59:28 a storyboard story is a good start 15:59:41 let's make sure we document reasons outside of IRC 15:59:43 has multiple problems 16:00:13 a story on the discovered problems? 16:00:13 Nothing is perfect and the intent was in the right direction 16:00:24 etingof: yes 16:00:41 ++ lets detail why 16:00:41 fixing these problems will be quite a time sink 16:00:47 okay 16:01:25 TheJulia, dtantsur: thanks, will think more about it :) 16:02:28 Thanks everyone! 16:02:42 One last question before I end the meeting 16:02:47 Do we want to have another SPUC this friday? 16:02:55 or the following week? 16:03:00 can be this week 16:03:15 SPUC? 16:03:31 cdearborn: Sanity Preserving Un-Conference 16:03:38 2nd edition 16:03:39 i.e. an ironic happy hour 16:03:48 ahhh - lol - +3 16:03:53 * dtantsur hopes he remembered the abbreviation right 16:04:00 close enough :) 16:04:17 The idea is get people to talk about things they are interested in and there may be coffee, tea.... or other tasty beverage 16:04:18 s/Preserving/Preservation 16:04:30 Oh, and we did silly hats 16:04:39 we need something new silly for this round :) 16:04:42 JayF totally won with his video overlay silly hat of silliness 16:04:49 oh yeah :D 16:04:54 ++ 16:05:04 best hat 16:05:06 * TheJulia delegates the silliness topic 16:05:17 best hat could work! 16:05:31 Most well dressed? 16:05:48 oh, I'm going to lose :D 16:05:56 heh I lose before even starting 16:05:57 * dtantsur will bet on TheJulia 16:06:00 3 16:06:12 TheJulia: we all know you have awesome dresses! 16:06:39 dtantsur: this is true... 16:06:55 Well, you guys decide on something! 16:06:59 and with that, thanks everyone! 16:07:02 Have a wonderful week! 16:07:03 thank you TheJulia 16:07:05 #endmeeting