Thursday, 2022-08-11

*** pojadhav|out is now known as pojadhav|rover02:49
*** tosky_ is now known as tosky09:33
fungiin trying to explain the tested runtimes for zed to someone, it dawned on me that our documented explanation for how we choose distro versions is at odds with what we're actually doing13:24
fungioh, or maybe not. was ubuntu 20.04 chosen for zed because 22.04 didn't exist officially until a couple of weeks into the cycle? looking harder at the schedule i guess that was the case after all13:26
fungistrangely, it looks like most devstack jobs for zed are running on 22.04 (ubuntu-focal) nodes13:31
fungieven though the pti says 20.0413:31
fungiso if we list 20.04 because it's required for upgrade testing from yoga, we should probably make the overlap policy clear in the pti pages13:33
fungii'm happy to push updates to those docs if we can get to the bottom of what it's supposed to say13:33
*** pojadhav|rover is now known as pojadhav|afk14:19
gmannfungi: I am not completely getting it. zed cycle runtime is Ubuntu 20.04 and that is what our most of testing it. 22.04 (jammy) is target for 2023.1 cycle 14:46
gmannrecently we added a new job on Ubuntu 22.04 (jammy) in devstack to be prepared for the next cycle14:47
gmannit seems doc and current testing are in sync. 14:51
gmanntc-members: weekly IRC meeting in an 9 min from now.14:51
clarkbfungi: focal is 20.04 not 22.0414:59
fungigmann: oh, thanks yes, 20.04 is focal. i was entirely confusing myself14:59
fungiso antelope will probably be jammy (22.04) with upgrade testing starting on focal (20.04)14:59
gmannyes15:00
gmann#startmeeting tc15:00
opendevmeetMeeting started Thu Aug 11 15:00:56 2022 UTC and is due to finish in 60 minutes.  The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot.15:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:00
opendevmeetThe meeting name has been set to 'tc'15:00
knikollao/15:01
rosmaitao/15:01
gmann#topic Roll call15:01
gmanno/15:01
spotzo/15:01
diablo_rojoo/15:01
gmannIn absence section:15:01
gmannarne_wiebalck (will miss 4 August and 11 August)15:01
gmannslaweq (will miss 11 August)15:01
gmannJay Bryant (jungleboyj) -- Will miss 8/11 due to Out of Office15:01
dansmitho/15:01
gmannlet's start15:02
gmann#link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda_Suggestions15:02
gmann^^ today agenda15:02
gmann#topic Follow up on past action items15:02
gmannno action item from previous meeting15:02
gmann#topic Gate health check15:03
gmannany news on gate health ?15:03
gmannI have not noticed any frequent failure this week15:03
gmanndansmith: rosmaita did you? or anyone else15:04
dansmithnope, haven't noticed any major problems this week15:04
rosmaitawe got hit by the update to flake8-logging-format, but fixed it15:05
rosmaitanot sure a lot of people use that extension15:05
gmannrosmaita: may be we should pin them in hacking, I remember we faced some flake8 extensions issue in past too15:06
gmannbut good to know it is fixed15:06
rosmaitait just hit the cinder pep8 jobs in most branches15:07
gmannok15:07
gmannwe can see if any extension is frequently causing failure on updates then we have option to pin it in hacking 15:07
rosmaitaok15:08
gmannbare recheck 15:08
gmann#link https://etherpad.opendev.org/p/recheck-weekly-summary15:08
gmannslaweq is not present today. we can skip it unless anyone else want to discuss someting on it?15:09
gmann#topic 2023.1 cycle PTG Planning15:10
gmannAs we all know that Oct PTG is not virutal. 15:10
fungiis now virtual15:11
gmannthere will be some update from diablo_rojo or someone from foundation on team/slot signup things?15:11
gmannsorry s/not/now :)15:11
diablo_rojoYes. 15:11
gmanndiablo_rojo: thanks. 15:11
diablo_rojoEssentially I just updated the survey15:11
gmannlet's wait for that to start the TC PTG planning15:11
diablo_rojoYou can use the same link15:11
diablo_rojoas before15:12
diablo_rojoDeadline is the 26th15:12
gmanndiablo_rojo: ok, but it will be good to send in on ML as projects also wanted to get more clarity that old team signup is needed or new one will come15:12
gmannat least that is what QA team was waiting. 15:12
diablo_rojoYes, I plan to do that, but I figured since I was here you can go ahead for TC15:13
diablo_rojoand now QA15:13
gmann+1, cool15:13
gmann#action gmann to fill PTG survey and team signup for TC discussions 15:14
spotzYeah I'm not sure I signed up D&I so maybe a mail of who's signed up already?15:14
*** pojadhav|afk is now known as pojadhav15:14
diablo_rojoIt is signed up already15:14
gmannI think we should ask eveyrone to signup again and ignore the old data as that signup was as per in-person meeting15:14
diablo_rojoI think with you as the modderator spotz 15:14
spotzSweet thanks:)15:14
fungiyeah, the original deadline was tomorrow so i had put off signing up the security and tact sigs because i procrastinate15:14
fungisorry diablo_rojo!15:15
diablo_rojogmann: I contacted all of the people that had signed up already to confirm their signups were still valid. Its been handled. 15:15
gmannand total attendance in survey or any other answer might be different for virtual PTG15:15
diablo_rojoPlease just use the same formstack survey to say if the team wants to attend. 15:15
gmannok if old data still valid for virtual PTG then ok15:16
gmannin yesterday call with Board members, one idea came up to include more operators and ops interaction in PTG.15:17
gmannidea is that project should plan a 'operator hours' slot in PTG where we can have operator to discuss about feedback etc.15:18
gmannas TC we should encourage projects to schedule 'operator hours' as a separate slot in PTG15:18
gmannand those 'operator hours' per project should avoid conflicts among other projects 'operator hours' slot15:19
diablo_rojoAnd that it would be good being coordinated by the TC as an invitation to operators. A show of good faith.15:19
spotzOnce we have a schedule for that I can send out via OPS Meetup twitter15:19
gmannthanks15:19
diablo_rojospotz: that would be super helpful- I think we are a ways from that, but definitely good to keep in mind!15:19
gmannonce we will have slot booking open, we will send this on ML and to PTL15:20
gmannanything else on PTG planning ?15:20
rosmaitai think we (or someone) need to propose what slots should be used by projects for operators15:21
rosmaitaso that we don't all schedule the same time15:21
spotzhehe15:21
gmannrosmaita: yeah, we need to take care of the conflict 15:21
diablo_rojorosmaita: Yeah I guess I am just imagining an etherpad that the TC puts together with a list of slots available each day and the projects signup for them? 15:21
diablo_rojoThat way the TC can keep an eye on it. 15:22
rosmaitai was thinking that maybe we suggest for the "dirty six" and the other projects that already expressed interest to diablo_rojo have a time they can use if they want15:22
gmannhow about asking projects to use Monday or Tuesday to use the operators hours ? those are the day most of the projects are not having their project specific slots15:22
diablo_rojoLOL15:22
diablo_rojo"dirty six"?15:22
rosmaitathats what mugsy calls the primary projects15:23
gmannany other day Wed to Friday is difficult to avoid conflict 15:23
diablo_rojogmann: sure, in that case- since the schedule will be the same as the other virutal PTGS- someone can setup the etherpad right now15:24
rosmaitai thought the idea was to use project specific slots so that the operators can meet with the dev team?15:24
spotzI might be a little in and out on Monday but then around the rest of the week 15:24
gmannrosmaita: yes, I mean giving at least one project specific slot on Monday or Tuesday where developers will be there. but if any other day work for project that is all ok15:25
gmannanyways let's have slot open and I will put this on ML and also to avoid conflict. and how project book then we can discuss if any conflcit 15:26
gmannwe were pretty good for cross project sessions in virtual PTG so the operator hours can also be arranged in coordinated way. like nova has operator hours at 10 AM wed so neutron can plan it some other time15:27
dansmithgmann: we should try to spread things out I think,15:27
dansmithso that there's a potential for all those projects to have a non-overlapping ops slot15:27
rosmaitadansmith: ++15:28
diablo_rojorosmaita: thats my thought too, but I think gmann wants to push them to the front of the week15:28
gmannyeah15:28
dansmithmaybe in the schedule, we could reserve two slots per day as "for ops" and have projects sign up for one?15:28
gmanndiablo_rojo: that was one idea but I think if TC can prebooked slot on every day and then project can choose them15:28
diablo_rojoI don't really care one way or the other about when they happen- we just need visibility as to when they are going to happen and to avoid collisions15:29
gmanndansmith: yeah, indeed 15:29
dansmithI think specializing days of the week (as we have in the past) is less good,15:29
dansmithbecause it ends up stacking similar things on fewer days15:29
gmannsure15:29
rosmaitai guess it could work either day, have ops day and each team comes for "nova time', "cinder time" , etc. or spread them out and have "ops time" at nova, "ops time" at cinder, etc15:29
rosmaitai mean "work either way"15:29
diablo_rojogmann: but then the schedule will say 'TC' every day and not just the project name?15:29
dansmithrosmaita: makes it pretty intense for ops for that day15:30
dansmithbut maybe they'd prefer that I dunno15:30
gmanndiablo_rojo: no, we can do what dansmith suggesting 'for ops' and ask project to fill it15:30
rosmaitayeah, not sure15:30
dansmithand if days are only four hours long, then that makes it pretty small slots for more than a few projects15:30
diablo_rojoSo the openstack service teams are going to go to the ops space then? and not the other way around?15:31
gmannI think reserving 2 slot per day should work. so at least 8-10 slots15:31
dansmithgmann: ++15:31
dansmithdiablo_rojo: I would expect the ops hour for nova to be in nova's "room" no?15:31
rosmaitaok, so ops teams would come to the dev team rooms15:31
*** pojadhav is now known as pojadhav|out15:31
dansmithespecially if we're trying to encourage the pattern of ops coming to dev things15:31
gmannyeah15:31
rosmaitadansmith: ++15:31
diablo_rojodansmith: thats what I was thinking15:31
knikollai think it might be easier to bring the dev teams to the ops room, than bring the ops to the teams room15:31
TheJuliamaybe going to the ops is better symbol of collaboration15:32
diablo_rojoThe operators would join the nova team or whomever15:32
knikollaconsidering the ops will have to switch rooms every other hour15:32
TheJulias/is/is a/15:32
diablo_rojoknikolla: especially since they arent as used to the whole virtual ptg thing?15:32
diablo_rojoNot as practiced anyway15:32
diablo_rojoTheJulia: +215:32
TheJuliaYeah, that would be a concern is getting lost and giving up15:32
gmannswitching room in zoom is not that hard15:33
dansmithyeah, I mean.. come on :)15:33
knikollait is an active action15:33
TheJuliafor someone with the context of doing so every so often15:33
gmannknikolla: just one times in a day15:33
knikollagoing to their room will allow passive participants to also participate15:33
TheJuliaknikolla++15:33
diablo_rojogmann: but using the bot and discovering the rooms might be a higher barrier to entry then they are familiar with15:34
gmannI think for virtual PTG, it is less concern about what room is for what. project also switch room many times15:34
dansmithknikolla: same argument works the other way right? ops get exposed to more than just the few devs that are going over to hear the ops?15:34
diablo_rojoSince it will effectively be their first virtual PTG15:34
knikollai imagine the dev group is smaller and more tightly knit. 15:34
gmanndiablo_rojo: humm, it is not that hard to learn :)15:34
diablo_rojogmann: its not, but we also have the luxury of having done it several times at this point. 15:35
diablo_rojoJust being devil's advocate. 15:35
dansmithI dunno, if we're trying to encourage the ops to feel welcome in dev-centric things in the future (like when we are in person) it seems like inviting them into the dev rooms when the bar is low sets the right tone15:35
gmannyeah that is what my impression is.15:35
fungidiscovering rooms should hopefully not be that hard regardless of virtual or physical, we publish it to a web page anyway15:36
dansmithobviously if clicking on a different zoom link is so onerous that it won't happen, then it's not that important, but I thought from the call that we're trying to mechanically connect *and* set the right tone15:36
knikollamaking them feel as guests vs. them running the show15:36
dansmiththat they are "part of the team"15:36
gmannI think  we are over thinking  for them :)15:36
diablo_rojoI do think it would be good to get the operators exposure to as many people in all the teams as possible and the best route, optimizing for that- is to get them into the services rooms15:36
TheJuliathey don't see the bar as low though, they see it as tight knit and in some cases impervious to their context15:36
dansmithknikolla: we discussed on the call them still running the hour, I don't think they need to be in their own room to drive the agenda15:36
diablo_rojogmann: quite possibly15:36
gmannthey might be saying yes it is all good15:36
diablo_rojoTheJulia: +215:36
TheJuliaWe need to build mutual context, so anything where we expect people to already have context is not easy and is just creating barriers15:37
dansmithTheJulia: by a low bar I meant "clicking on a zoom link" versus "fly to a site, walk into a room full of people they don't know"15:37
diablo_rojodansmith: except the in person format is what they are used to, so that is easier for them15:37
gmannspotz: can you check with them if switching zoom room once or twice in a day is barrier for them to join PTG ?15:37
TheJuliastill a room full of people they don't know at the end of the day15:37
dansmithdiablo_rojo: you really have a low opinion of these people it seems :)15:37
diablo_rojothe ptgbot site and having to play a zoom hop game is not anything like what they have done before. 15:38
diablo_rojodansmith: not at all15:38
spotzgmann Yeah I can't think it would be especially if we're not one on top of the other. If we are I'd think it makes sense to just use one main one15:38
gmannwe can ping them the zoom link in IRC or in ML and they just join it like any ohter meeting they join15:38
dansmithbut whatever, the important part is the collab. if "which zoom link I clicked on" defines who's "turf" it is, then we might be in more trouble than I thought15:38
diablo_rojoI just have the context of having actually been to meetups before and also having seen new teams come into the virtual PTG and struggle with the PTGbot interface + irc 15:38
gmannspotz: ok, conflict we will avoid for sure15:38
knikollaat least to me, it's more of a psychological barrier. i have avoided going into places where i have no shared context / not know a lot of people. and when i do, i tend not to speak up.15:39
TheJuliaknikolla: exactly15:39
diablo_rojobecause if they aren't used to using IRC regularly, the bot, while well documented and clear in my opinion, is not super easy. 15:39
knikollai have found that to extend to zoom meetings as well.15:39
dansmithknikolla: that's part of the thing we're trying to break down for the future right?15:39
diablo_rojoknikolla: +215:39
gmannknikolla: idea is to improve such interaction and knwoing peoplr15:39
dansmithops staying corralled in their own place and then simultaneously feeling left out?15:39
gmannwe are trying to make PTG for them same as ops meetup, you are a separate entity and separate room/people.... 15:40
dansmithright15:40
gmannas spotz mentioned it is not issue for most of them I think we should try to welcome them in developers room and see how it goes15:40
spotzThey should be welcome and invited everywhere, but if we have back to back meetings it just makes sense to have a set room folks go to. Kinda like in Speed Mentoring it's quicker for a mentor to change tables then all the mentees15:40
TheJuliaIt goes back to what knikolla said. We need to build shared context and trust first15:41
dansmithspotz: they wouldn't be back to back though.. two slots per day15:41
spotzdansmith: Ahh then yeah they can open a new zoom when the time comes:)15:41
knikollalet's ask them what would they prefer.15:42
gmannyeah, that way they feel like yes 'we attended developers meeting and it was all easy and good, we will talk more with them'15:42
TheJuliaknikolla: +215:42
dansmithlet's please just move on.. we've argued about who owns which zoom turf for a lot of the meeting.. if its critical that we show up to the ops rooms, then fine. I think it doesn't progress us towards a better situation in the future,15:42
gmannok.15:42
fungialso if there are back-to-back nova+ops and nova meetings, but also back-to-back nova+ops and cinder+ops meetings, someone's room-hopping either way15:42
dansmithbut neither does this conversation :)15:42
gmannspotz: can you check with them on ML or someway how they prefer?15:42
diablo_rojoSo theoretical question: Nova and Cinder are going to meet at the vPTG, how do you decide who goes where?15:42
rosmaitathe ptls fight it out15:42
diablo_rojorosmaita: I would like to see that15:43
TheJuliarosmaita: this is a t-shirt opportunity15:43
fungitwo teams enter, one team leaves15:43
dansmithdiablo_rojo: yeah we don't argue15:43
rosmaitait usually depends on who is recording15:43
spotzgmann Yeah just let me know secifically what the ask is. Changing Zoom rooms, back to back vs non-back to back meetings?15:43
dansmithnor do we consider it to be either one's turf when we switch :)15:43
rosmaitaanyway, we can schedule the times for ops/teams meetings and decide later on where15:43
TheJuliafungi: two teams enter the collaboration-atron15:43
gmannspotz: I think we are (like any other discussion) interpreting them in all differnet way, and we really need few of them to talk for what they want/like15:43
spotzOk I'll do and email and a twitter poll15:44
gmannspotz: basically two option 1. join ops hours in the project PTG room and meet/discuss with developer 2. book ops slot in same room/day and ask developer to come to that room and discuss the things15:44
dansmithTBH, I'm less interested in a poll or asking them, because what we're trying to do is draw them out of their shell(s)15:45
dansmithyou don't ask a hermit if they want to go outside,15:45
dansmithyou show up outside with ice cream and try to lure them into the sunlight :)15:45
diablo_rojoAlso, I would like to just point out how much of a struggle organizing anything is because all of our teams are SO different, so, in the future please remember this conversation when something about how things are organized isnt exactly perfect for everyone :)15:45
diablo_rojo(getting meta for a moment :) )15:45
rosmaitadansmith: that sounds like another t-shirt opportunity15:45
gmanndansmith: agree but I am seeing lot of different voice on their behalf so it is good to ask them directly 15:45
spotzIce cream!15:45
gmannspotz: adding action item for you15:46
dansmithsee, I just got spotz' attention with ice cream. it works.15:46
diablo_rojodansmith: so thats what we need to do for you? :D15:46
TheJuliadansmith: "yes, we'll fix that bug" might work better :)15:46
TheJuliaAt least, virtually :)15:46
spotzI think some of the main feedback from Berlin was that just sending 1 OPS wasn't helpful as they're all different in their needs. I think as long as we minimize crossover sessions we'll be ok15:46
dansmithTheJulia: yeah, its just as black and white as all these bugs that are straightforward but we don't fix because we're evil right? :)15:46
gmann#action spotz to check with operators about what format they want in virtual PTG. options: 1. join ops hours in the project PTG room and meet/discuss with developer 2. book ops slot in same room/day and ask developer to come to that room and discuss the things15:47
gmannanything else on PTG things than the ops slot format?15:47
gmannbased on what spotz finds out, we will discuss and finalize it next meeting15:47
TheJuliadansmith: eh... I've always kind of let the harder opinions roll off my shoulders as I've tried to build mutual context.15:48
TheJuliadansmith: but that doesn't prevent us from humming "when your evil" in the background...15:48
diablo_rojoI would also make sure, spotz, that they know that even if they go to the projects meeting rooms, they can also still have their own dedicated time and space15:48
TheJuliadiablo_rojo: ++15:48
spotzYeah I'll frame it for the Dev/ops interactions15:48
gmanndiablo_rojo: yes, and even they can be moderator if they want15:48
diablo_rojoJust because they pick option 1, doesn't preculde them from having their own track in addition15:48
gmannbut joining developers room will improve the interaction to community 15:49
diablo_rojogmann: correct. 15:49
diablo_rojo+215:49
spotzYep which is why we were hoping ops folks would come for the whole week15:49
dansmithspotz: when you poll,15:49
gmannspotz: yes15:49
dansmithyou might say that the developers want more ops interactions *in their project rooms* in general15:49
spotzYep15:50
gmann+115:50
diablo_rojospotz: happy to help with phrasing15:50
spotzok I'll run it by you before posting diablo_rojo most likely later today15:50
diablo_rojospotz: sounds good- I'm around :_) 15:50
diablo_rojo:)15:50
gmannspotz: if you would like, you can add the sentence in etherpad also in case we miss anything ? but up to you15:51
gmannlet's move to the next topic15:51
spotzok15:51
gmannthanks 15:51
gmann#topic Open Reviews15:51
gmann#link https://review.opendev.org/q/projects:openstack/governance+is:open15:52
gmannneed one more review in the python testing template update #link https://review.opendev.org/c/openstack/governance/+/85246315:52
gmannother than that, diablo_rojo any consensus on 'Environmental Sustainability SIG' ? #link https://review.opendev.org/c/openstack/governance-sigs/+/84533615:53
diablo_rojoWe are moving forward being a foundation level wg 15:53
diablo_rojoI just need to abandon the patch and put other things in motion15:54
gmanndiablo_rojo: ok, thanks for the updates.15:54
fungiwill there be a ptg session?15:54
diablo_rojoNo problem15:54
diablo_rojofungi: I certainly hope so :) 15:54
fungithanks!15:54
gmannother than that all other reviews are in good shape, either waiting for author or deps15:54
rosmaitaalso, don't forget this patch that's not in that search:15:54
rosmaita#link https://review.opendev.org/c/openstack/project-team-guide/+/85067515:54
gmannrosmaita: ack. thanks 15:55
gmannthat is all from agenda today, we have ~5 min left. anything else to discuss ?15:55
rosmaitagmann: for https://review.opendev.org/c/openstack/governance/+/852463 , do we do rollcall or code-review on that?15:55
fungirosmaita: i suppose we could approve 850675 but at the moment it's only got +2 votes from foundation staff, which is probably why it hasn't been yet15:55
rosmaitaoh, ok15:56
gmannrosmaita: rollcall is fine15:56
gmannI will review the 850675 today15:56
gmannfungi: as you are here, please check this project-config change needed for openstack-helm-addons retirement  #link https://review.opendev.org/c/openstack/project-config/+/85185715:57
gmannthat is all for today, thanks everyone for joining. 15:58
fungithanks15:58
gmann#endmeeting15:58
opendevmeetMeeting ended Thu Aug 11 15:58:07 2022 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:58
opendevmeetMinutes:        https://meetings.opendev.org/meetings/tc/2022/tc.2022-08-11-15.00.html15:58
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/tc/2022/tc.2022-08-11-15.00.txt15:58
opendevmeetLog:            https://meetings.opendev.org/meetings/tc/2022/tc.2022-08-11-15.00.log.html15:58
spotzThanks everyone!15:58
diablo_rojoThanks15:59
gmannfungi: saw few comments from you on 850675, you want them to be fixed or i can  approve it as it is and we can fix them later in follow up if needed?16:02
fungigmann: yeah, i tried to make that clear in my review comment, "I'm good with merging this as a faithful replica of what's on the wiki, and then we can iterate with additional changes to the document once merged."16:04
fungithat helps separate our changes to the original text from our importing of it16:04
gmannfungi: got it, thanks16:05
opendevreviewMerged openstack/governance master: Add Python3 zed unit tests  https://review.opendev.org/c/openstack/governance/+/85246316:07
*** gibi is now known as gibi_pto16:24
opendevreviewMerged openstack/project-team-guide master: Add Legal Issues FAQ  https://review.opendev.org/c/openstack/project-team-guide/+/85067516:26
*** tosky_ is now known as tosky18:16

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!