15:59:49 #startmeeting tc 15:59:49 Meeting started Wed Mar 8 15:59:49 2023 UTC and is due to finish in 60 minutes. The chair is knikolla[m]. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:59:49 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:59:49 The meeting name has been set to 'tc' 16:00:17 gmann: just to make you be late I started it 1 minute early :) 16:00:26 :) 16:00:29 #topic Roll call 16:00:30 o/ 16:00:31 o/ 16:00:32 o/ 16:00:35 o/ 16:00:39 Oj 16:00:39 \o 16:00:55 o/ 16:02:52 I'm removing discussion on TripleO deprecation and PyPI maintainer list to give us more time to talk about leaderless projects and the vPTG planning. Any opposition? 16:03:07 nope 16:03:11 I think we need to decide on TripleO deprecation 16:03:25 ++ 16:03:43 we have all the info we need and it just need to take decision 16:03:49 Yeah, I'd say to avoid confusion we'd better to deprecate master early 16:03:58 +1 16:04:14 ++, we have all the info and we discussed extensively. We just need to formalize a proposal that we can decide on. 16:04:22 Anyone volunteers to take that on? 16:04:33 For example - we in OSA kind of depend on that as we'd love to drop out tripleo CI jobs from our master jobs 16:04:39 That we have for some repos 16:04:45 I can do but let's discuss in today meeeting 16:04:51 Late but here 16:05:08 gmann: thanks. Keeping the item in agenda. 16:05:12 thanks 16:05:22 #topic Deciding on meeting time 16:05:31 This week I’ll be sending out a poll to vote on a new meeting time for the TC weekly meeting. 16:05:40 This will be effective from the first week in April. So, after the vPTG and after all daylight savings changes, so please vote accordingly. 16:05:50 This will also give us a bit more time to try to make something work for everyone. 16:06:02 That wasn't the case last time unfortunately. 16:06:10 knikolla[m]: I'm quite fine with this time at the moment 16:06:27 why should you want to modify the time ? 16:06:39 bauzas: we need to check time as we have new TC member|s 16:06:41 I wasn’t good with it and with the time change next week this time is a disaster 16:06:50 bauzas: it's conventional when there are new members to see if the time still works for the majority 16:06:52 bauzas: We have new TC members. And schedules of people change. 16:06:59 okay 16:07:16 We need to get rid of daylight savings:( 16:07:24 bauzas: it can end up with same or new time depends on majority of availability 16:07:45 spotz: well, all the upstream meetings use UTC in general 16:07:46 spotz: That's also why I'm delaying it to April for the new possible meeting time. 16:08:12 spotz: so everyone should just know about the daylight savings for every tz 16:09:07 Bauzas other orgs and companies don’t 16:09:40 You're missing out on the fun if you don't know all the tzs. 16:10:16 But because our meetings are in UTC, people need to be aware of DST in their own TZ only. 16:10:17 spotz: look at https://meetings.opendev.org/ 16:10:28 all of the meetings are using UTC 16:10:41 I am fully aware of that page bauzas 16:11:05 So starting from next week the TC meeting would be at noon ET, rather than 11am, which is today. 16:11:28 Anyhow, I'm moving on to the next topic. Please keep an eye on for your IRC pings. 16:11:28 9 AM for me, little better 16:11:29 it's 11 there? 16:11:43 oh right, sorry my finger math is wrong 16:11:48 dansmith: yep, 11am in blacksburg, too 16:11:49 Time in DC, yes. 16:11:51 lol 16:11:59 knikolla[m]: that's why in general the meeting chair pings like 1 hour before for making sure people remember about the time 16:12:02 it's been a long morning :D 16:12:21 dansmith: do not check EU time :) 16:12:28 bauzas: which i missed last time, but did today :) i'm getting better. 16:12:48 anyhow. 16:12:49 #topic Gate health check 16:12:59 Always a fun topic. Any updates? 16:13:05 Things are incrementally improving still 16:13:13 not seen much frequent failure this week 16:13:16 it seems to be much better, at least from my experience 16:13:19 still plenty of fail, but life is worth living again 16:13:31 multiple projects seem interested in the mysql memory footprint flag 16:13:35 ++ 16:13:36 slaweq: did you guys enable that somewhere? 16:13:44 and also number of rechecks before patches are merged are going down 16:13:52 ah, excellent data point 16:13:54 For us it's way better, we've almost didn't do rechecks last week 16:13:55 we broke stable gate with jsonschema constraints in Tempest but not it is reverted and green 16:14:24 dansmith yes, I proposed patch but it's not merged yet https://review.opendev.org/c/openstack/neutron/+/876556 16:14:30 *now it is reverted 16:14:32 slaweq: ack 16:14:36 +1 16:14:50 and seems that it works pretty good so we will go with this patch most likely 16:15:07 cool, we were going to flip that to default at some point 16:15:07 we've been a bit constrained on job capacity at heavier times of day, but have been working on trying to tune some things to squeeze a little more out of problem regions or get them back online to help with the volume 16:15:38 people have definitely been observing some backlogs on getting test results though 16:15:50 Ghanshyam proposed openstack/election master: Fix setup_election_config for combined election events https://review.opendev.org/c/openstack/election/+/876443 16:17:15 anything else, or anything that requires a decision? 16:17:29 nothing else from me 16:17:35 nothing from me 16:17:44 nope 16:17:46 I haven't seen other issues yet this week 16:17:52 Awesome, great work on getting things to improve! 16:17:53 like from Tempest 16:18:07 #topic 2023.2 cycle Leaderless projects 16:18:12 #link https://etherpad.opendev.org/p/2023.2-leaderless 16:18:22 We have 7 projects without any candidates: Monasca, Rally, Sahara, Swift, TripleO, Vitrage, Winstackers. 16:18:22 And 3 projects with late candidacies. Charms, Senlin, and Mistral. 16:18:44 Mistral is basically moving from DPL to PTL 16:18:50 Please vote on the 3 patches linked in the etherpad for the late appointments. 16:19:22 And please provide comments on the projects without any candidacies. Otherwise our default behavior will be to switch to DPL. 16:19:47 I think we need to check with team before moving to DPL 16:20:05 because most of them might not have any volunteer so retirement might be the option for them 16:20:15 Yeah 16:20:27 but yes, we need to check their situation and decide on available options we have in etherpad 16:20:47 Please write your name on the above linked etherpad if you volunteer to reach out to any one of the teams above. 16:20:56 but now a days, PTL missing is not because no one want to be PTL it is because there is no active maintainer left 16:21:18 Or if you propose a different path of action. 16:21:34 End of week I will reach out to all the teams that don't have a volunteering TC to reach out to them. 16:21:47 Seems sensible? 16:21:54 +1 16:22:32 knikolla I will try to help with that and will put my name on some of those projects 16:22:34 +1 16:22:45 slaweq: fantastic! thank you. 16:23:20 #topic Deprecation process for TripleO 16:23:28 #link https://lists.openstack.org/pipermail/openstack-discuss/2023-February/032083.html 16:23:50 Last week we discussed extensively about this, but if I’m now wrong, we didn’t quite have decisions or proposals. 16:23:54 so as per latest information, it seems no volunteer to maintain stable/zed #link https://lists.openstack.org/pipermail/openstack-discuss/2023-February/032393.html 16:24:16 we are good for master deprecation but were waiting for stable/zed decision 16:24:43 I still think we should not drop stable/zed branch 16:24:45 regardless 16:24:46 I will propose to keep stable/zed open (do not mark deprecated) and if anyone come to take it then they can continue the maintenance 16:24:49 +1 with noonedeadpunk 16:25:02 noonedeadpunk: that can create confusion who using stable/zed 16:25:32 gmann: Um. You proposed exactly the same just in case 16:25:37 I will say let's keep it open from where it is today and anyone using it and want to maintain can help 16:25:44 ++, thank you for formulating a proposal. I also agree on keeping it open. 16:25:53 yup 16:25:56 noonedeadpunk: oh. sorry missed to read 'not' 16:26:13 We should at a minimum let the security sig know that there are no maintainers on it. 16:26:14 yeah so basically mark master deprecated, keep stable/zed open as it is 16:26:29 ++ 16:26:45 With some readme adjustment? 16:26:48 yeah we can inform then and if anything fail need fix we will say no maintainers and any help is welcome 16:26:59 +1 readme can be good place 16:27:02 That's really the only thing that we NEED to keep supporting. 16:27:29 Voting? 16:27:51 * bauzas stays silent so (but likes the readme thing) 16:27:56 i can propose the patches and we can vote on gerrit unless we want to vote here ? 16:28:02 Does this require a vote on IRC or a vote on a proposal to governance? 16:28:13 gmann: I prefer Gerrit. 16:28:22 Well, branches removal will be releases patch 16:28:25 gerrit will be good IMO 16:28:27 not governance 16:28:56 since I think this is unprecedented, I'm not sure which procedure to apply, but I too am fine with gerrit :) 16:29:06 noonedeadpunk: there will be needed-by / depends-on on governance patch also and we can see how it goes. 16:29:14 I think a resolution? 16:29:15 that is same for any deprecation process 16:29:22 no resolution is needed i think 16:29:24 yes, for master. 16:29:38 But should we conclude that they should not touch zed? 16:29:43 for stable/zed we are just updating readme right? 16:30:09 Yeah, but there we won't have rollcall or anything 16:30:24 yeah 16:30:29 So I'm just unsure if gerrit is enough to mandate keeping zed 16:30:41 in that case let's do voting on IRC for stable/zed as formal agreement 16:30:50 isn't that a matter of declaring a branch on EM ? 16:30:52 and I guess that's the biggest question we had 16:31:00 noonedeadpunk: you have good point, there are many stackholder like tripleO team release team etc 16:31:50 if we keep zed (which I'd like), what would be the status of its branch from a project perspective ? 16:32:07 bauzas: that's the whole question. Zed is not EM and we do EM kind of coordinated 16:32:07 bauzas: great question. was just about to ask that to formalize the voting question. 16:32:19 bauzas: we discussed the EM option but stable/wallaby for tripleO is not EM i think and it can create more confusion having stable/zed EM but stable/wallaby open 16:32:34 bauzas: but they're having independent release cycle so they should not have stable/zed at all. But since they do... 16:32:39 wallaby is not EM? 16:32:51 I'm quite OK with saying Zed is "Maintained" without contributors 16:32:53 I thought it was, they were just really M'ing it a lot :) 16:32:56 Even if it will officially not be EM, technically it will be in that state anyway 16:33:02 I mean for TripeleO I need to check 16:33:02 which is the actual situation 16:33:05 yeh, true 16:33:09 hence the README 16:33:10 wallaby is already em 16:33:21 but I wanted to clarify the branch state 16:33:30 for TripleO also ? they want to keep support as maintained i think 16:33:30 It will be maintained beause we need to push out security fixes, unless I'm misunderstanding our commitment to releases. 16:33:49 then I'm cool with the outcome 16:33:55 Well, they can keep maintained in EM :) 16:34:05 but my information on stable/wallaby need more clarification. anyways stable/zed can be just open to anyone to maintained 16:34:15 ++ 16:34:17 gmann: projects transition branches to em state together. that's how it's designed. the stable/wallaby branches of all projects are no longer receiving normal maintenance 16:34:45 fungi: right but Tripleo is special where independent release can have stable/zed 16:35:00 that doesn't make it officially maintained state though 16:35:03 But on W they were not independent yet 16:35:06 so I am not sure how they are maintaining stable/wallaby as EM or full maintenance 16:35:12 stable/wallaby is no longer officially maintained by openstack, full stop 16:35:40 let's vote on stable/zed things and master deprecation is already under our defined process 16:35:45 right, so only EM by the tripleo team right? 16:35:57 some people from red hat are patching stable/wallaby of tripleo under extended maintenance 16:36:03 There's maintained by the team and downstream, and there's maintained by us as a coordinated release body 16:36:04 yeah 16:36:12 those are separate things. 16:36:35 I'm cool with stable/wallaby being EM and stable/zed being Maintained (as other projects do), provided we set a README clarifying the exact worldcrisis discussion 16:36:44 yeah 16:38:03 That works for me. Keeping stable/zed as maintained, and providing in the readme details on talking to the TC or security team, since there isn't a team anymore. 16:38:20 that works for formulating a vote here on IRC as the proposal? 16:38:44 then, I assume no vote is actually required, right? 16:39:02 As we're up with the regular process? 16:39:05 +1, we can add TripleO master will be deprecated as existing process 16:39:09 And no exception is made? 16:39:18 I think as long as it defines the constrained of what maintained means in that specific case for stable/zed, it should work from my pov. 16:39:24 that's just a bunch of contributors who decided to turn down their feet and the TC saying it's gonna write something in the repo to clarify the lack of contributors despite the project official status 16:39:26 bauzas: and..... you would be correct, I think. 16:39:54 so to be clear, 16:40:12 we're saying zed is supported/maintained which is what the tripleo team said the would not do, 16:40:33 yes 16:40:33 but we're hoping that if any CVEs come along, we'll be able to convince them to fix that, or hope someone else from the community will jump in and try 16:40:35 correct? 16:40:41 dansmith: thanks, that was the heart of the concern I was thinking of 16:40:48 that is why I want to say 'it is open to maintain but no maintainers for now' 16:40:55 yes, or us. considering the reputation we need to uphold for releases. 16:41:08 yeah, "maintained but no maintainers" is the short story I think 16:41:17 :) 16:41:18 yeah 16:41:24 or "supported but no maintainers" perhaps is better 16:41:24 I just wanted to clarify it 16:41:32 dansmith: I effectively see that as "the TC being the emergency maintainers" 16:41:45 humm 16:41:49 and hunting people down to do something about what crops up 16:42:04 and yeah, since there is no antelope series, people have to understand that the CVE fix they gonna land will first be on stable/zed 16:42:05 I think if anything come as urgent to fix and no volunteer then we can just retire it 16:42:09 saying it's being kept under stable/maintenance as a matter of policy because those branches were already created. there would need to be a deprecation policy change to make it possible to officially abandon them, but having nobody proposing or reviewing stable branch changes technically isn't that unique of a situation for an openstack project (unfortunately) 16:43:02 if a security fix for those stable branches becomes important to someone, the tc can give them access to approve such fixes 16:43:11 fungi: yeah true, we have some non-deprecated projects that are basically in that state I guess 16:43:33 how about this to vote on? "TrieplO master will be deprecated as normal process, keeping stable/zed as maintained, and providing in the readme details on "it is supported but no maintainers ". " 16:43:55 "TrieplO master will be deprecated as normal process, keeping stable/zed as supported, and providing in the readme details on "it is supported but no maintainers ". " 16:44:04 or "seeking interested maintainers" maybe 16:44:04 s/maintained/supported 16:44:28 gmann: wfm 16:44:46 that leaves us about 13 months to figure out whether this is a good outcome :) 16:44:52 gmann: i really want a "talk to TC" part in that readme 16:45:02 and in 13 months, Zed will be EM \o/ 16:45:09 but that works for me. 16:45:28 knikolla[m]: sure, or there will be tripelo PTL till it is retired so we have point of contact 16:45:36 How do I open a vote? I forgot the syntax. 16:45:43 startvote 16:45:56 with a question and a question mark with the answers 16:46:03 like startvote blah ? yes, bo 16:46:07 knikolla[m]: #link https://docs.opendev.org/opendev/system-config/latest/irc.html#voting 16:47:16 #startvote For TripleO: Follow the normal deprecation process for the master branch, keep stable/zed as supported, and update the readme in the repos to describe the lack of a team and to contact the TC for urgent matters? yes, no 16:47:16 Begin voting on: For TripleO: Follow the normal deprecation process for the master branch, keep stable/zed as supported, and update the readme in the repos to describe the lack of a team and to contact the TC for urgent matters? Valid vote options are yes, no. 16:47:16 Vote using '#vote OPTION'. Only your last vote counts. 16:47:35 #vote yes 16:47:36 #vote yes 16:47:39 #vote yes 16:47:45 #vote no 16:47:49 #vote yes 16:47:52 * bauzas abstains (despite his nod) 16:47:55 #vote yes 16:48:01 #vote yes 16:48:25 are we missing anyone? 16:48:37 Jay is absent 16:48:40 despite the clear majority. 16:48:47 those present here all voted 16:48:50 #endvote 16:48:50 Voted on "For TripleO: Follow the normal deprecation process for the master branch, keep stable/zed as supported, and update the readme in the repos to describe the lack of a team and to contact the TC for urgent matters?" Results are 16:48:50 yes (6): dansmith, spotz[m], noonedeadpunk, knikolla[m], slaweq, gmann 16:48:50 no (1): rosmaita 16:49:18 rosmaita: we can discuss after the meeting if you'd like about your disagreement? 16:49:31 rosmaita: did I miss you arguing for something else 16:49:31 ? 16:49:39 knikolla[m]: sure 16:49:56 as next step, I can propose the patches in governance and stable/zed or wait until rosmaita concern discussion ? 16:49:57 dansmith: no, i don't have any alternative proposal, i just don't like the precedent this sets 16:50:15 rosmaita: oh I don't like it either 16:50:22 but it's the practical thing to do 16:50:23 me too :) 16:50:34 heh, I don't see a lot of fun in this discussion 16:50:40 i had the luxury of "voting my conscience" because i knew i was going to lose! 16:50:41 damn, can I change my vote to be with rosmaita ? #vote yes-but-I-hate-it ? 16:50:45 and something to discuss in PTG to avoid such situation in future 16:50:46 haha 16:50:55 Hehe 16:50:56 It's situation we never should have found ourselves, but I assume honest mistake was made when stable/zed was created 16:51:08 true 16:51:26 all options were terrible, we chose the least worse. 16:51:35 alright. moving on. 16:51:35 can we maybe say an independent release model needs to not use a specific release name from the upstream cadence if so ?. 16:51:54 Yes, exactly this ^ I've also proposed 16:52:17 #link https://review.opendev.org/c/openstack/governance/+/875942 16:52:21 this isn't a quite-independent-skip-some-releases model AFAIK 16:52:25 thank you gmann 16:52:41 knikolla[m]: rosmaita you want me to go ahead on governance changes or wait more? 16:53:02 gmann: i think go ahead 16:53:03 I really want to close this task 16:53:13 rosmaita: ok thanks 16:53:14 But they deployed that version of OpenStack so in the case it made sense 16:53:29 I'm moving on directly to the vPTG agenda item. And skipping the 2 ones in between. 16:53:33 +1 16:53:40 #topic Virtual PTG Planning 16:53:49 March 27-31, 2023, there's the Virtual PTG. 16:54:18 #link https://etherpad.opendev.org/p/tc-2023-2-ptg 16:54:35 We need to decide on the amount of time 16:54:41 time 16:54:45 and agenda items 16:55:11 I think 4 hrs slots for two days work well in past PTGs 16:55:12 Please propose agenda items in the etherpad above 16:55:22 and 17-19 UTC slots on Thursday and Friday worked well last time. even ptgbot did not support those not in this PTG too 16:55:53 any other feedback on 17-19 UTC slots? 16:56:10 at least it worked fine for me to attend project specific discussion 16:56:39 which just to be clear is a 3 hour block (ends at 2000, not 1900) 16:57:12 makes for a long day, but i agree it worked well to participate in other stuff, and have other people join us 16:57:17 Thursday/Friday works for me 16:57:25 it will be a bit late for me after daylight change (22:00) but I will handle that if that works for others 16:57:38 rosmaita: I think it was 15-19 UTC 16:57:38 +1 ^ 16:57:45 not 20 UTC 16:58:01 #link https://etherpad.opendev.org/p/tc-2023-1-ptg#L18 16:58:17 * dansmith has a hard stop in 70 seconds 16:58:20 that's the problem with virtual meetings, you have to fight the time differences + the personal reasons 16:58:41 In last PTG, it ended at 19 UTC 16:58:43 pick anytime you want, I'll try to attend (hardly tho) 16:58:48 ok, so we are just talking about two two-hour blocks 16:58:58 gmann 19 UTC is better for me :) 16:59:30 I'm okay with the same times and days as last PTG. 16:59:31 rosmaita: I mean to say 15-16 UTC as normal slot and 17-19 UTC as extended slot where most of projects does not have their discussion schedule 16:59:41 15-17 UTC as normal 16:59:52 just be aware that scheduling discussion during the break defeats the purpose of having a break 16:59:58 and as a reminder, US and European daylight savings occur *before* the PTG, keep it in mind when you decide ;) 17:00:15 after US, before European 17:00:18 I think, no? 17:00:23 after EU 17:00:30 nope, europeans shift the sunday before the ptg 17:00:33 it will be just weekend before PTG 17:00:35 After both, then, cool! 17:00:36 +1 17:00:38 us changes this weekend, eu changes two weeks later 17:00:46 Alright, I'll schedule the times and send an email 17:00:54 knikolla[m]: one more thing but we are on time 17:01:12 do we want to schedule 'operator-hour' this time? 17:01:32 last time, nova operator hours were crickets. 17:01:42 despite heavy communication 17:01:44 gmann: great question, we can talk about that in the channel after the meeting. 17:01:45 or we can discuss it in next meeting as it might need more time to discuss 17:01:48 I'm closing it now. 17:01:51 knikolla[m]: sure 17:01:54 I did reach out to Kendall to maybe get a list of projects they want to meet with 17:02:00 thx, I have to go now 17:02:01 o/ 17:02:13 thanks spotz 17:02:18 #endmeeting