14:00:07 #startmeeting tc 14:00:08 Meeting started Thu Dec 5 14:00:07 2019 UTC and is due to finish in 60 minutes. The chair is evrardjp. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:11 The meeting name has been set to 'tc' 14:00:18 #topic roll call 14:00:22 o/ 14:00:23 o/ 14:00:24 #link https://media.giphy.com/media/2YtTdoSEl4m4/giphy.gif 14:01:03 o/ 14:01:09 o/ 14:01:11 oh hey, this is happening :) 14:01:37 exciting 14:01:46 yes indeed 14:01:54 O/ 14:01:58 one missing for quorum 14:01:59 woot 14:02:02 we have quorum! 14:02:21 just in case: tc-members the meeting has started! 14:02:36 let's first start with the follow up action items from previous meeting 14:02:42 #link https://media.giphy.com/media/xUA7b2z8NcgvyIthGU/giphy.gif 14:02:58 :) 14:03:00 #topic follow up previous action item (ricolin): SIG guideliness 14:03:01 \o 14:03:22 I meant guidelines ofc 14:03:31 evrardjp, we already got a patch up for that and looks might get landed anytime soon 14:03:31 https://review.opendev.org/#/c/695366 14:03:46 nothing else to do than reviewing? 14:04:27 I assume only needs final approve. And will try to provide it to new SIG and see if that guideline helps 14:04:56 that all 14:04:58 ok. Do you want to report here if that helped, at the next meeting? 14:05:06 or do you want to give a little more time? 14:05:29 report what? 14:05:32 In the latter, I will scratch that action item, and I will have that as a "longer term" action item 14:05:47 report on whether it helps or not :) 14:05:55 evrardjp, yes 14:05:57 sure 14:05:59 ok 14:06:03 let's move on then 14:06:06 #topic follow up previous action item (ttx): large scale sig 14:06:31 The group was created and had its first meeting 14:06:49 that's good news! 14:06:52 #link https://wiki.openstack.org/wiki/Large_Scale_SIG 14:07:03 is the initial attendance as expected? 14:07:04 Two short-term objectives have been defined 14:07:12 - Scaling within one cluster, and instrumentation of the bottlenecks there 14:07:19 - Document large scale configuration and tips &tricks 14:07:36 yes, pretty good turnout, essentially APAC and Europe for now 14:08:00 Now we'll see if the excitement holds 14:08:14 sounds good 14:08:15 I'll help drive it until it can fly by itself 14:08:33 Next meeting in two weeks 14:08:34 maybe jroll could help on the America's side? 14:08:36 ok 14:08:52 do you want to do a status report in a few months? 14:08:54 o/ 14:09:04 * ricolin expect Line team will shows up in Large scale SIG:) 14:09:07 Well for now it is a bit convenient not to have anyone from teh US from a meeting org perspective, so I'm not exactly chasing more down 14:09:15 I see 14:09:22 LINE, YAhoo!Japan, CERN, China Mobile... 14:09:37 :) 14:09:43 StackHPC 14:09:50 OVH 14:09:56 yeah definitely europe/apac 14:10:08 report next month to see how this flies? 14:10:22 I like that those were not necessarily very invested upstream before 14:10:25 So we'll see 14:10:29 sure 14:10:32 that's good news indeed :) 14:10:59 I will keep the topic for next month if you don't mind 14:11:06 wfm 14:11:09 that looks important to keep an heartbeat on 14:11:13 ok next 14:11:15 is gmann there? 14:11:29 or mnaser? 14:12:08 I will skip the two topics of them for now, moving to ricolin again then :) 14:12:11 #topic ricolin report on multi-arch sig and other sigs 14:13:20 For multi-arch we got a great number of people shows their interest on, which means there's definitely value to form it asap 14:13:31 nice to hear 14:13:53 would that be joined by the power folks, combined by the people you know focusing on ARM? 14:14:05 rigth now we have Linaro to provide around 10 server, which might get online early next year 14:14:17 Nice. 14:14:25 nice indeed 14:14:31 still looking for more:) 14:14:46 evrardjp, that's good idea, I should ask them 14:14:48 did you ping tonyb for his opinion on this? He might be able to provide you contacts 14:15:11 evrardjp, yes, he also replied the ML too 14:15:14 #link http://lists.openstack.org/pipermail/openstack-discuss/2019-November/010970.html 14:15:20 awesome! 14:15:24 he's onboard! 14:15:25 I missed this, sorry 14:15:28 amazing 14:15:29 NP 14:15:33 oh yeah I remember now 14:15:48 ok cool 14:16:03 is there anything else that we should be aware or that we can help, regarding SIGs? 14:16:13 (not necesarily the multi-arch one) 14:16:14 I will take action to contact to Infra team to see if we got more ARM server in our Nodepool 14:16:30 We got two thing going on in SIGs 14:16:40 that might be good to discuss with mnaser about that too, as he is already very involved in infra 14:17:01 eh, timezones are confusing. hi, i'm here. 14:17:02 (he is kinda our liaison to deal with the static hosting story that I wanted to chat here) 14:17:03 One is we're tagging SIGs to match their current state https://review.opendev.org/#/c/695625/ 14:17:09 #link https://review.opendev.org/#/c/695625/ 14:17:31 another is the combine of self-healing SIG and auto-scaling SIG discussion 14:17:46 #link http://lists.openstack.org/pipermail/openstack-discuss/2019-November/010989.html 14:17:55 turns out most of static hosting was actually done by andreas, i'm working on figuring out the next steps. 14:18:14 mnaser: yeah we'll come to that soon, we're finishing the report on what's going on in sigs 14:18:50 (if there is more to say) 14:18:57 nope 14:19:08 that's all 14:19:12 haha ok 14:19:25 thanks for the work there ricolin 14:19:30 also ttx been promote to meta-sig core 14:19:35 oh great! 14:20:10 do we need to track the merger of those sigs (self healing and auto scaling) ? 14:20:23 I suppose you could do a little summary at next meeting? 14:20:34 (you = ricolin or ttx) 14:20:48 I will do it since I'm driving that atcion too 14:20:54 ok perfect 14:21:03 nice to see progress there! 14:21:30 ok let's go to one of the many topics for mnaser then 14:21:37 #topic mnaser report on the infra liaison 14:21:43 what's happening with the static hosting? 14:22:02 is there more to say than what you said above? 14:22:06 there's not much to report, turns out andreas has done a lot of the work and we were blocked for a while with the afs volume being pending creation and some infra patches pending to merge 14:22:16 so there's not much progress but we've also been blocked for waiting for a bit too 14:23:05 can anyone around here help on that, or is there tribal knowledge that will be prevent people from this team to step up? 14:23:14 or any other blocker 14:23:24 I am just curious on how to help this forward 14:23:45 it was blocked on the infra team because of afs things and job creation (i.e. openstack/project-config patches) 14:23:55 they didn't have the infrastructure ready for us (yet) 14:24:00 oh I see 14:24:33 I will keep this in the topics for next meeting, so we track if it's still blocked on infra side, ok for you? 14:24:41 sure 14:25:03 in the meantime, I suppose you're pinging on this to track the status update? 14:25:26 i'm sorry, i don't follow 14:25:34 (just to know that both teams aren't waiting for each other at some point) 14:25:48 no, i'm not pinging the infra team, afaik they just finished their work a day or two ago 14:25:51 "oh but we did that, you can go ahead now" 14:26:05 I see 14:26:19 so this work can be unblocked soon then? 14:26:30 hopefully 14:26:31 (sorry if I ask stupid questions, it's because I am not sure to have full context) 14:26:32 ok 14:26:33 cool 14:26:38 let's talk about that next meeting 14:26:43 it's being taken care of :) 14:26:55 yeah, sorry for checking! :p 14:27:11 ok next 14:27:21 #topic mnaser report on sync with swift team 14:27:25 I had this topic pending 14:27:29 i think that should have been removed a while back 14:27:34 it was on py3, I think we can remove 14:27:38 ok 14:27:41 sounds good 14:27:59 next 14:28:00 #topic ttx report on Technical vision reflection update 14:28:15 #link https://media.giphy.com/media/Y4tXwMTNHStfxDxduD/giphy.gif 14:28:22 I have not started that yet. I hope to get it going early January 14:28:34 ok 14:28:37 so that I can report about it at next meeting 14:28:51 I might start it end of December tho 14:29:11 that would be good, if you want to discuss about it in January 14:29:16 holidays and stuff :) 14:29:32 ok 14:29:59 next 14:30:02 #topic mnaser summary of the maintain issue with Telemetry 14:30:26 it seems like catalyst picked up ceilometer work and ran away with it 14:30:28 I still have that in the topics, it was discussed in the PTG, continued over ML 14:30:30 like, literally, ran away with it 14:30:41 other projects suggested working together to leverage their backends 14:30:55 and the current roadmap for ceilometer is "add the api again, add mongodb support again, go back in time" 14:31:18 :/ 14:31:19 I remember a tweet pointing to that ML thread indede 14:31:24 yes. 14:31:35 https://storyboard.openstack.org/#!/story/2006893 14:31:51 https://storyboard.openstack.org/#!/story/2006894v 14:32:06 mnaser: did we start a conversation with them, to engage into not running back in time? :p 14:32:24 i tried to politely steer the convo in the ML 14:32:39 #link https://storyboard.openstack.org/#!/story/2006894 14:32:42 #link https://storyboard.openstack.org/#!/story/2006893 14:32:46 fwiw, I know we discussed putting metrics directly in prometheus at some point, I did some poking at that, and the python client is... not awesome. gets weird with multithread/process wsgi runners, and doesn't do any sort of quantile metrics (I was surprised this is up to the client, not the server) 14:32:52 and i think i got a "this is what the ceilometer team is doing." 14:33:01 but just a data point, we'd need to do some work there first 14:33:08 and then there was a few other proposals from other teams like monasca to integrate 14:33:34 jroll: this was the next topic as a smooth transition :) 14:33:39 that were ignored, i asked for an update on what was the final decision, no answer, yoctozepto did as well (2 days ago) -- nothing. http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011337.html 14:33:44 mnaser: :/ 14:33:50 aha, I shall wait 14:34:11 http://lists.openstack.org/pipermail/openstack-discuss/2019-November/011236.html 14:34:14 asked on nov 28 14:35:05 I see it didn't really result in new stance 14:35:11 http://lists.openstack.org/pipermail/openstack-discuss/2019-November/010972.html 14:35:35 that is where it starts to feel concerning tbh 14:35:51 "we're going to maintain and support our use case" 14:36:13 I wonder if they're unwilling to support monasca/etc, or just saying it works well 14:36:16 yes i get it, this is open source, no one is entitled, but we also ask people to do whats best for openstack, not whats best to maintain your legacy piece of code 14:36:23 isn't that fair though? Are they refusing code from others? 14:36:27 wonder if lxkong is around 14:36:32 its probably late there 14:36:34 yeah 14:36:42 3.30am 14:36:52 jroll: agreed, that's basically my question 14:36:59 anyways, i don't want to sit and send emails to the entire ML every other day if there's no reasonable outcome that that the tc has any influence obn 14:37:04 I don't think that means they will refusing others 14:37:19 if you're going to invest all that time 14:37:20 on one hand, it's fair to say that if nobody else is contributing, they'll contribute what they want to contribute 14:37:23 mnaser: not being an official project is the only thing we can do, but this is not something I want to arrive to right now 14:37:33 and that's only if it's not open to contributions 14:37:37 might as well as invest it in the right way 14:37:39 zaneb, true 14:37:49 zaneb: yeah personally im torn on that 14:37:58 on the other hand, who is going to show up and contribute when they're making it clear that they're only interested in maintaining their legacy stuff from 5 years ago? 14:38:01 "do the right thing if you're gonna do the work anyways" vs "anything better than nothing" 14:38:11 zaneb: yes indeed 14:38:58 also if it works for them, could that work for others? just curious 14:39:18 I think we should discuss with them whether they could continue to meet their goals as an unofficial project 14:39:29 zaneb: it will not be the only project to be single vendor in openstack :p 14:39:41 adding the api back is a serious change in direction 14:40:02 because I don't see that they're getting great benefit from it being in OpenStack, and for other people it's only creating confusion 14:40:23 evrardjp: we could also force a PTL change, but that's also nuclear and it's not as if we had likely candidates 14:40:32 I agree, but as TC we probably shouldn't intervene on the project, but helping on the ecosystem, and I haven't seen evidence it's not helping the ecosystem as a whole right now 14:40:47 if they're ok with it leaving OpenStack officially, then our job is relatively easy 14:40:56 Yeah. 14:41:05 if not then we have some hard discussions to have 14:41:13 Is there a precedent for forcing at PTL change? 14:41:17 I don't want to go there though -- I know it's a radical change for users, but is it for a bad thing for the ecosystem? 14:41:25 I been promised by telemetry PTL of a status report, we should ask some questions to them to clarify the current status 14:41:35 lets not go down the forced PTL change route discussion 14:41:44 yes, let's not 14:41:45 mnaser: ++ 14:41:47 o/, sorry i missed the time change thing. 14:41:50 ricolin: good idea, and if it still benefits everyone in the community would be a nice question 14:41:52 mnaser++ 14:41:56 Ok, wanted to make sure that others felt that way. 14:42:06 mnaser: agreed 14:42:21 we're discussing the issue much more than trying to gather an update 14:42:28 does anyone wanna come up with some next steps 14:42:34 ricolin: just did 14:42:51 well 14:42:55 ok, so ricolin has an action item of getting a status update, id hope before the next meeting cause that would mean a month of work (in maybe) not the ideal direction. 14:42:55 between the lines :) 14:43:17 ok you want a status update before next meeting 14:43:27 that sounds good, can we discuss this at next week's office hours? 14:43:29 i dont want one, i think it would be beneficial, because a month is a long time 14:43:34 that's fair 14:44:14 Wondering if there's any other resources/info. we need from them? 14:44:43 besides status update 14:44:50 ricolin: I guess what would matter for mnaser would be having a good chat of what's going on, why the decision -- while people seem so eager to not do that 14:45:05 wow my english is terrible 14:46:04 I think users need answers on the direction taken, to know how they can contribute or not. And if it's not the direction that some people want to take, ensuring that the voices are heard 14:46:10 evrardjp, it's better than your chinese:) 14:46:12 given my constant complaints about magnum (which ended up cleaning up a lot of it's stuff), i did really end up feeling pretty awful in the whole thing because it pushes people 14:46:35 and the employer of this PTL is the same as the magnum one at the time, so id rather sit and talk from the sidelines 14:46:45 ok 14:46:58 I wonder if someone should not raise the situation again (not mnaser) to show that the concern is a TC concern 14:47:16 i assume that s/not/now/ 14:47:36 now raise, not mnaser yes 14:48:07 I will take that action tomorrow and asking for the status report too from PTL:) 14:48:09 ricolin: I can do that with you, as my chinese is awful :p 14:48:18 let's sync on that 14:48:25 yes, definitely 14:48:27 it can only help if the person is close to teh APAC tz, so +1 14:48:39 ttx evrardjp sweet! 14:48:55 #action ricolin evrardjp talk with PTLs about the direction of telemetry (see meeting log for detailed AP) 14:49:08 #topic mnaser report on oslo metrics project 14:49:18 jroll: you also wanted to say a few words about this? 14:49:41 I'll let mnaser speak first, I just have some recent experience with prometheus here 14:49:52 i didn't progress on this, i think bnemec is not as involved in openstack-y things and he spoke with the LINE team, maybe ttx can help with the intro/sharing of the code :< 14:50:18 was LINE using prom, IIRC? 14:50:24 yeah 14:50:28 i think? 14:50:31 cool, I'd love to see the code 14:50:33 Yeah, we talked in Shanghai. We're waiting on them to propose the new library and make the code available. 14:50:41 ok 14:50:48 * ttx checks notes 14:50:48 as I mentioned before, the official python client for prom has some problems that make it not ideal for usage in openstack 14:50:56 but they seem relatively solvable 14:50:59 they got a presentation for their structure, back in summit 14:51:16 jroll: can you push from your side to see the code, so I can ask you in a month how is this going? 14:51:19 that would be amazing 14:51:31 bnemec, is there any update from them now? 14:51:32 yes, os.metrics instrumenting oslo.messaging to send to Prometheus 14:51:43 * ricolin knows it still early to ask 14:52:05 evrardjp: I have no contacts there, all I could do is ask on the ML, not sure if that's helpful 14:52:13 argh 14:52:22 I don't think they're using prometheus from the OpenStack side, they're just exporting metrics in a way it can consume. 14:52:56 jroll: maybe bnemec has :p 14:53:11 and he seems just here 14:53:22 :D 14:53:23 They did a presentation during the forum. Can probably contact those folks. 14:53:29 awesome! 14:53:49 I think if you can contact them, and introduce to jroll that would be the first step 14:54:10 jroll: you can reach out to dinesh (he used to be on irc, i can check his irc name) 14:54:15 I don't want intros, only code 14:54:16 heh 14:54:31 fair 14:54:35 I am happy to reply in an email chain about how excited I am to see the code, if that will motivate them 14:54:37 but those ppl can give you links 14:54:54 ok let's settle for that then 14:55:28 #action jroll to use his email client to see code 14:55:33 ahem ahem 14:55:36 -.- 14:55:53 oh come on, that was hilarious. 14:56:01 "." 14:56:04 tough crowd 14:56:15 ok we have so many things 14:56:18 let's continue 14:56:29 #topic gmann report on community goals for U/V/py2 drop/rolling to py3/goal select process schedule 14:56:30 good thing we have an office hour to go deeper 14:56:39 *sad_trombone.wav* 14:56:46 jungleboyj: :) 14:56:57 I will give you gifs 14:57:00 #link https://media.giphy.com/media/3oEhn3lV6Yk1Ku6WRi/giphy.gif 14:57:01 Ussuri goal update: 1. py2 drop - this is going on, many services has merged the patches and other are in progress. This is in good progress. 2. PTl and Contributor guide- goal is not yet merged - https://review.opendev.org/#/c/691737/ 14:57:20 oh wait, a better python drop 14:57:22 #link https://media.giphy.com/media/zzkFADpR7k2NG/giphy.gif 14:57:30 I have only one concern left there is to have template for PTL guide so that projects can maintain the conssitency 14:57:37 consistency 14:57:44 diablo_rojo: ^ 14:58:18 gmann: ++ 14:58:24 gmann, I think that there were concerns that having as its own separate document might be too time consuming for a small amount of people 14:58:41 because not all read the expectation till we have the defined template . for example projects does not read the py drop schedule and start dropping the py2 from python client lib also and i have to -1 them every time 14:58:44 I also think that there may not be that much different than needs to be mentioned? 14:59:04 Also that its fine living in CONTRIBUTING.rst because being a PTL is form of contribution 14:59:18 gmann: arrgh 14:59:33 diablo_rojo: i think common PTL guide we have currently does not reflect all PTL works. 14:59:38 (I think I also wrote all/most of this in the review of the template yesterday) 14:59:58 gmann, no? Well that should probably get updated with all the common work that PTLs do them 15:00:10 Only the project specific stuff should live in the template 15:00:14 ok let's talk one topic at a time, now talking about the U - PTL goal part 15:00:17 ok, we can discuss that in office hour. 15:00:26 we are in the office hour basically :) 15:00:27 Lol sorry 15:00:30 :) 15:01:04 V goal- zuulv3 goal is under review which we decided to have as pre-selected as V cycle 15:01:24 ok that's good 15:01:25 goal schedule - I am going to propose the schedule up by today. 15:01:31 thanks 15:01:38 that's all from my side 15:01:44 for all those points (before we circle back to the problems), do you need any help? 15:02:05 we are good on these. 15:02:09 ok 15:02:26 we will discuss more on PTL goal next in office hour 15:02:42 you mean next week? 15:02:53 or is my agenda broken? 15:03:05 today 15:03:08 or after the meeting 15:03:09 ok 15:03:11 got it 15:03:14 yeah after meeting 15:03:22 sounds good, I need to accelerate then 15:03:47 #action gmann discuss about the issues in the U goals at the next office hours 15:03:50 #topic jungleboyj update on the blog post about the analysis of the Foundation user survey 15:04:09 :-) I have this in my sprint plan for the week. 15:04:13 jungleboyj: are you there? Do you have an eta? :p 15:04:15 woot 15:04:17 awesome 15:04:26 So, should have something for this next week. 15:04:31 I am writing that down for next meeting 15:04:38 evrardjp: Sounds good. 15:04:46 mugsie: are you there? 15:04:53 it's for the next topic which also takes time 15:04:55 #topic mugsie update on release naming progress 15:05:30 I want to put this to bed 15:05:35 it's been far too long 15:05:36 ++ 15:05:49 evrardjp: ++ 15:05:57 I thought we reached consensus during PTG, but I was wrong 15:06:01 What's the standing objection at this point ? Not enough votes? 15:06:24 gmann: thinks it's not respecting opinions of the people, if I am not mistaken 15:06:52 which leads to me to think it's possible to invert the things: tc proposes names and community votes 15:07:09 instead of the opposite: community propose names and tc votes 15:07:18 though it will still make the life of the voting complex 15:07:22 I liked the community proposed named better. 15:07:28 *names 15:07:42 smcginnis: Agreed. 15:07:44 smcginnis: +9million 15:07:51 evrardjp: that would be a ton more work for the TC 15:07:53 me too, as it was also transparent for which tc member has voted, so it was still transparent for members 15:07:57 I thought that that is the important thing. 15:08:01 ok 15:08:07 so we are all in agreement again 15:08:19 I am glad. Let's vote in majority for this and make it 15:08:26 only issue I have is let community decide the final name and anything else is all fine to me(who choose the name etc) 15:08:31 speaking with my community member hat on, the fun part is looking for names 15:08:38 zaneb: yes 15:08:48 wait what are we voting in majority for? 15:08:48 zaneb: ++ 15:08:51 voting is boring and the outcome is usually terrible no matter who chooses it 15:09:04 also conducting a vote of the entire community has a lot of rough edges 15:09:07 We may get some heat for the votes, but that is part of the work of leadership. 15:09:09 zaneb: ++ 15:09:13 mnaser: community propose names and tc votes 15:09:17 Could care less about voting on it. 15:09:23 mnaser: I know you disagree :) 15:09:41 ill make my own little twitter poll 15:09:43 Community elects TC, TC picks the name. 15:09:51 and then the community can have my vote if im on tc 15:09:51 mnaser: sure, that works too 15:09:51 :P 15:09:57 And the community can lobby for their preferences. 15:10:05 that is what current proposal in review is so what is difference 15:10:20