15:00:05 #startmeeting tc 15:00:05 Meeting started Thu Aug 19 15:00:05 2021 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:05 The meeting name has been set to 'tc' 15:00:08 #topic Roll call 15:00:12 o/ 15:00:34 bonjour 15:00:35 o/ 15:00:44 o/ 15:00:54 o/ 15:00:54 o/ 15:01:27 absence: Belmiro Moreira (belmoreira) 15:01:46 o/ 15:01:49 #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda_Suggestions 15:01:52 ^^ Today agenda 15:01:58 let's start 15:02:08 #topic Follow up on past action items 15:02:13 three action item 15:02:25 gmann to drop skyline pre-check topic from agenda - done 15:02:40 ykarel to add centos8 vs centos8-stream testing for old stable in PTG etherpad - done 15:02:59 mnaser to send the murano retirement proposal n openstack-discuss ML - this is also sent 15:03:09 yup :) 15:03:10 and we will talk about it in next topics 15:03:19 thanks ykarel mnaser 15:03:30 Gate health check (dansmith/yoctozepto) 15:03:34 dansmith: yoctozepto any updates 15:03:38 #topic Gate health check (dansmith/yoctozepto 15:03:49 so I haven't really been experiencing any systemic issues lately, 15:04:03 glance is struggling with something that seems specific to glance's tests, 15:04:06 so we've been doing a bunch of rechecks 15:04:15 and so things seem pretty stable overall 15:04:21 and for py36 especially 15:04:45 there was one thing i found interesting that i can bring up if that's ok 15:04:54 I am wondering if we should continue on py36 testing or move to py37-py38 in next release? 15:05:09 gmann: I dunno, but it's not like this is actually py36, 15:05:16 k 15:05:23 what's the default python3 in centos-8 stream at the moment? 15:05:32 I think it's just because it's 18.04, different mysql versions, etc and that's making it happen 15:05:46 yeah py36 jobs run on 18.04 15:05:49 it's likely a legit race and that's just where it's manifesting right now 15:05:52 mnaser: bring it on :D 15:06:04 http://lists.openstack.org/pipermail/openstack-discuss/2021-August/024236.html 15:06:09 if centos is still going to be python 3.6 then dropping 3.6 support next cycle could be problematic 15:06:26 this seems like the sort of thing that would be good to help drive i think 15:06:34 centos 8 stream is python3.6 15:06:58 will it ever move to newer python minor releases or does that require centos 9? 15:07:16 mnaser: thanks, I'm coordinating them already 15:07:35 mnaser: agree. in our gate tempest tests use cirros image (default) in most of the job so clarkb suggestion make sense to try in that 15:07:41 fungi you'd have too moove to a newer release 15:07:43 yoctozepto: +1 15:07:54 fungi, i don't think it will move to new python3 minor, need to go with centos 9 stream , which i think includes python3.9 15:08:00 but anyway, claiming we target centos/rhel means we probably need to continue not breaking python 3.6 15:08:04 yoctozepto: great, i just wanted to put some eyes at as well 15:08:09 fungi: +1 15:08:29 and also agreed on needing to continue to support 3.6 15:08:32 agreed we should not drop py36, and especially not just because of this glance test race thing 15:08:35 or else tripleo world would break i guess 15:08:45 because it has nothing to do with py36 most likely 15:08:46 yoctozepto: mnaser and i think some tempest test do modify it at runtime for image meta or so. 15:08:57 there is however some potential struggle coming, since pip and friends want to drop python 3.6 support when it reaches eol in a few months (and are already breaking 3.6 compatibility and not concerned with fixing the regressions there) 15:09:02 we can't reproduce it on 18.04 outside of the gate either 15:09:10 ok 15:09:20 yeah, py36 is going eol in the python community soon 15:09:26 let's continue on py3.6 15:09:27 yoctozepto: when? 15:09:48 4 months apparently 15:09:50 23 dec 2021 15:10:00 if py36 goes eol and that actually prevents us from testing, then that's fine, and it's on RH to be supporting such an old thing anyway, but no reason to drop it until it's broken, IMHO 15:10:03 "3.6 will receive bugfix updates approximately every 3 months for about 24 months. Sometime after the release of 3.7.0 final, a final 3.6 bugfix update will be released. After that, it is expected that security updates (source only) will be released as needed until 5 years after the release of 3.6 final, so until approximately 2021-12." 15:10:19 src: https://www.python.org/dev/peps/pep-0494/#lifespan 15:10:21 yeah 23 Dec 2021 15:10:40 there will come a point when we want to continue working on python 3.6 but our dependencies and toolchain utilities are dropping support for it, which will probably force us to carry a lot of pins or support multiple versions of some deps 15:11:03 the pypi annotations for supported python versions help a lot with that 15:11:15 if you get apckages to properly set those values it simplifies things greatyl 15:11:16 we already have some 15:11:17 i think the natural involvement of rh/tripleo will help navigate those 3.6 problems given they have to continue to support it 15:11:36 yes, that may make it not necessary to pin things, since pip will try to avoid downloading versions of things which say they don't support your version of python 15:12:00 we already pin a lot with u-c 15:12:05 it's just that we will need to pin pip 15:12:17 pin the pinmaster 8-) 15:12:22 and possibly tox and so on 15:12:32 I think we should discuss it in PTG as dec is not so far 15:12:33 yeah, possibly virtualenv, setuptools too 15:12:41 but it's fine, we know the drill 15:13:13 though for Yoga it trulye makes little sense to test py36 15:13:27 why? 15:13:32 centos 8 stream has 3.6 15:13:32 gmann, agree, better comes up with plan for community guide 15:13:35 yeah not getting that 15:13:58 mnaser: testing on something going eol in 2 months? ;d 15:14:02 if we have centos8-stream keep supporting it then no need to drop 15:14:12 but I agree that we keep it as distro has it 15:14:25 anyways we have lot of topic today let's discuss this in detail in PTG? 15:14:28 yeah if we're going to keep centos 8 stream in listed of supported/tested distros, we need to test py36 15:14:33 but i agree with gmann wrt ptg discussion 15:14:35 #action gmann to add py3.6 testing plan (after its EOL -Dec 2021) in PTG etherpad 15:14:45 perhaps we need to reword that page 15:14:48 #topic Murano project health (gmann) 15:15:02 #link http://lists.openstack.org/pipermail/openstack-discuss/2021-August/024120.html 15:15:03 #link http://lists.openstack.org/pipermail/openstack-discuss/2021-August/024120.html 15:15:08 :) 15:15:09 two chairs 15:15:12 :) 15:15:16 old habits die hard 15:15:20 +1 15:15:36 so murano PTL replied to maintain it, continue maintain it 15:15:58 last reply #link http://lists.openstack.org/pipermail/openstack-discuss/2021-August/024240.html 15:16:14 yeah, it seems he strongly wants to keep it alive 15:16:21 I feel we should let them to decide on retirement and if they want to continue on maintaining it it should be ok 15:16:39 i think the reason we wanted to do a 'tc-initiated retirement' is because we didnt get any responses 15:16:45 community wide goal patches are merged for murano 15:16:57 mnaser: yeah for many months 15:17:15 it is a very edge case 15:17:31 but this is good trigger to convey the message to keep up the basic maintenance at least 15:17:36 i think there are some good points that were brought up 15:17:40 which is quality of the software we ship 15:17:48 mnaser: ++ 15:18:15 but that diverges the topic 15:18:29 not drastically I guess? 15:18:36 i worry that if they don't respond until the tc starts a thread about retiring the project, that's not a particularly compelling reason to expect it to be actively maintained 15:18:47 fungi: ++ 15:18:53 which leaves two: 1) do we continue with the retirement of murano and 2) do we need to come up with a discussion of a 'standard' for openstack shipped software 15:19:00 my point is that we might want to explain to the PTL 15:19:07 that they can still develop this software 15:19:11 using opendev tooling 15:19:23 but outside openstack itself 15:19:24 yeah, there might be lot more project in that category 15:19:34 but per tc policy they will have to fork it to another namespace once it's retired 15:19:37 because perhaps they are worried 15:19:45 that it's just going to vanish with a click of a button 15:19:46 mnaser: +1 on 2nd point and take it from there and drop the murano retirement now 15:19:50 like it's happened to others 15:20:01 fungi: yeah 15:20:08 but we can link there 15:20:09 because imho if we're going with the "this thing doesnt make sense in the current landscape" as retirement reasons 15:20:25 this would mean a lot of other openstack projects might see a similar story 15:20:33 and that is not in our policy of now 15:20:38 right 15:20:46 mnaser: although they exhibit better observable behaviour 15:21:16 but maybe that's just because they have a bit more contributors 15:21:29 yup, more general interest 15:21:58 i agree that 'our current policy is not to drop projects that dont fit the landscape or seem inactive' and we've gotten a response from murano 15:22:07 so maybe drop the retirement for now but open this for a ptg discussion 15:22:17 yeah make sense 15:22:19 I agree with mnaser 15:22:32 mnaser: would you like to add it in PTG etherpad ? 15:22:40 sure 15:23:06 I will respond to ML on we are fine on not to retire murano 15:23:11 thanks 15:23:37 and i added to ptg list 15:23:44 cool thanks 15:24:02 anything else on this topic ? 15:24:23 #topic New project application: 'Venus' 15:24:30 there is new project application 15:24:31 #link http://lists.openstack.org/pipermail/openstack-discuss/2021-January/019748.html 15:24:38 #link https://review.opendev.org/c/openstack/governance/+/804824 15:25:09 ML is on Jan month but I think they put all the info and answer the yoctozepto query on gerrit too 15:25:27 please review or ask query if you have any 15:25:33 yeah, they are responsive 15:26:00 Two thing to note is: 1. current repo are not in openstack namespace and can be added later or before our agreement on gerrit 15:26:18 2. if we add this projetc now, there will not be any release in Xena 15:26:25 both looks ok to me 15:26:36 it still feels very vague to me, i don't quite understand fully what it is exactly 15:26:44 yeah, renaming the repos into the openstack namespace will need a gerrit outage scheduled, like any rename maintenance 15:26:54 it seems like some sort of mix of monasca... 15:27:00 mnaser: query them 15:27:07 i recommend not planning for renaming until after the tc has approved inclusion of the project 15:27:13 the more we know, the better 15:27:18 fungi ++ 15:27:18 yeah, ML or gerrit will be good to add query 15:27:28 I suggest focusing on gerrit 15:27:38 And if it won't make Xena maybe noot rename until after rellease? 15:27:46 fungi: yeah we are getting TC agreement first and then we will ask them to propose renaming 15:28:32 yeah i think the rename is the easy bit :) 15:28:50 yeah, let's focus on the why 15:28:58 yeah 15:28:58 right, more suggesting that you don't consider the fact that the repos are in a different namespace as a blocker to the governance change 15:29:06 ++ 15:29:16 I *think* I understand their perspective 15:29:17 so please put your query on ML or gerrit. 15:29:22 it should be governance approval first, then rename maintenance, then update governance for the new repo names 15:29:23 to have log analysis tailored to OpenStack 15:29:29 without the complexity of Monasca 15:29:40 and we will continue this topic in case anything we need to decide on that or any strong objection 15:29:46 yoctozepto, yeah, I guess something like that 15:30:04 I am looking forward to seeing this working with kolla 15:30:10 but we are working on that with inspur now 15:30:19 +1 15:30:21 or more so: they are working on answering my queries 15:30:51 yoctozepto: did you decide on non-openstack repo to include in kolla as you said you will discuss in kolla meeting 15:30:59 or just trying to see if it work 15:31:35 gmann: yeah, we decided we can add it even if it was unofficial as long as it actually does something useful 15:31:48 ok. +1 15:32:02 it looks healthy as it's brand new and inspur poured a lot of work into it 15:32:14 but we'll see of course 15:32:21 yeah 15:32:33 I understand it can get popularity in China due to mothertongue marketing :-) 15:33:08 let's see 15:33:09 anything else on this topic? 15:33:32 not from me 15:33:35 #topic PTG Planning 15:33:36 * yoctozepto initially happy 15:33:39 #link https://etherpad.opendev.org/p/tc-yoga-ptg 15:33:48 gmann going fast 15:33:55 :) 15:33:57 like a tempest :D 15:34:05 we are copiously adding topics there which is great 15:34:20 please keep adding if you find something to discuss there 15:34:29 nothing else to discuss on this 15:34:42 #topic Moving weekly meetings to video/voice calls. 15:35:02 there is cinder feedback on this 15:35:11 i'm a big supporter of this in terms of increased engagement of meetings 15:35:33 i think the longer an irc meeting goes on, the more.. things slow down 15:35:35 this is for the tc meeting presumably? 15:35:49 ^ yes 15:35:52 #link http://lists.openstack.org/pipermail/openstack-discuss/2021-August/024260.html 15:36:07 dansmith: yes only TC meeting 15:36:08 I think it makes sense to do it monthly 15:36:12 oh I see, cinder feedback on video meetings in general 15:36:22 you know, even just to see each other 15:36:24 yeah I'm cool with some video meetings 15:36:27 logging is bit issue as brian mentioned 15:36:36 I am also in favor of that 15:36:52 we can mitigate logging slightly but picking topics wisely 15:36:56 by* 15:37:00 by* picking 15:37:07 and try to summarize more in the summary email 15:37:18 currently I just add logs link 15:37:18 yeah, i just think it's a good way to increase our overall meeting engagement 15:37:21 yeah, your mails, gmann, are being praised ;-) 15:37:23 (i sound very business-y saying that, ha) 15:37:36 (mnaser: you do indeed!) 15:37:46 (but we need this layer too) 15:37:53 we will get better sYnErGy 15:38:02 yoctozepto: happy to see someone reading that and it is helpful :) 15:38:08 mnaser: omg, I'm melting 15:38:08 mnaser: +1 15:38:10 no, but i really do think having a video meeting will be more engaging overall 15:38:17 ++ 15:38:24 so seems no objection on video call 15:38:25 We can do video meeting +etherpad (PTG like) 15:38:28 engaging tc members, or engaging non-tc-members? 15:38:39 so question is should we do every week or monthly first ? 15:38:42 As long as we have better notes then we get from the video meetings we do for RDO:) 15:38:43 and see how it goes 15:38:47 fungi: i think perhaps both 15:38:52 gmann: monthly 15:38:55 We do ours first meeting of the month 15:38:59 folks might join in and listen 15:39:22 mnaser, fungi: yeah, we need to at least let them join as well 15:39:33 yeah 15:39:48 as mentioned last week, i will certainly be less engaged, because i'm double and sometimes triple-booked, so while i can opportunistically participate in irc meetings, i will skip the videocalls. i may be in a minority however 15:40:11 I guess we could optimise the schedule? 15:40:32 I have a CentOS Cloud SiG conflict first week of the month 15:40:33 i may reply in irc to things people log here as meeting notes, so maybe that's "good enough" 15:40:34 in the worst case, you will be losing one meeting a month 15:40:42 let's try it on first meeting of every month. which will be 2nd sept meeting on video. and if it goes well then we can make weekly video call too 15:40:42 fungi: ++ 15:40:44 But that's IRC 15:40:59 gmann, +1 15:41:14 i have a few suggestions 15:41:43 sure 15:41:52 google meet has excellent transcription, and it works in a way of recognizing who is speaking, so it is _SUPER_ accessible for anyone who's not a native speaker and rather read if they don't hear something well 15:41:57 so it'll actually be like 15:42:09 Rico: ....., Jeremy: ..., etc 15:42:19 and there are chrome extensions which save those transcripts to publish them 15:42:49 NEC has some in-progress scripting tool to convert video call to text, at some time I can try to see if we can use that 15:42:53 so we can have a _full_ transcript of the discussion, and it helps with accessiblity for those who might not be able to understand some things 15:43:09 also remember that if you're going to hold the call on a google service, you will be excluding participants from mainland china (or expecting them to break national laws in order to participate) 15:43:14 i don't know whether that's a concern 15:43:26 yeah, that is my main worry about google meet 15:43:30 fungi: ah right, i wonder if dial-in works for them.. 15:43:35 i've not found zoom to be good when it comes to transcription 15:44:16 ok well 15:44:23 we've got another meeting in place to be able to nail down the logistics 15:44:55 yeah and first can try on zoom or so and with the some feedback we can also discuss the best tool in PTG 15:45:03 TBH it's 11:00 pm for China, so you shouldn't expect to have much access from China anyway 15:45:04 but having few meeting before PTG will be good to know pros and cons 15:45:30 ricolin: and you can access google meet right? 15:45:36 yes 15:45:38 cool 15:46:07 i'm happy to host the link/etc because we pay for it 15:46:16 so no issues with time limits / etc wrt gmeet 15:46:27 so we agreed on monthly video call and in next meeting we can finalize the channel 15:46:32 +1 15:46:41 +1 15:46:43 +1 15:47:00 anything else on this topic ? 15:47:03 gmann: do we want to make that more official in a governance change to get the proper tc okay? 15:47:15 i dont want to go through too much back and forth in a governance change 15:47:25 but just so we have it put down, unless we want to make this informal 15:47:25 mnaser: humm, I was thinking to do it after PTG with few try feedback 15:47:31 sounsd good to me 15:47:39 yeah, let's start informal 15:48:03 and as per byLaw we still have more meeting now as weekly so monthly on video should be ok in any case 15:48:12 indeed 15:48:24 #topic Board informal Brainstorming sessions about "community health and resource management" 15:48:32 egg-cellent 15:48:44 :) 15:48:57 Allison sent invite for brainstorming session 15:48:59 #link http://lists.openstack.org/pipermail/foundation/2021-August/002998.html 15:49:20 its on Aug 24th, 2021 14:00 UTC 15:49:50 i should be there 15:49:56 thanks 15:50:09 all the details are in ML including the etehrpad link 15:50:12 etherpad 15:50:26 #link https://etherpad.opendev.org/p/board-scratchpad-2021-08-24 15:50:29 my ac is broken down and they're scheduled 'between 8 and 4pm' (love that) so hopefully they don't show up during it 15:50:44 hehe 15:51:03 :) 15:51:06 #topic TC Election 15:51:21 #link https://governance.openstack.org/election/ 15:51:42 as you know, we have tc election for 4 seats 15:51:47 nomination is going on 15:51:56 0 candidates so far 15:52:00 worrying 15:52:12 I've been waiting for someone else to do theirs so I can copyit 15:52:13 please check/encourage people to run for it or add your nomination if you are thinking to re-run 15:52:30 I think I'm up, aren't you also gmann ? 15:52:39 I will re-run, might add nomination tomorrow or weekend 15:52:44 dansmith: yeah 15:52:55 yep alongside diablo_rojo_phone and jungleboyj 15:52:59 I can send message out for most Asia user groups tomorrow and see how that goes:) 15:53:13 I received a message from someoonoe interrested in maybe running and had suggested they attend a meeting to see what we do but their nick isn't in the channel list:( 15:53:24 :-( 15:53:31 ricolin: that will be great 15:53:58 spotz: +1, please guide them if they need any help 15:53:59 I'll do mine today 15:54:07 dansmith: thanks 15:54:53 and PTL nomination is also going on, so other action item for us until election official want to bring to us 15:55:15 #topic Open Reviews 15:55:17 #link https://review.opendev.org/q/projects:openstack/governance+is:open 15:55:24 gmann Always. Bad part is nominations will end before we meet again 15:55:39 mostly we talked about open review but if you have not voted/review yet please do 15:56:04 spotz: yeah, but feel free to bring in between also. we do not need to wait for meeting 15:56:15 we have one topic from ricolin to cover 15:56:36 time is less but we can continue that for couple of min if needed 15:56:37 but not enough time 15:56:40 true we're just more talking during meetings so a better idea of what we do for someone who doesn't knoow 15:56:42 ricolin: go ahead 15:56:48 #link https://etherpad.opendev.org/p/pain-point-elimination 15:57:31 tc-members, so we have a lot of projects answering pain point collect action, we kind of need some brain storming about how should we keep that forward 15:57:48 dang, that document is wild 15:57:50 and how we can make community goal out of it 15:57:55 mnaser, exactly 15:58:46 even we didn't make a goal, still is something we can try to targeting on 15:58:54 ricolin: +1 on brainstorming. should we have few of them before PTG. some informal video call sessions ? 15:59:13 gmann, yeah, I do like that a lot 15:59:22 and in PTG we can check if anything common and we can get some goal wide topic/things to improve 16:00:13 I guess it's also up to us to figure out how to help teams to get a goal out out that etherpad 16:00:45 I'd be curioous thoughts on the Client as we were moving towards fully functional and then some projects went back to using their own 16:00:51 yeah, I feel goals depends on the common issue every or most of the project facing 16:01:26 i thought the suggestion which originated this exercise was that projects would focus on their individual user pain points rather than expecting the tc to provide a common goal across projects 16:02:07 * yoctozepto off; thanks gmann and others 16:02:08 spotz: yeah that is one of the key issue we need to continue working . I do not have clear or current status on that 16:02:13 Dan Smith proposed openstack/election master: Add Dan Smith TC candidacy for Yoga https://review.opendev.org/c/openstack/election/+/805247 16:02:14 anyways 16:02:18 I think time is less to discuss or plan those. so let's continue it in next meeting, I will add it in agenda. 16:02:30 ricolin: is it ok? 16:02:37 yes 16:02:41 thanks gmann 16:02:42 cool 16:02:51 thanks everyone for joining 16:02:53 I will added it to next week's agenda 16:03:00 ricolin: +1 thanks 16:03:02 #endmeeting