Thursday, 2018-08-02

*** edmondsw has joined #openstack-tc00:07
*** clarkb has quit IRC00:09
*** edmondsw has quit IRC00:12
openstackgerritIan Y. Choi proposed openstack/governance master: Update PTI around docs build with additional logics
openstackgerritIan Y. Choi proposed openstack/governance master: Update PTI around project doc translation support
*** mriedem has quit IRC00:29
openstackgerritIan Y. Choi proposed openstack/governance master: Update PTI around project doc translation support
openstackgerritIan Y. Choi proposed openstack/governance master: Update PTI around project doc translation support
*** EmilienM has quit IRC01:02
*** EmilienM has joined #openstack-tc01:03
*** edmondsw has joined #openstack-tc01:29
*** zaneb has quit IRC01:43
*** edmondsw has quit IRC01:50
*** edmondsw has joined #openstack-tc01:51
*** edmondsw has quit IRC01:55
*** e0ne has joined #openstack-tc02:23
*** e0ne has quit IRC02:23
*** tellesnobrega has quit IRC02:29
*** ricolin has joined #openstack-tc02:36
*** zaneb has joined #openstack-tc03:00
*** edmondsw has joined #openstack-tc03:19
*** edmondsw has quit IRC03:24
*** tosky has joined #openstack-tc07:33
*** gouthamr has quit IRC07:36
*** jpich has joined #openstack-tc07:38
*** flaper87 has quit IRC07:46
*** tbarron has quit IRC07:46
*** eandersson has quit IRC07:46
*** flaper87 has joined #openstack-tc07:46
*** cdent has joined #openstack-tc08:49
ttxRather than dumping here, been posting my thoughts on the leaderless teams on the ML09:01
eumel8ttx: just wondering, what's the problem with the Trove PTL?09:15
ttxno authored commit yet09:25
eumel8ttx: ok, have you talked with Dariusz? From my knowledge  Samsung (PL) wants to step into Trove with a complete new team, so I would expect some more people there. If the plan is upstream developement.09:35
ttxNo, I just have second-hand reports from the Rocky PTL09:35
eumel8okay, I can contact him. We had a chat during the last OpenStack Day in Krakow09:37
ttxthat would be great, yes09:38
eumel8ttx: just found:
*** edmondsw has joined #openstack-tc09:48
*** tbarron has joined #openstack-tc10:09
*** ricolin has quit IRC10:17
*** dtantsur|afk is now known as dtantsur10:32
*** cdent has quit IRC11:59
smcginnisdhellmann: Do we have a finalized plan for the Sunday before the PTG?12:35
smcginnisdhellmann: Oh, and good morning. :)12:36
dimsgood morning y'all12:36
dhellmannsmcginnis : good morning! we have a room, and based on the polling it looks like folks wanted to limit the meeting to the afternoon. I suggested 1-5 but that may make lunch a little more challenging to manage. I'll add details about where the meeting will be to the etherpad after breakfast.12:37
smcginnisMaybe an optional, informal lunch gathering for folks that are there already, official time 1-5, then a group dinner and drinks afterwards?12:42
* TheJulia double checks when she arrives12:45
dhellmannsmcginnis : yeah, that's more or less what I was thinking12:46
dhellmannok, the etherpad has those details now
dhellmanncan someone make sure that dropbox link to the map is actually public, please?12:52
scasit is12:52
dhellmannthanks, scas12:52
dhellmannand jroll12:52
smcginnisHeh, I see you're note. Was just about to say that room looks a little snug.12:53
dhellmannyeah, it's the best they could do for us12:54
ttxWe could take it to the bar if too snug12:54
dhellmannwe'll have a bigger space on friday so we should consider that when we decide which topics to cover12:54
smcginnisWe do seem to always end up there.12:54
dhellmannttx: I think the bar was actually reserved by the other conference in the hotel that weekend? I'd have to go look back at Kendall's emails12:55
ttxah? You know more than I do :)12:55
*** mriedem has joined #openstack-tc12:57
dhellmannI'm not sure where I got that. So yeah we could see about moving to the bar if we need to.12:57
scasi'd be interested in being present at the in-person events, if i could justify it. given chef openstack's diminished contributor base, i was never able to justify to the existing contributors the effort to travel from their respective locations. now that it's down to me as the active core, i'm not certain what the conversations would be, other than continuity and despair13:01
*** cdent has joined #openstack-tc13:02
*** jaypipes has joined #openstack-tc13:02
dhellmannscas : in one of the Vancouver sessions there was some discussion about whether it would be possible to build tools to make implementing deployments with chef, puppet, ansible, etc. easier because some of the work would only have to be done once13:05
dhellmannI don't know how realistic that is, or what exactly those tools would be13:05
scasdhellmann: i've had one of those tools noodling in my head since peak openstack ate my cores the first time13:06
scaswith the dearth of time, it's been mainly thoughts and notes at this point13:06
dhellmannmaybe if you publish them, it will spark some interest13:09
scasi broke my publishing method, which i intend to get working this weekend. it's about time for another state of the kitchen13:12
openstackgerritMerged openstack/project-team-guide master: Follow the new PTI for document build
scasi thought it bad taste to send out a 'hey, queens is out' given the time of the cycle, so i've held it for the state address. i'll be in a position very soon to address getting releases properly handled13:19
*** jaosorior has quit IRC13:20
scasmaking the cookbooks generally available has felt like the vogons jumping to hyperspace as time has gone on13:22
*** tellesnobrega has joined #openstack-tc13:24
*** cdent has quit IRC13:27
openstackgerritMerged openstack/governance master: teach check-review-status to report verification failures
openstackgerritMerged openstack/governance master: update link from python3-first goal to storyboard
scasthis might also be a good time for me to get on storyboard. nobody to really onboard but myself for now13:29
*** jaosorior has joined #openstack-tc13:33
openstackgerritDoug Hellmann proposed openstack/governance master: fix markup for links
openstackgerritDoug Hellmann proposed openstack/governance master: remove placeholder file from stein goals
zanebanybody else curious whether Dariusz is the team lead or a manager for that Trove team?13:59
dhellmannhmm. interesting question.  does it matter?14:00
zanebwell, the requirement to have merged a patch is presumably because we want people who are actually doing the work to be PTLs, not their managers14:00
dhellmannI always thought it was because we wanted people to know something about the project, but I guess it amounts to the same thing.14:01
dhellmannthe whole situation with trove makes me uncomfortalbe14:01
*** cdent has joined #openstack-tc14:02
dhellmannI would almost rather move it out of governance and then let the new team take over and re-apply14:02
dhellmannI have no idea what the ramifications of that are, though14:02
dhellmannand I guess that approach doesn't exactly demonstrate a lot of trust on my part, so maybe that's not a great way to go about it14:03
scastrove is a special one. it's undergone multiple shakeups. if the new team is looking to take a different direction, that harkens to refactoring in-place14:08
dhellmannit's not really clear what they intend to do14:09
cdentnot yet, no14:09
smcginnisI'm on the fence on that. I kind of like the idea of moving them out and seeing if the new team can get back to a good enough state, but I also wonder if that would be too heavy handed.14:09
scasas a maybe-openstack-project, the intent needs to be clear14:09
scasi'm speaking about myself at this point14:09
cdentI share your concerns dhellmann. The idea of having them reapply has a lot of appeal, but also feels punishing14:09
* cdent jinxes with smcginnis. again. as usual.14:10
scasif i had others come in and want to take chef openstack a different direction, i wouldn't not be amenable, but i'd also reconsider the choices of old14:10
dhellmannat the same time I'm reluctant to simply set aside our governance rules for expediency14:10
smcginnisDifferent topic - do we have an etherpad or something to track the decisions or discussions on PTL-less decisions.14:10
dhellmannsmcginnis : i had that on my list of things to start today14:11
scasin my current state, i'm reconsidering some significant choices of old, short of rehoming chef openstack14:11
scasbut i'm also not changing the mission14:11
dhellmanntechnically it was on my list yesterday, too, but I didn't get to it14:11
smcginnisLooks like we may have an outcome for packaging-rpm at least.14:11
smcginnisHah, I know how that goes.14:11
scasthat's my armchair take on it, anyway14:12
scasif it's taking a wholly separate direction from how it came in, then different paths should be considered, no matter how distasteful14:13
*** ricolin has joined #openstack-tc14:14
scasif the mission is still openstack-focused, then there is still alignment. if it's becoming agnostic, perhaps it has a chance of standing alone14:16
*** hongbin_ has joined #openstack-tc14:16
scasi believe that's what happened with gnocchi, no?14:16
dhellmannscas : are you talking about trove changing its mission?14:16
scasdhellmann: in a slightly rhetorical sense, at this point, yes. apologies, breakfast kicking in14:17
dhellmannok. I didn't see anything about that happening, so I couldn't tell if you were thinking out loud or if I missed something.14:18
scasif the intent from the new team is to keep things on the current heading, i see it like health insurance. check the boxes, and the machine may not reject the claim14:19
cdentdhellmann: you in any kind of contact with ian bicking: I was long ago, but I'm not sure I have quick entree14:19
scasadmittedly i need to reread up on the whole trove thing. i've been following it somewhat14:20
dhellmanncdent : I think Ian switched to javascript a long time ago, so I don't know if he has anything to do with paste.deploy any more14:20
cdentthat's kind of the issue in a nutshell14:20
cdentit looks no one else does14:22
smcginnisThere's been some good work done in keystone to move to flask and get rid of paste. Maybe a future cycle goal?14:23
* cdent sends some email14:23
cdentindeed, but it's currently broken under 3.7 so there's interim ground to cover14:23
scaswhen chef openstack joined as a project, there was some initial, well-intended scrutiny. but, we all followed the process out in the open. the comment, at the time, was that we were conditionally approved on grounds that we adapt to the then-big-something-not-to-be-named. being that it's a matter of not having landed any commits, i'd say that process-wise the ship has more or less sailed. that said,14:28
scasexceptions are currently being considered for latecomers14:29
scasone possibility for continuity could be that the rocky ptl be more or less continued in stein, with 100% of the tasks delegated to the individual seeking to take the reins, then an 'official' transition in T14:31
scasshort of out-n-back-in, that seems like the least dramatic thing14:31
scasanother possibility is a tag of incubation and that become the new-old-stackforge for a probation type thing. again, options.14:37
scasmy perspective comes from framing it if the situation happened to me, where someone wanted to take things over without having followed the existing processes14:42
scaslicense and code-wise, anyone can fork and do as they please. in the arena that OpenStack: The Project operates, process is like... breathing14:44
*** pabelanger has joined #openstack-tc15:01
cdenttc-members, that time again15:01
fungihey howdy15:01
mnaserhello :>15:01
fungistill catching up on 100 lines of scrollback in here, be gentle ;)15:01
*** gouthamr has joined #openstack-tc15:01
mnaseri have been trying to dig the email about the trove PTL in lists.o.o but i havent been able to15:01
scasohai is my line!15:01
cdentseems like "what to do about PTLs" and "so, this paste thing" are issues of the day15:03
ttxTo clarify, for Searchligth and Winstackers, i was not suggesting no PTL, I just did not have any easy answer15:03
pabelangermnaser: ?15:04
ttxdhellmann: do you have that etherpad ready or should I create it ?15:04
mnaseri think there was an email that was more recent but in that thread so ill look15:04
mnaserit was yesterday or today or something15:04
mnaser"An HTML attachment was scrubbed..."15:05
dhellmannttx: I haven't created one yet15:05
ttxon it15:05
* dhellmann is fighting with the new expense report tool15:05
ttxsounds superfun15:05
dhellmannit has lots of "features" to "help" make things "easier"15:05
scasthere's another part of the thread15:06
scasthe two parts appear to be detached at the above links15:07
mnaserso about trove15:07
mnaseri'd be more comfortable with anyone on that team that has openstack contributons15:07
*** clarkb has joined #openstack-tc15:08
zanebme too tbh15:08
mnaser was the first and only patch15:09
dhellmannit appears that no current trove contributor is willing or able to serve15:09
TheJuliamyself as well15:09
mnaseri guess the question i want to ask myself (and other tc members) is.. what can come out of this if we hand it over?15:09
TheJuliaWhich makes me wonder... is it time to perhaps remove it from governance, and let them do their own thing and see where it goes?15:10
mnaseri mean, i doubt they'll change the mission of the project.. and it's better than removing them (as that feels a bit punishing)15:10
dhellmanndo we have precedent for the situation?15:10
cdentAs much as it pains us, I think we need to recognize that trove needs to reset itself and the best way for it to accomplish that may be outside of governance15:10
mnaserlooking in positive light this could be *the* team that can bring trove back and maybe we might lose out by shutting them down15:10
scashasn't the trove project already signaled a direction change in the past?15:10
ttxtwice yes15:11
scasmy wet memory seeems to recall that from yestercycle15:11
mnaseri think so too15:11
TheJuliascas: indeed :(15:11
mnasermy question is -- what bad can come out of handing it over?15:11
cdentIf none of the corporations out there were willing to take care of trove, than trove isn't "okay" and we shouldn't have projects that aren't "okay"15:11
ttxIBM took over from Tesora, then a diverse group of Chinese contributors took over15:11
ttxbut now Samsung Poland has a group more committed... just the PTL election coming at the wrong time (slightly too soon)15:11
mnaserwhile it makes me uncomfortable, i don't think there's anything bad about this, and if we notice anything weird (i cant imagine anything), we can step in.15:11
zanebit's like a serial collaboration at this point15:12
mnaserso subjectively i don't see any issue, i don't see them trying to take over the project or change the mission15:12
smcginnisI do wish trove would have gone with my suggestion of keeping their current PTL just as an advisor to help the new team get onboard.15:12
ttxThere is potential for OVH to help too, once they are done migrating their DCs15:12
cdentI don't know. I want being official to be meaningful and signal something to the outside world, not just to the internals of the community15:12
scasthe intent, to me, seems to be for continuity15:12
zaneba company steps in, runs it like an extension of their internal development for a bit, then steps out and passes it on to someone else who does the same15:12
ttxand OVH has their dev teams in Poland, so synergy possible here15:12
smcginniscdent: ++15:12
scasnot necessarily to come in with grand ideas, but keep their own use case alive. i do that.15:12
scasbut that's extrapolating and inferring a bit15:13
dimscontinuity can be there even without being in governance right?15:13
ttxI'd rather give them a chance to save it rather than punish them for having their involvement badly timed with our election process.15:13
cdentdims: I think so, yes15:13
mnaseri agree with ttx15:13
scasdims: quite. gnocchi did the same, no?15:14
ttxWe did the same when IBM took over15:14
cdentttx it's not really punishment for now, it's more finally catching up for our lack of action long ago15:14
ttxThey had a vanity commit in to pass the ATC bar15:14
cdentbut yes, that is likely how it would feel15:14
zanebttx: how did that work out?15:14
TheJulia+1 to ttx's position, lets see where they go, keep an eye on it, and go from there15:14
dhellmannttx and mugsie are the liaisons attached to trove; how about if we have them actively monitor the project?15:14
smcginnisDoes Dariusz at least have any patches proposed that might be close to being merged?15:14
ttxzaneb: let's say that the failure might have had more to do with the versatile nature of the company investing15:14
cdentI feel like we've been saying "lets see where they go, keep an eye on it, and go from there" about trove for years (?)15:15
mnaserjust blocked based on CI, smcginnis15:15
TheJuliacdent: good point15:15
mnaserand a patch that seems to have been throughly reviewed too15:15
ttxzaneb: here I see potential to have a group forming, with one product vendor (Samsung) and one user (OVH)15:15
scasthe failing tests look like cruft to me15:15
ttx+ the group in China still being involved15:15
scasmysql, postgresql15:15
ttxzaneb: the big difference is that Tesora dropped it on IBM and stopped doing stuff15:15
scas'openstack' has firmly centered on mariadb at this point15:16
ttxzaneb: in this case Zhao Chao said he would stay around to help15:16
dhellmannbut he does not have time to serve as even nominal PTL?15:16
ttxhe just feels his involvement will not match what Samsung is ready to invest15:16
zanebttx: but not around so much that they would take smcginnis's suggestion to be the figurehead PTL15:16
ttxdhellmann: he might be convinced of being a placeholder PTL... maybe15:16
smcginnisI would feel better if Zhao took the role for one more cycle.15:16
mnaseri mean, i'm sure we've asked him a few times about being placeholder PTL, but he might have his own personal reasons15:17
dhellmannmy impression from one of the threads was that he had already said no to that15:17
fungion the trove front: looking at our roll generation there are 69 individuals who qualified to run as trove ptl. and yet none of them volunteered15:17
mnaserwhat about asking Maciej?15:17
mnaserhe seems to be a core reviewer on the change that Dariusz pushed up15:17
clarkbscas: the idea is to be able to deploy those databases for users with trove regardless of what openstack itself is running15:17
fungithe tacker and senlin elections have smaller contributor sets and they had multiple candidates15:17
ttxmnaser: could be an option15:17
mnasersometimes we just need to ask and OVH is much more openstack-based so it could be a lot more justifiable15:18
scasclarkb: too early to technology still. right.15:18
zanebtbh I'm just very sceptical of any company's promises to invest in anything. I want to see PTLs with a track record of actually being involved in the project15:18
scaspromises without proof are the stuff of politicians15:18
mnaserzaneb: the thing is, technically, they could have had a single commit the whole cycle, nominated themselves and still would have gotten PTL.15:18
mnaserso while i agree abotu the 'track record PTL' .. dunno how much that really means, especially when not a lot of people are stepping up15:19
mnaserhow long is our cycle now?15:19
scasthe timing of the election may not be ideal, but someone wanting to take the wheel needs to have tangential awareness of the existing process15:19
zanebscas: you're aware that Trove is a project that deploys databases, right?15:19
dhellmannmnaser: did that not happen because of timing? or because of a lack of available reviewers?15:19
zaneb(re <scas> 'openstack' has firmly centered on mariadb at this point)15:19
scaszaneb: as my brain is waking up, i'm remembering that. i haven't deployed trove with chef in years15:19
ttxI feel like it's a mismatch between what Samsung is planning to do and what Zhao can dedicate to coordinate15:19
mnaserdhellmann: it seems like a change was pushedu p here on july 27th and i'm not sure if that qualifies --
ttxIf we explained that he could be PTL in the book but delegate everything...15:20
fungiscas: in trove's case, i think they want to test that it can provide mysql and postgresql, not necessarily use them internally15:20
scasright. i misspoke.15:20
zanebmnaser: they could have, and still didn't manage it :)15:20
smcginnisttx: I kind of did that, but maybe we could reinforce the idea.15:20
dhellmannI'm not sure asking someone to be PTL and then delegate everything to someone else really solves the problem of having an active contributor serve in the role15:20
fungibecause that's the service trove is in the business of (and dropping those would mean loss of backward compatibility for users relying on them)15:20
ttxsmcginnis: the subtlety might have been lost in translation. Worth a new try15:20
mnaserhow about we ask him to be PTL for a month or two, and delegate everything.15:21
fungioh, and clarkb beat me to that point15:21
mnaserif samsung folks keep up, we hand things over15:21
mnaserif samsung folks don't keep up, we take out of governance15:21
* fungi is still about 20 lines behind but catching up15:21
TheJuliamnaser: like that idea15:21
dhellmannasking Zhao to serve in name only feels like us fitting the situation to the rules, and not actually solving the problem15:21
mnaserso we're giving a chance but we're not handing a project.  feels like a reasonable comprimise15:22
dhellmannif he's not going to be actively involved in the project, then we have an absent PTL and a defacto leader without that responsibility15:22
* mnaser notices Bartosz who's at samsung15:23
cdentdhellmann: I agree.15:23
ttxwhat I'm saying is that we should see if the new team is successful before removing the project, rather than remove the project because we expect it will fail. Election is not the only time we can demote a project15:23
smcginnisIf Zhao is unable or unwilling to continue, I'm on the fence between removing from governance, or just letting Samsung take over without being ATC and keeping an eye on things.15:23
cdentIf we want to fiddle the rules every single time something happens, let's just go back to doing everything on a case by case basis...15:23
dhellmannttx: I agree with that.15:23
dhellmannthe APC rule has to do with the elections; we have no such restriction on appointments15:24
ttxdhellmann: if the election had happened two months ago or two months later, we would probably not be having that discussion15:24
dhellmannso it's not "fiddling" the rules, it's dealing with a situation that is outside of the norm15:24
smcginnisttx: To be fair, I am not considering removal because I expect it to fail. More that they would have more flexibility not being under governance until they get things in good shape.15:24
mnasersmcginnis: i agree but i think they'd feel like an outsiders with that15:24
dhellmannare any of them contributors to other projects?15:25
ttxdhellmann: How about we wait a bit before appointing someone?15:25
scasgovernance has allowed chef openstack to exist with a shrinking team for six cycles, five of which are with a high fragility factor. if trove is asked to leave and reapply to conform to the process, that spotlight invariably is cast in directions like mine, unless i just clam up and push code15:25
mnaserif anything, it would probably feel like "ah, well, we got our welcome to openstack by being 'kicked' out of governance" heh15:25
mnaserdhellmann: i found who is on the core team in trove..15:25
ttxi.e. not ask Zhao to sign up for 6 more months, but for one more month15:26
ttxtime for the transition to happen or fail15:26
ttxfixing the timing mismatch15:26
mnaserwhat if we just leave no ptl till october (ptg time)15:26
cdentwhat are the objections to appointing Dariusz and waiting and seeing?15:26
fungiyes, with the trove ptl candidate situation, appointing the volunteer we have would be entirely within our purview. normal election process and candidate qualification exists to provide a default mechanism where the tc doesn't need to intervene. one of the reasons we're here is to step in when consensus and process break down15:26
scasif i had a full team approaching me, i'd want to allow a soft transition. probably not a full cycle, because i'd feel like i'd be more smothering than soloing now15:26
ttxcdent: that would work for me as well15:26
zanebso they have a core reviewer on the team, but they want some other guy with a token commit submitted last week to be the PTL15:26
mnasercdent: none by me15:26
mnaserzaneb: the core reviewer has 3 commits15:27
mnasernon-merged and non-reviewed15:27
smcginnisOK, I'm slight on the "Vote for Dariusz" team at this point.15:27
mnaseri have an idea15:27
ttxsmcginnis: with e caveat that we might step in and interbvene if that goes nowhere15:27
mnaserwhy don't we push up a governance change to add dariusz and we can vote on it15:27
smcginnisAnd we could discuss in Denver and see if we need any course corrections.15:27
zanebmnaser: ah, ok. they did have another contributor whose commits have actually merged I believe15:27
cdentWith [t 2aJt] I wasn't suggesting it. I was asking.15:27
purplerbot<cdent> what are the objections to appointing Dariusz and waiting and seeing? [2018-08-02 15:26:29.485883] [n 2aJt]15:27
smcginnisOnly real objection is he is not an ATC, and that was one of our rules.15:28
cdentI continue to think that trove probably should have expired months ago.15:28
dhellmannzaneb : we know that writing code and leading teams are different, and that not everyone wants to do both. So I would rather we consider willing volunteers than draft someone15:28
cdentATC is for elections, not appointments?15:28
smcginnisTechnically true I think.15:28
ttxcdent: the Rocky team was pretty active keeping it afloat15:28
dhellmannright, there are no hard rules about appointments, specifically to allow us flexibility15:28
smcginnisSo I suppose we are within bounds to appoint Dariusz.15:29
ttxsmcginnis: totally15:29
ttxwe did apoint people that were not APC before15:29
cdentGiven that we have someone who is actually willingly volunteering, that's important, right?15:29
dhellmanncdent : yes, I think it is15:29
cdentPlenty of people who are PTLs in long running semi-healthy projects do it reluctantly15:29
smcginnisI would hate to squash someone's desire to get involved.15:30
ttxsmcginnis: ++15:30
pabelangersmcginnis: agree15:30
zanebI'm coming around to the idea of giving it a shot15:30
ttxsmcginnis: even if I think we need to warn that we are watching the situation unfold15:30
ttxbecause it's rather unusual15:30
smcginnisNext step? Patch?15:30
cdentzaneb: was your original resistance because a lack of proven openstack experience?15:30
dhellmannso if we appoint Dariusz, are there any conditions we would want to apply? such as extra monitoring of the team?15:30
zanebcdent: basically, yeah, especially when there appear to be alternatives who do at least have some15:31
mnasercan we be clear on what is 'extra monitoring' for example15:31
mnaseri.e.: they should contribute code?15:31
ttxdhellmann: if they end up not picking it up, not running open meetings whatever... we would come back on our decision15:31
ttx(or make trove unofficial)15:31
smcginnisdhellmann: I would be fine just putting it on the agenda to review the situation at the PTG.15:31
mnaserok so, doing things in the open15:31
dhellmannmnaser : I mean having the TC liaisons keep an eye on how the team operates, since one of the alternatives we discussed was removing Trove from governance.15:31
mnasersmcginnis: that seems like a reasonable timeframe for us to follow back on the status15:32
dhellmannsmcginnis : I think we'll want to do it more often than that, over the course of steein15:32
mnaserdhellmann: ack, so make sure meetings run, etc15:32
ttxAs one of teh TC liaisons for Trove, that would be my recommendation15:32
smcginnisYeah, probably multiple check-in's, but at least a TC review at the PTG.15:32
zanebcan we document somewhere the criteria under which we'd decide to make them non-official?15:32
dhellmannI wonder how much they'll actually have time to do between now and then15:33
fungii'm not in favor of saying an appointed ptl should commit code. that sounds backwards. code changes should happen for good reasons and not to qualify someone for a position. extra-atcs can run for ptl without ever contributing code changes15:33
zanebit seems clear that elections are not the right way to be triggering such decisions15:33
ttxif Poland in August is anywhere like France, probably not much15:33
ttxzaneb: agreed15:33
smcginnisfungi: That's a good point too.15:33
dhellmannthe election situation is another example of issues with this particular team, so while it's not reason enough to remove them from governance it's a trigger for the monitoring15:33
ttxthat's why I think we should just appoint whoever steps up (within certain bounds)15:34
ttxand ten apply extra caution15:34
mnaseri'm in agreement with TheJulia15:34
mnaserer, ttx , but also TheJulia  in general :p15:34
ttxI'm in agreement with TheJulia15:34
smcginnisAlways agree with TheJulia15:34
ttxsafe bet15:34
ttxI also blame flaper8715:35
dhellmannso, do we have consensus that we would accept Dariusz as PTL and that we would have ttx and mugsie keep an eye on trove and report back periodically?15:35
TheJuliaI agree15:35
ttxworks for me15:35
ttx(planned to keep an eye on it anyway, as I'd like to get OVH involved too)15:36
dhellmannok, maybe we can get Dariusz to propose the patch to make himself PTL then15:36
cdentwith the caveat that at some point the length of our "keep an eye on trove" needs to timeout and error15:36
cdentotherwise it starts seeming a bit desperate and cloying15:36
dhellmanncdent : technically we're supposed to be keeping an eye on all teams always15:36
zanebcdent: ++15:36
cdentdhellmann: indeed15:36
dhellmannttx and mugsie : can you talk to Dariusz about proposing that patch to governance?15:37
ttxyes with the health tracking stuff at least we have a timeout/error loop15:37
ttxdhellmann: eumel8 was planning to talk to him. I can help there15:37
dhellmannok, thanks15:38
dhellmannnoted in the etherpad15:38
ttxI just need to sync with eumel8 first to avoid collision15:38
dhellmannhow about Dragonflow?15:38
ttxDragonflow sounds like no brainer15:38
dhellmannOmer is the current PTL15:38
smcginnisI think this isn't the first time they've missed the election, but I agree.15:39
ttxwe always have groups missing the deadline15:39
dims+1 to Omer15:39
zaneb+1 that ones seems easy15:39
dhellmannsince he's the PTL already, there's no patch to write. How do we affirm him?15:40
dhellmanndo we go through the motions of removing and re-adding him? or is there some less silly-sounding way to vote? :-)15:40
pabelangerremoving / re-adding is just an extra patch right?15:41
* dhellmann taps mic15:41
dhellmannwell, we wouldn't normally remove someone if they were serving another term15:42
smcginnisWhere are these officially recorded after the election?15:42
dhellmannI'm thinking of openstack/governance/reference/projects.yaml15:42
zanebwouldn't the election officials submit an update after the election removing anyone who failed to nominate?15:43
pabelangerthat's what I was just about to ask :)15:43
dhellmannthey normally submit the update for any changes, but I don't know about removing a name from a team when there was no candidate15:43
dhellmannwe could certainly ask them to do that15:44
smcginnisSince that doesn't call out which cycle, I think we can just decide here or on the mailing list and not need to do anything with projects.yml15:44
dhellmannsince this is election-related, I would feel more comfortable with a more formal vote in place15:44
smcginnisIt should be recorded here probably:
dhellmannI would have to look at the yaml validation for that repo to see if it's even possible to have the PTL contact info be empty15:45
smcginnisMaybe a patch to add that with a note that it was TC appointed?15:46
dhellmann*that*the governance15:46
smcginnisAnd +1 votes from a majority of TC.15:46
dhellmannthat might work, what do others think?15:47
cdentfair enough15:47
ttx  - elected: TC-APPOINTED15:48
dhellmannok, so I'll propose a patch to add any appointments in the openstack/election repo15:48
smcginnisYay, precedent.15:48
*** AlanClark has joined #openstack-tc15:48
dimswelcome AlanClark15:48
dhellmannnext up, Freezer15:49
dhellmannttx and mugsie are the liaisons there, too15:49
smcginnisNo one has stepped up for that yet, right?15:49
AlanClarkthanks dims - sorry for not joining earlier had passwd trouble15:49
dhellmannsmcginnis : not as far as I know15:50
dhellmannthe current ptl is szaher15:50
dhellmannis this another case of someone not paying attention to the calendar?15:51
fungialso there's no rush on the ptl appointments in my opinion. we can propose one or more changes with our appointments on top of the change from the election officials noting the elected ptls15:51
dhellmannttx, mugsie, have you talked to that team or checked in on them yet?15:51
smcginnisdhellmann: Probably15:51
dhellmannfungi : I agree15:52
fungiwe easily have the remainder of election week to deliberate on appointments15:52
fungiand new information _could_ come to light between now and when polling closes on the tacker and senlin elections (doubtful but possible)15:52
smcginnis++, we definitely shouldn't rush.15:53
dhellmannwhat do we want to do next for freezer? have our liaisons contact szaher?15:54
smcginnisProbably a good next step for that one.15:55
dhellmannditto for loci, I assume?15:56
smcginnisYeah, looks like it.15:56
ttxdhellmann: unfortunately no15:57
ttx(I planned to check on them next week)15:57
dhellmannok, I laid out what I think are next steps for each team in except for RefStack15:58
dhellmannttx, mugsie, pabelanger, dims, cmurphy|vacation, zaneb, mnaser, cdent : your names all appear next to action items there15:59
mnaseri will do that, i dont thin kthey responded to my reaching out email initially too15:59
dhellmannlet's track status in that etherpad, so when you email them make a note, etc.16:00
* ttx runs16:00
dhellmannif we have to escalate to emailing the whole core team or something we can deal with that later16:00
zanebcmurphy is away so I will take the packaging-rpm one16:00
dhellmannzaneb : thanks, please cc cmurphy on that (teamwork!)16:00
dhellmannso, what do we want to do with refstack?16:01
dhellmannchanging the WG to a SIG and having them own the repo feels popular, but we haven't heard from the interop team about that yet16:02
fungiwgs can also own repos, fwiw16:02
dhellmannok, that's another approach16:02
dhellmannfungi and TheJulia : as the liaisons, can you get in touch with the interop folks and see what they want to do?16:02
fungii wouldn't necessarily tie the wg->sig move for inetrop together with winding down the refstack team and reassigning their repos16:02
fungidhellmann: i've been in touch with them about this since well before the election, notes are in the health tracker about it16:03
dhellmannyeah, if we can do something simpler that's good16:03
* fungi waves to mugsie's cat16:03
* mugsie 's yubikey :)16:04
mnasersearchlight's PTL has not had a commit in 2 months, did not respond when i contacted for the check-in16:04
dhellmannfungi : ack, thanks, so it looks like there's a minimal RefStack team. How does the Interop WG look?16:04
dhellmannAlanClark : for context, the situation we have is that no one stepped up to serve as PTL for the RefStack team, so we're trying to figure out what to do16:04
fungithe sticking point when i spoke with the refstack team last is that the service project itself was a few patches away from being deemed "feature complete" and put into maintenance16:05
dhellmann fungi : sorry, now I see the interop group notes separate from the refstack notes16:05
fungiand so it seemed they wanted to get that last bit behind them before folding refstack into the interop wg16:05
dhellmannis someone actively working on that? should we just carry over with the same PTL?16:05
fungiyeah, hogepoge and mguiney were working on that piece16:06
fungithey may be done now, it was a few weeks ago16:06
fungithe other bit is the tempest autodiscovery library which is being used by refstack but also has gained some life outside refstack/interop concerns entirely16:06
fungithat deliverable doesn't necessarily make sense to put under the niterop wg16:07
fungier, interop wg16:07
dhellmannthat feels like something the QA team might be willing to take?16:07
fungithat's the thing... it wasn't under qa because the tempest maintainers felt it was very much not a feature they wanted16:07
fungiso if it got moved to qa it would likely be manitained by entirely separate people from tempest anyway16:08
fungiit _may_ make sense to form a new official team around that library, or... host of other possibilities ranging from unofficial to mubmle-mubmle-oslo-something-mubmle16:09
TheJuliaPerhaps worth looking at the contributors to that library and see where they perceive it best fitting?16:09
dhellmannI'm reluctant to encourage the idea of oslo being a dumping ground for things other teams don't want16:09
dhellmannit was meant to be a set of tools common to many teams, not a set of things no one else would maintain16:10
fungiyeah, i talked to some of them in the refstack meeting. they were unsure but they've had a few weeks to think it over more since then so we should revisit16:10
TheJuliaThat sounds like a plan then16:10
*** eandersson has joined #openstack-tc16:10
dhellmannfungi : ok, I'll make a note for you and TheJulia to do that and we can talk about it again when you have newer info16:10
fungii agree, putting their team under oslo makes little (or at least less) sense than qa. and both seem like an excuse for not just applying as a new team or renaming/rescoping the refstack team16:10
fungior if they feel havnig a dedicated official openstack team for that lib is overkill, just make it unofficial16:11
dhellmannyeah, although it also feels a bit weird to have one team producing a tool to fill a gap that another team doesn't want filled16:11
dhellmannsure, that's another option16:12
fungii don't think it's so much that the tempest team doesn't want that gap filled, they just don't want it filled in tempest because they deem it out of scope/off mission and so also won't contribute to maintaining it16:12
dhellmannah, ok16:12
dhellmannI guess I don't think of the QA team as "the tempest team" so much as "all the test tools team"16:13
fungii don't think they have any problem with some other team unrelated to tempest maintaining an autodiscovery solution for generating tempest configs16:13
dhellmannwell, that's good, at least we aren't working at cross purposes16:13
fungiright, it's all a question of at what point does forcing all qa-oriented teams to be subteams of one master qa team take us back to the old "program" days16:14
dhellmannfair point16:14
dhellmannlet's see what the contributors have to say about it16:15
fungithe split in the refstack team is a pretty natural one though, the people maintaining refstack and refstackclient aren't the same people maintaining python-tempestconf16:17
cdent(this raises the point of devstack and grenade not being anywhere near as healthy as tempest, from what I can tell (but I've not had a response from my queries to gmann yet re health))16:17
*** spotz has quit IRC16:18
zanebto my mind it's unfortunate that the QA team sees themselves as effectively the "Nova QA" team. it becomes a self-fulfilling prophecy.16:19
* dhellmann makes a note to remind tc-members to try to complete those health checks before the ptg16:19
cdentzaneb: I'm glad you said that.16:20
fungizaneb: returning to their roots? ;)16:20
toskywell, also the QA team suffers from a lack of contributors too16:21
*** jpich has quit IRC16:21
toskyAndrea started the effort to port also grenade to zuul v3, but he probably stopped16:21
zanebtosky: hence the self-fulfilling prophesy comment16:21
dhellmannI'm torn on that. On the one hand we had a team like Documentation which was trying to do more than they really could. On the other hand we wanted cross-project teams to focus on helping and building tools, rather than doing.16:21
* dhellmann wanders off to find lunch16:22
zanebdhellmann: the QA team works like the old documentation team, where they have a small set of projects they care about and manage themselves, and everybody else is on their own16:23
zanebbut anyway, let's not reopen that whole debate from February16:23
toskyfungi: but many (if not all) relevant features of tempestconf since the last PTG were implemented as requirements for the refstack integration16:24
mugsiezaneb: it is starting again - just look at patrole (the RBAC test tool)16:25
toskyzaneb: didn't you just describe the new documentation team, which manages the infrastructure and everything else (the real documentation) is in the hands of each project?16:27
fungitosky: yeah, it seems like that's a phase though, since refstack isn't planning any further development but python-tempestconf will likely continue to grow?16:27
zanebtosky: no?16:27
zanebnova/cinder/glance/neutron tests are in-tree in Tempest. if they were all plugins managed by the projects themselves then it would be like the new docs team structure16:28
pabelangerdims: dhellmann: talking with #openstack-loci, it seems SamYaple will step up again for stein to be PTL again. I'll update etherpad, it seems he is also going to propose a patch to elections again16:30
dimspabelanger : gotcha16:31
toskyfungi: that's true, the status was different during the last PTG16:33
toskyzaneb: and swift and keystone; but that was for interoperability tests (keystone, cinder and neutron have additional tempest plugins)16:34
pabelangerokay, yes. SamYaple does look to want to be PTL for loci for stein, he does say he lost the majority of his small team 3 months ago16:34
zanebpabelanger: that's concerning. isn't openstack-helm also dependent on loci?16:35
pabelangerzaneb: I'm not sure how dependant helm is on loci ATM, but I can look into it a little.16:38
zanebmy understanding was that kolla-k8s (based on kolla, obvs) shut down, in part because openstack-helm (based on loci) had sucked up all the oxygen16:39
*** tosky has quit IRC16:41
*** ricolin has quit IRC16:44
mugsiezaneb: wasn't openstack-helm the AT&T team that now is doing airship?16:44
dtroyermugsie: basically, yes16:45
*** dtantsur is now known as dtantsur|afk16:46
mugsieyeah, stackalytics has a lot of common names for the 216:47
*** zaneb has quit IRC16:48
portdirectopenstack-helm started out from the same group within AT&T, and we have significant overlap for sure16:48
portdirectairship is a way of managing the life cycle of clusters, running helm packaged applications16:49
portdirectborn out of a need for us to be able to manage them - OSH being the 1st and prime example.16:49
portdirectone point of correction i would make however, is that OSH supports both LOCI and Kolla docker containers16:52
portdirectand have voting gates for both16:52
pabelangerdhellmann: cdent: I've sent an email out to rocky PTL for winstackers, and updated etherpad16:56
cdentthanks pabelanger17:01
openstackgerritHongbin Lu proposed openstack/governance master: Add openstack/os-ryu to neutron
*** AlanClark has quit IRC19:25
dhellmannthanks pabelanger19:28
*** tellesnobrega has quit IRC19:28
*** cdent has quit IRC19:35
*** masayukig has quit IRC20:18
*** tosky has joined #openstack-tc20:50
openstackgerritSean McGinnis proposed openstack/governance master: Add Stein goal for upgrade checkers
dimspabelanger : dhellmann : openstack-helm seems to depend on loci master (,openstack-helm-addons,openstack-helm-infra )21:04
smcginnisCinder block-box does as well.21:18
*** mriedem is now known as mriedem_afk21:28
*** hongbin_ has quit IRC22:39
*** tosky has quit IRC22:59
*** edmondsw has quit IRC23:08
*** jaypipes has quit IRC23:14
*** jaypipes has joined #openstack-tc23:16
*** ricolin has joined #openstack-tc23:51

Generated by 2.15.3 by Marius Gedminas - find it at!