Monday, 2019-03-11

*** ricolin has joined #openstack-tc00:59
*** jamesmcarthur has joined #openstack-tc01:41
*** jamesmcarthur has quit IRC01:48
fungiyep, that's the plan02:06
fungii'm waffling on the idea of including that list as subject tags, but that would be one crazy long list02:07
*** whoami-rajat has joined #openstack-tc03:07
fungiokay, reminder sent03:39
*** jamesmcarthur has joined #openstack-tc03:39
fungii think at least cloudkitty seems to have had a prospective candidate e-mail their platform to the openstack-discuss ml but not propose it to the election repo yet03:39
fungialso a candidate just nominated themselves for tricircle while was sending that03:40
fungioh, and zun also has one waiting for a second official to confirm03:53
fungiso we're down to 29 missing now including those03:53
*** marst has joined #openstack-tc04:11
*** jamesmcarthur has quit IRC04:32
*** marst has quit IRC05:00
*** jaosorior has joined #openstack-tc05:52
*** ianychoi has quit IRC06:32
*** ianychoi has joined #openstack-tc06:32
*** ianychoi has quit IRC06:35
*** ianychoi has joined #openstack-tc06:36
*** e0ne has joined #openstack-tc06:51
*** e0ne has quit IRC07:02
*** gouthamr has quit IRC08:18
*** e0ne has joined #openstack-tc08:27
*** dtantsur|afk is now known as dtantsur08:38
*** tosky has joined #openstack-tc08:41
*** zbr has joined #openstack-tc08:54
*** jpich has joined #openstack-tc08:57
*** jpich has quit IRC08:57
openstackgerritMerged openstack/governance-sigs master: Remove lyarwood from upgrades SIG  https://review.openstack.org/64196208:58
*** needssleep has quit IRC09:01
*** jpich has joined #openstack-tc09:02
*** e0ne has quit IRC09:22
*** e0ne has joined #openstack-tc09:42
ttxtc-members: we'll have another open infra newsletter this week, if you'd like to propose openstack news snippets to include please see https://etherpad.openstack.org/p/newsletter-openstack-news10:12
ttxI was thinking of mentioning being past feature freeze, and the TC election results / PTL election start10:13
ttxsmcginnis: maybe a one-liner about how the Ops meetup went?10:14
*** e0ne has quit IRC10:43
*** e0ne has joined #openstack-tc10:47
dtantsurttx: one thing that came from the ops meetup: they may need your help (e.g. via this newsletter) to reach out to moar people10:48
*** ricolin has quit IRC10:59
*** gouthamr has joined #openstack-tc10:59
asettledtantsur, that's interesting esp with the Rackspace layoffs. Lots of stackers went from ops and dev. Might be worth trying for a combined effort?11:01
dtantsuryeah11:01
evrardjpttx: lgtm11:35
*** e0ne has quit IRC11:37
*** edmondsw has joined #openstack-tc11:47
*** marst has joined #openstack-tc12:03
*** e0ne has joined #openstack-tc12:30
*** jaypipes has joined #openstack-tc12:31
*** ijolliffe has joined #openstack-tc12:34
*** jamesmcarthur has joined #openstack-tc12:48
*** dangtrinhnt_x has joined #openstack-tc13:08
*** jaypipes has quit IRC13:08
smcginnisttx: That might be nice to mention the ops meetup. Frank has written up a nice article (with lot's of pictures) for superuser. But highlights were 30 attendees, great venue and lunch provided by Deutsche telecom, and feedback that it was a useful event and worth going.13:09
*** marst has quit IRC13:19
dhellmanntc-members: there are some interesting ideas about consensus vs. agreement in this post: https://thenewstack.io/this-week-in-programming-rough-consent-for-project-management/13:20
*** zaneb has joined #openstack-tc13:21
evrardjpthanks dhellmann13:21
*** ianychoi has quit IRC13:28
*** ianychoi has joined #openstack-tc13:29
*** jamesmcarthur has quit IRC13:30
*** jamesmcarthur has joined #openstack-tc13:31
*** lbragstad has joined #openstack-tc13:35
*** jamesmcarthur has quit IRC13:36
*** mriedem has joined #openstack-tc13:39
mnasertc-members: i'd like to look into merging https://review.openstack.org/#/c/641878/ soon if no one objects (esp if fungi can give a +2 given his involvement in the py3 stuff :))13:39
fungisure, will take a look once i'm caught up on ptl nominations13:40
*** lxkong has quit IRC13:42
*** marst has joined #openstack-tc13:43
*** jamesmcarthur has joined #openstack-tc13:45
*** jamesmcarthur_ has joined #openstack-tc13:49
*** jamesmcarthur has quit IRC13:53
*** dangtrinhnt_x has quit IRC14:14
*** e0ne has quit IRC14:14
mnasertc-members: there seems to be a lot of questions raised about our usage of using a template name linked with release name.  some are very valid.  do we want to consider bringing this up for discussion?14:15
zanebmnaser: we discussed this for months on the review of the resolution (which everybody ended up voting for)14:16
fungii didn't raise it as a concern in the original discussion because it seemed like an implementation detail14:16
evrardjpwhat's the concern?14:17
evrardjpI might have missed that14:17
mnaserreview comments here https://review.openstack.org/#/c/641878/ -- about using release name in project template14:17
fungifigured when it came time to actually implement, we'd seek input from job configuration reviewers on the most efficient implementation14:17
mnaseri do feel it is an implementation detail as well, and given zuul's powers and efficency, it would be beneficial to get input from the infra team about the best way to go about getting our end result14:18
mugsiewe will be keeping in on stable branches in the future14:18
fungibut like i said, i'm willing to give it a try. just thinking it's going to be a massive change of thousands of lines we have to script every time we transition between cycles when it could just be a few lines14:18
mugsieso stable/train will have these jobs forever (or  until we eol the branch)14:19
evrardjpit seems everyone agrees, so I am moving on :p14:20
mugsieI like dhellmann's idea of applying it centrally, but that can follow on14:20
fungimugsie: i don't see why that makes any difference though, whether it's multiple templates with per-release names (again though, why aren't the branch filters in the template?) vs one template name with multiple variants for the different branch filters14:20
zanebfungi: it could be if we were prepared to break lots of projects' gates just to change there unit test config - something that is 100% not worth it, and which we discussed about 50 times during the resolution review14:21
fungithe latter allows it to be applied once to relevant projects and then adjusted centrally by creating a new variant of that name each cycle with the new branch filters14:21
evrardjpmugsie: I think we discussed this on the initial implementation14:21
evrardjpzaneb: I seem to remember that too14:21
fungizaneb: okay, so that means dhellmann's suggestion of applying centrally is out of the question, since instead of one change with thousands of lines you're now talking about hundreds of changes14:21
mnaserthat's going to take a huge toll on our infra overall, eh?14:22
zanebI think people are misinterpreting dhellman's comment14:22
fungiand i don't think we want the central config reviewers reviewing those, so they would have to be applied in each project14:22
dhellmannno, I think fungi understands exactly what I meant14:22
dhellmannI think we want 1 template per series, and I think we want 1 patch per series to apply that to all projects at one time14:23
*** e0ne has joined #openstack-tc14:23
dhellmannif that breaks jobs, then projects can have their in-tree overrides change the voting rules until they can get things fixed14:23
zanebdhellmann: oh, I thought you were saying the templates were good because we can do stuff like (the future equivalent of) drop py35 across the project without having to push updates to individual repos14:23
dhellmannthey are, yes14:23
fungiif we're applying it to all projects at the same time anyway, we can do that with one variant per series but keep the same template name so it doesn't have to be added to and removed from every project independently14:23
dhellmannI think if any part of this requires us to change a setting in a project tree it's going to fail14:23
dhellmannfungi : I want a repo under TC control that lets us add jobs to projects14:24
dhellmannwe may delegate that to the qa or infra team, but I want a place where we can say "you must run these test jobs"14:24
fungidhellmann: yep, that's what the current project-config repository should eventually become, since it grants some body of configuration reviewers control to enforce job selection across all projects in the "openstack" tenant14:25
dhellmannotherwise we're going to continue to have the issue we have with projects dragging their feet to meet the testing specs14:25
dhellmannok, cool, if we don't need a new repo that's fine14:25
dhellmannwhat I also want is to not have to learn zuul's branch specifier rules in order to look at a config file and see what jobs are run for $series14:26
fungithe idea is that we can't lean on the infra team indefinitely to keep on top of what the job policies are for each tenant, so the tc would delegate reviewing the tenant configuration for openstack to a group of people who are familiar with openstack's testing policies14:26
openstackgerritMerged openstack/governance master: members: convert to yaml  https://review.openstack.org/64170014:26
dhellmannI would be happy with that14:27
dhellmannperhaps the qa team is a good candidate for that responsibility14:27
smcginnisI'm still partial to this approach - https://review.openstack.org/#/c/637866/14:27
fungia comment in each variant saying what series it's for would probably suffice, but also the job browser/api for zuul is nearly to the point where you can show per-branch views of jobs run on a project (right now there's a ui issue being addressed where it displays the wrong branch name, but i expect that will be rectified in the next few days)14:28
dhellmannsmcginnis : that looks pretty similar to zaneb's patch https://review.openstack.org/#/c/641878/14:28
dhellmannhow are they different?14:28
smcginnisNot just py3, though that's a very minor distinction.14:29
dhellmannfungi : it's not just a matter of understanding what is already configured, it is a matter of understanding how to add the configuration needed for a new series14:29
dhellmannsmcginnis : oh, yeah, I think there was a comment on zaneb's patch about that. I agree we should have templates for everything14:29
evrardjpdhellmann: I like the idea on shared ownership of code with TC being an active hand -- if that repo could be having cores from infra/QA/TC that would be great, IMO.14:30
smcginnisIt made sense to me, if we are going to have per-cycle templates, to include the whole set of expected jobs we want run for that cycle.14:30
dhellmannsmcginnis : ++14:30
evrardjpbut isn't that a little too forward thinking?14:30
smcginnisHeh, I don't think we've been accused of being too forward thinking too often. :)14:30
evrardjpI meant on dhellmann's comment about share ownership, smcginnis :)14:31
zanebsmcginnis: it's not a bad idea, but it only applies for the next 2 releases anyway. and by keeping py27 separate, we can allow projects that are py3-only to still use the templates14:31
smcginnisBut we shouldn't have any py3-only projects until all of them are in U, right?14:33
smcginnisAnd it could be expanded to include others. Say we want to have all projects run tox -docs or something.14:33
zanebofficial projects14:34
*** e0ne has quit IRC14:36
*** e0ne has joined #openstack-tc14:38
*** weshay68802228 has joined #openstack-tc14:44
dhellmannadjutant is 3 only, isn't it?14:46
dhellmannmaybe we solve that by special-casing them14:46
fungiwith the most recent flurry of nominations approved, we're down to 21 teams with no ptl candidate yet: Adjutant Blazar Cinder Congress Glance Heat I18n Ironic Magnum Monasca Neutron Octavia OpenStack_Helm PowerVMStackers Puppet_OpenStack Tacker Telemetry Tripleo Trove Vitrage Zaqar14:55
smcginnisI know Jay was going to do Cinder again but was waiting to give someone else a chance to put their name in first.14:56
fungiaround the 24 hour mark that count will hopefully be small enough we can reasonably add them as subject tags in the next reminder14:56
evrardjpfungi: should we worry about Octavia?15:00
mugsieI think we should ...15:00
evrardjpI can ping in OpenStack helm15:00
evrardjpI suppose rico will continue on heat?15:01
dhellmannI think we'll find there is no one willing to do it for Telemetry15:01
fungii think we should worry about any which don't have a candidate tomorrow. there's still ~1.5 days remaining in the nomination period15:01
evrardjpdhellmann: :(15:01
evrardjpfungi: Ex PTL was Rackspace, not sure about the current status :(15:02
evrardjpI mean current PTL ofc15:02
evrardjpfuture ex, whatever the term is :)15:02
mugsieoutgoing ;)15:02
evrardjpthanks!15:02
dhellmannthere's also only 1 contributor in qinling, iirc15:03
fungidhellmann: if that's the case, then i guess it's the ptl candidate who is nominated since nobody else would qualify15:03
mugsieoctavia have a plan apparently - so they should be good15:03
gmannyeah, there are many active contributor in Octavia15:04
evrardjpgood to hear15:05
evrardjpAnyone has pinged Erno?15:05
mugsieI did15:05
evrardjpgood :)15:05
mugsiehe is running15:05
evrardjpgreat15:05
mriedemcan i run as ptl for 31 other projects?15:18
mugsieif your an atc, but, the TC can take this as you volunteering to be appointed to 31 projects :D15:19
mugsieyou're*15:20
mriedemi would be like a god15:21
mriedemdrunk with power15:21
mugsieif it was me I would just be drunk15:22
smcginnishah15:22
gmann:)15:31
fungii don't think there's enough booze in the world to get me to volunteer to ptl 31 (well, now 21) projects15:32
dtantsurJulia just confirmed she plans to run again15:33
*** jamesmcarthur_ has quit IRC15:38
*** jamesmcarthur has joined #openstack-tc15:38
mnasergreat leader mriedem can be ptl for 21 projects15:44
dtantsur++15:52
*** weshay68802228 is now known as weshay15:53
*** jamesmcarthur has quit IRC15:57
*** jamesmcarthur has joined #openstack-tc15:57
*** jamesmcarthur has quit IRC16:00
*** jamesmcarthur has joined #openstack-tc16:01
mriedemtime for a serious query, where in the train schedule does the community wide goal deadline land? https://releases.openstack.org/train/schedule.html16:13
mriedemor is there no deadline if we can't agree on goals?16:13
mriedemor is it April 11 to align with the goal research item?16:14
smcginnismriedem: We've set the completion deadline, but we haven't done a selection deadline yet.16:17
smcginnisI think we've always just had a few lined up and made sure to pick some early-ish in the cycle.16:17
smcginnisWe could formalize it though and put something on the calendar to set a deadline for selection if that's what folks want.16:17
mriedemi was just wondering because i put the 2 that are open into the nova meeting agenda and that got me thinking about urgency16:18
*** marst has quit IRC16:18
mriedemif it's roughly april 11 at this point, then there is a month16:18
mriedembut not a big deal16:18
mriedemi just know not many people are looking at these proposals16:18
smcginnisI think an ideal situation would be to have them selected before the cycle starts.16:18
mnaserlbragstad: had a suggested timeline actually16:19
smcginnisI don't think that will change though. We never seem to get too much feedback on the goals.16:19
smcginnisUnless someone absolutely doesn't want to do them.16:19
mnaserso that he wanted that projects knew the goal ahead of time to slot time in for their ptg16:19
mnaserttx, fungi: do we have someone at the foundation that can help apply to this? https://opensource.googleblog.com/2019/03/introducing-season-of-docs.html16:20
ttxMight be a stronger statement if it comes from our Docs team? /cc diablo_rojo16:25
fungimnaser: reading https://developers.google.com/season-of-docs/docs/admin-guide and https://developers.google.com/season-of-docs/docs/organization-application-hints it seems like something anyone in the community could do16:25
lbragstadmnaser correct16:25
fungii agree with ttx, i was also thinking this makes more sense for the application to come from a representative of the docs team16:25
lbragstadmy primary concern with the goal selection process is making sure we capitalize on the opportunity of face-to-face communication16:25
lbragstadas opposed to selecting the goals too late and not being able to work on them during the PTG16:26
ttxI mean we can definitely facilitate, but yeah we should do that with the Docs team16:26
mnasermriedem, smcginnis ^ about the goal selection 'deadline'16:27
lbragstador having a "damn, i wish we knew these were going to be the goals two weeks ago during the PTG!" moment16:27
mugsiettx: I think the season of docs mentors should be docs / community people, but the admin may fit in the foundation (if there is people hours availible to do it of course)16:32
*** e0ne has quit IRC16:41
lbragstadflaper87 would you be opposed to taking a quick look and leaving feedback on https://review.openstack.org/#/c/641784/ since you were the original author? I'll happily incorporate the feedback into subsequent patches.16:41
ttxmugsie: diablo_rojo trackes mentorship and other on-boarding activities, so she should be able to help with coordinating that application with the docs team16:43
*** e0ne has joined #openstack-tc16:57
mnasertc-members: does anyone who enjoys some sort of 'event planning' would like to take on planning a TC dinner on saturday? seems like most of us come in that day anyways :) -- not that it would be a "tc only" event but at least figure out some logistics of a "place to go" and "attendees"17:10
*** e0ne has quit IRC17:13
lbragstadlast september a few of us when to https://www.denvermilkmarket.com/17:13
lbragstadwent to*17:13
*** tosky__ has joined #openstack-tc17:14
*** tosky has quit IRC17:14
*** e0ne has joined #openstack-tc17:14
lbragstadlooks like there are plenty of options close by though17:15
*** tosky__ is now known as tosky17:19
*** dtantsur is now known as dtantsur|afk17:19
*** ijolliffe has quit IRC17:20
ttxmnaser: I can organize something.17:23
lbragstadno shortage of places close-by https://tinyurl.com/y5fhypd517:25
*** jpich has quit IRC17:29
*** e0ne has quit IRC17:31
*** mriedem is now known as mriedem_afk17:31
mnaserthanks ttx.  i'm pretty boring when it comes to that stuff, better let others do it ;)17:35
*** jamesmcarthur has quit IRC17:40
*** jamesmcarthur has joined #openstack-tc17:41
*** jamesmcarthur has quit IRC17:45
dhellmannmnaser : maybe talk to stephenfin?18:12
dhellmann(about the season-of-docs)18:12
*** jamesmcarthur has joined #openstack-tc18:20
*** e0ne has joined #openstack-tc18:23
mnaserdhellmann: i posted to the ml, will try to follow up if i dont get a response18:24
*** e0ne has quit IRC18:25
*** jamesmcarthur has quit IRC18:26
*** diablo_rojo has joined #openstack-tc18:40
*** jamesmcarthur has joined #openstack-tc18:47
*** e0ne has joined #openstack-tc18:57
*** mriedem_afk is now known as mriedem19:04
*** e0ne has quit IRC19:40
*** e0ne has joined #openstack-tc19:40
e0nesmcginnis, mnaser: hi. I've got few questions about horizon cross-project activities19:43
mnasere0ne: sure19:43
e0neI would like to introduce integration tests job as a required to be listed as official maintained and supported plugin19:43
e0neI implemented few such jobs for several plugins already19:44
e0nebut it would be good to involve plugin teams too:)19:44
e0neI know that plugins are not maintained well because of lack of resources:(19:44
mnasere0ne: i think this is interesting and ties into similarly how cinder now requires third party ci for drivers .. somewhat.19:45
e0nefrom one side we don't have CI at all, from others, teams don't have capacity to fo it19:46
e0nemnaser: cinder did a great job on CI19:46
e0nemnaser: do we have some official procedure to start such activities?19:47
e0nemnaser: of course, I'll start with mailing list and note in the horizon documentation19:47
gmanne0ne: to be more clear, that integrated-job you mean to install all the plugins and perform some sanity test or even more OR only install and perfrom plugins specific testing19:50
e0negmann: tests like install plugin and open the page will be enough for the beginning19:51
gmanne0ne: ok that is not covered on plugin CI yet ?19:52
e0neI do understand that all teams have very limited capacity to implement it19:52
e0negmann: there are a lot of plugins from this list https://docs.openstack.org/horizon/latest/install/plugin-registry.html  without any kind of CI19:53
gmannhow we do in case of tempest plugin is. we do have a single n-v job on tempest which does install all plugins to check no installation error  so that it would fail the whole system due to any buggy plugin.19:53
gmannso this kind of job on horizon side can be useful for you to monitor such active/buggy plugin etc19:54
gmannbut yes i agree with you to add it on plugin side too.19:54
e0negmann: it's a good idea to have such job on horizon side, thanks:)19:56
gmannone way to make the job implementation easy is 1. horizon implement such job which does install all detected plugins and open page for them and 2. add that in a template 3. add that template  entry in plugins gate pipeline19:58
gmannthis way any plugin running this job will have only that plugin installed and testing it.19:58
gmannand plugin would not need to implement the job, just add the template entry in their CI19:58
*** lxkong has joined #openstack-tc20:00
e0negmann: it's not so easy to do for horizon, but the general idea sounds good20:00
gmannok, not much idea on horizon plugin mechanism . but implementing job on each plugin side is complex and leads to inconsistency for long term.20:03
gmannfor ref. this is how we do for tempest plugins - http://git.openstack.org/cgit/openstack/tempest/tree/.zuul.yaml#n34320:04
e0negmann: thanks for the help20:10
*** e0ne has quit IRC20:11
*** e0ne has joined #openstack-tc20:12
*** e0ne has quit IRC20:15
*** jamesmcarthur has quit IRC20:56
*** e0ne has joined #openstack-tc20:56
*** e0ne has quit IRC21:04
*** e0ne has joined #openstack-tc21:12
*** whoami-rajat has quit IRC21:25
*** e0ne has quit IRC21:38
*** marst has joined #openstack-tc22:04
*** marst has quit IRC22:16
*** tosky has quit IRC22:26
*** lbragstad has quit IRC22:40
*** lbragstad has joined #openstack-tc22:43
*** mriedem has quit IRC22:45
*** TheJulia has joined #openstack-tc23:01

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!