15:00:19 <portdirect> #startmeeting openstack-helm
15:00:20 <openstack> Meeting started Tue Mar  5 15:00:19 2019 UTC and is due to finish in 60 minutes.  The chair is portdirect. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:23 <openstack> The meeting name has been set to 'openstack_helm'
15:00:41 <portdirect> hey - lets give it untill 5 past for people to arrive
15:01:08 <portdirect> the Agenda for this week is here: https://etherpad.openstack.org/p/openstack-helm-meeting-2019-03-05
15:01:14 <portdirect> #topic rollcall
15:01:42 <evrardjp> o/
15:01:47 <hogepodge> hi
15:01:52 <howell> o/
15:02:08 <srwilkers> o/
15:02:10 <roman_g> o/
15:03:37 <itxaka> o/
15:03:54 <dwalt> o/
15:04:08 <cheng1> o/
15:04:11 <georgk> hi
15:04:51 <portdirect> ok - lets go :)
15:04:55 <mattmceuen> o/
15:04:55 <portdirect> #topic Multiple distro support for images
15:05:13 <evrardjp> for images or for helm charts too? ;)
15:05:42 <evrardjp> I know someone who will hate me soon
15:05:46 <portdirect> i think they are one and the same no? we are not an image building project (as much as possible anyway!)
15:06:00 <evrardjp> I am just toying with you portdirect :)
15:06:16 <portdirect> theres been some really good input on this this morning in the etherpad - thanks evrardjp
15:06:29 <portdirect> but i think we probably need more time here
15:06:30 <evrardjp> don't mention it, it's my job
15:06:47 <portdirect> as until today theres only been myself contibuting to this
15:06:48 <hogepodge> portdirect: if only there was an image building project person here...
15:06:59 <evrardjp> hogepodge: :D
15:07:07 <portdirect> or three :D
15:07:09 <srwilkers> ^
15:07:51 <evrardjp> anyway -- what deadline do we set portdirect for input, before going to spec?
15:08:01 <portdirect> we had tried for today
15:08:11 <portdirect> but that was depenent on things being hashed out
15:08:14 <evrardjp> sorry, what deadline do we set, portdirect, for receiving inputs of everyone*
15:08:26 <portdirect> next week i think
15:08:37 <portdirect> so we can review at the next metting and then go to spec
15:08:42 <portdirect> work for you?
15:08:48 <srwilkers> another week seems sane to me
15:08:59 <evrardjp> that's long  :p but it doesn't prevent us to implement stuff in charts already
15:09:27 <portdirect> not that long - hopefully the we can work on the etherpad in the meantime
15:09:27 <evrardjp> I think I would like to expedite this, so I might even start on a draft spec in the same etherpad -- so I can gather more feedback
15:09:45 <portdirect> i know everyones been busy - buts is a shame no one had time to work on it for 6.5 days ;)
15:10:03 <portdirect> evrardjp: feel free to start a deft in the pad
15:10:07 <evrardjp> portdirect: that's how life works sadly
15:10:39 <evrardjp> will do tomorrow -- so that everyone gets still today to have a first view of my etherpads' comments
15:10:39 <portdirect> anyway - we ok to come back to this next week?
15:10:46 <evrardjp> sure thing
15:10:55 <portdirect> thanks - lets move on then
15:11:06 <portdirect> #topic Zuul credentials and next step for image publishing/consumption
15:11:13 <portdirect> evrardjp: you're up :)
15:11:32 <evrardjp> oh yeah, so I worked with the infra team, and refactored the jobs last week or so
15:11:45 <evrardjp> we should be ready with the latest patch of portdirect
15:12:01 <portdirect> nice - thats great evrardjp
15:12:05 <evrardjp> it means we'll have to do a first PR on the relative images to build
15:12:21 <evrardjp> and then they will be published for consumption in dockerhub in openstackhelm namespace
15:12:34 <evrardjp> I will not delve into the technical details here
15:12:59 <evrardjp> so I will continue with the plan: the next step would be consuming them
15:13:23 <evrardjp> For the images that are published, that will be easy, just use the right path if it's not the right one already
15:13:36 <evrardjp> But the interesting part is the depends-on
15:14:12 <evrardjp> I have a spec that explains all of that, which could have some eyes. mattmceuen had a +1 but I manually rebased and lost it.
15:14:27 <portdirect> do you have a link to that?
15:14:37 <evrardjp> https://review.openstack.org/#/c/615623/
15:14:38 <evrardjp> sorry
15:14:42 <evrardjp> #link https://review.openstack.org/#/c/615623/
15:15:22 <portdirect> roger - lets get some eyes on this
15:15:46 <portdirect> is it some variation on the themes of maintaining a equiv of this: https://github.com/openstack/openstack-helm/blob/master/tools/overrides/releases/pike/loci.yaml, that we then source in the gates?
15:16:03 <srwilkers> that would be my preference i think
15:17:27 <evrardjp> it is exactly that
15:17:53 <evrardjp> it's just that pike is a rolling target (latest pike)
15:18:13 <evrardjp> the spec goes a little beyond by having pike-myownthing and pike-0.1.0 for example
15:18:19 <alanmeadows> o/ sorry I'm late
15:18:26 <portdirect> awesome - that sounds good evrardjp
15:18:40 <evrardjp> it would mean ofc different overrides files based on context, default keeping it rolling.
15:19:39 <evrardjp> that's all I have.
15:21:02 <portdirect> ok - thanks evrardjp for all your work here
15:21:14 <portdirect> its really helping move us forward
15:21:29 <portdirect> #topic Helm sync meeting
15:21:43 <portdirect> hogepodge: the floor is yours
15:22:07 <hogepodge> I spoke with Lachlan at KubeCon Seattle
15:22:37 <hogepodge> He's open to having periodic meetings with the team to help out with 3.0 development
15:23:02 <portdirect> hogepodge: this would be great
15:23:05 <hogepodge> I don't necessarily need more meetings, but it's something I can facilitate if the team would find it useful (as it did a few years ago)
15:23:17 <portdirect> the ones we had for helm 2 were incredibly useful
15:23:31 <alanmeadows> Would this be a sep set of calls from the recurring Helm 3 Dev calls Chris?
15:23:38 <alanmeadows> or just sort of a slot on there
15:23:58 <hogepodge> there are helm 3 dev calls?
15:24:02 <portdirect> yes
15:24:35 <hogepodge> I think that the best approach would be to try and invite Lachlan or an appropriate Helm rep to those meetings and not add another spot to the calendar
15:24:48 <portdirect> hogepodge: that would work
15:25:03 <portdirect> howell has been going to them a bit i think
15:25:09 <howell> yes
15:25:35 <howell> i've been attempting to get them to add new issues for the community to work on
15:25:53 <hogepodge> ah, so the whole thing may be solved already? :-)
15:26:08 <evrardjp> hogepodge: impressive efficiency there
15:26:17 <hogepodge> (I'm assuming you work on helm, howell? )
15:26:32 <alanmeadows> well, its rather divorced from the underlying workload
15:26:33 <howell> i've been working on it, yes
15:26:49 <howell> (though i wouldn't call myself a rep)
15:27:00 <alanmeadows> if we're looking for more of what we were doing
15:27:04 <alanmeadows> "application needs forum"
15:27:09 <alanmeadows> thats not exactly the helm 3 call
15:27:09 <portdirect> i think alanmeadows makes a good point here
15:27:12 <alanmeadows> as it is
15:27:26 <hogepodge> Ok, well I'll take the meeting time back to the Helm team in general and see if we can carve out some specific feedback time
15:27:58 <hogepodge> alanmeadows: to the point I just made, could we slot in helm 3 specific agenda during that meeting time?
15:28:22 <evrardjp> alanmeadows: I would like to distill what you just said into knowledge for others, can you come back to this conversation after the meeting in the #openstack-helm channel?
15:29:17 <alanmeadows> My only point was I think -- even though it may mean more meetings - you're likely going to what something separate from the helm 3 recurring dev call that exists, or get their permission to tack on an item for application owners to voice
15:29:37 <alanmeadows> evrardjp: sure
15:31:16 <evrardjp> thanks
15:31:38 <hogepodge> portdirect: alanmeadows: I should get some more information from you on the helm 3 call and see what the best solution is, and also coordinate with the helm team. It sounds like the interest is there, and we should pursue it regardless of the forum.
15:31:54 <alanmeadows> Those meetings were always very useful
15:32:03 <alanmeadows> For all involved
15:32:16 <portdirect> ++
15:32:25 <hogepodge> I have to drop out of the meeting to prep for the next staff meeting, but I can ping portdirect offline to figure stuff out.
15:32:48 <portdirect> hogepodge: sounds great - thanks dude
15:33:11 <portdirect> i think this brings us nicely to the next suprise topic
15:33:26 <portdirect> #topic meeting times and office hours
15:34:00 <evrardjp> oh great!
15:34:03 <portdirect> whe have reviewed meeting times on a couple of occeasions, and each time come to the conclusion that the earth not being flat is an issue
15:34:30 <portdirect> I tried to flatten it last weekend, though this was largely unsuccessful
15:34:42 <evrardjp> yeah if we could all get the sun at the same time it would be a solution too
15:34:58 <evrardjp> portdirect: did you fall on it enough hard?
15:35:07 <evrardjp> maybe we need to sync there.
15:35:34 <portdirect> should we hold another poll to get feedback on any potential alternat times?
15:35:47 <alanmeadows> Its always good to reassess
15:35:53 <evrardjp> correct
15:36:06 <portdirect> also would it be worth considering having meetings at alternate times?
15:36:09 <evrardjp> I like this meeting time tbh, but I think we lack office hours in different timezones
15:36:46 <portdirect> eg - flip flopping between a westcoast(usa)/asia and eastcoast(usa)/eu focused time?
15:37:13 <portdirect> or would office hours be better for this?
15:37:34 <srwilkers> I don’t mind the idea of flip flopping
15:38:30 <evrardjp> I like the idea of flip flop with a focus on APAC and West coast
15:38:57 <evrardjp> that would give jayahn more natural times for feedback
15:39:05 <portdirect> exactly ^
15:39:24 <evrardjp> however I think we should separate the two conversations -- one to set the meeting time (maybe with a poll)
15:39:35 <evrardjp> then another to set the office hours slots
15:39:41 <portdirect> agreed
15:39:45 <srwilkers> That would be nice, but I’d also propose reevaluating that after a period of time to determine whether it’s making a difference
15:39:49 <evrardjp> they have two different purposes
15:39:52 <portdirect> I'll put together an email for the ml and get that out today so we can take this convo there
15:40:03 <evrardjp> portdirect: great
15:40:19 <portdirect> ok - so lets move on
15:40:25 <portdirect> #topic roundtable
15:40:27 <evrardjp> srwilkers: that's fair. but We need to take a few time, not reassess after 2 weeks
15:40:45 <evrardjp> maybe a cycle or something
15:40:59 <portdirect> i'd say probably half a cycle
15:41:07 <srwilkers> evrardjp: agreed, but I’m also saying it shouldn’t be 8-12 weeks before we do
15:41:10 <evrardjp> due to where we are in the cycle, that's more like it portdirect
15:41:34 <portdirect> ok - sorry i jumped the gun with the context switch
15:41:49 <portdirect> we have anything else we want to discuss this week?
15:42:01 <evrardjp> I have nothing to add
15:42:22 <DavidSull> Hey, figured I'd check here. Could someone review this change https://review.openstack.org/#/c/638249/ ? Thanks
15:42:55 <evrardjp> thanks DavidSull
15:43:11 <portdirect> DavidSull: yes!
15:43:41 <DavidSull> Much appreciated
15:43:42 <portdirect> sorry this dropped of my radar, but looks great to me, I'll just a have a quick check over of the syntax after the meeting but overall lgtm
15:47:12 <cheng1> evrardjp, portdirect could you help review this patch, I have updated it today https://review.openstack.org/#/c/639087/
15:47:56 <portdirect> cheng1: will do - the code was fine to me before, so should be pretty straight forward
15:48:24 <cheng1> portdirect, thanks :)
15:49:47 <evrardjp> cheng1: will do
15:50:01 <portdirect> ok - lets wrap up and give everyone 10 mins back
15:50:07 <portdirect> thanks everyone!
15:50:11 <portdirect> #endmeeting