14:00:21 <mihalis68> #startmeeting Ops Meetup Team 14:00:24 <openstack> Meeting started Tue Nov 21 14:00:21 2017 UTC and is due to finish in 60 minutes. The chair is mihalis68. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:28 <openstack> The meeting name has been set to 'ops_meetup_team' 14:00:28 <mihalis68> I don't see any regular attendees 14:00:37 <mihalis68> perhaps people are already on vacation 14:00:51 <mihalis68> #link agenda https://etherpad.openstack.org/p/ops-meetups-team 14:01:50 <mihalis68> hello 14:02:02 <shintaro> hi 14:02:07 <mihalis68> there's nobody around 14:02:26 <shintaro> oh. shall we wait for few minutes 14:02:39 <mihalis68> yeah 14:02:44 <mihalis68> thanks for filling out the agenda 14:03:00 <mihalis68> so sponsoring breakfast is estimated at $2k? 14:03:16 <shintaro> yes. including morning coffee 14:03:23 <mihalis68> nice. I'll ask if we can cover that 14:03:43 <mihalis68> the question about the eventbrite funds will have to go to a foundation member. 14:03:52 <med_> \o 14:03:56 <mihalis68> hi! 14:04:02 * med_ just woke up, not a work day 14:04:18 <shintaro> I will ask Clair about the eventbrite funds 14:04:23 <mihalis68> you beat all but 2 here! 14:04:35 <shintaro> hi 14:04:57 <mihalis68> #topic actions 14:05:18 <mihalis68> last week we only had one - emccormick to join the LTS debate on the mailing list. COMPLETE 14:05:33 <mihalis68> that debate has taken off on the mailing list, and on an ether pad (see link in agenda) 14:05:52 <mihalis68> there's talk of converting the topic into a SIG instead of spanning operators mailing list and developers mailing list 14:05:52 <med_> that they dragged that discussion to -sig was kind of annoying. 14:06:24 <mihalis68> somebody else commented privately to me that is sometimes a way to play nice but effectively bury an issue 14:07:20 <shintaro> I wonder how many people are in the SIG-ML 14:07:25 <med_> well, just meant I had another ml to join and also didn't have older messages in my mail queue. 14:07:31 <mihalis68> if anyone else had older actions, feel free to put them on the agenda. It's too hard to go back and scrape old meeting minutes to find incomplete actions 14:07:45 <mihalis68> I wasn't subscribed, certainly (now fixed) 14:08:35 <mihalis68> #action shintaro is going to contact foundation (Clair) about using Eventbrite funds for T-Shirts 14:08:36 <shintaro> I need to check the Foundation about the graphics and logos for the OpsMeetup 14:09:24 <shintaro> I will also ask how many have registered so far. 14:09:42 <mihalis68> seems we've moved on to the tokyo event, so #topic Tokyo Ops Mid-Cycle Meetup 14:10:07 <mihalis68> is breakfast+coffee the main sponsorship opportunity? 14:10:10 <med_> #topic must be first on the line I think 14:10:33 <mihalis68> #topic Tokyo Ops Mid-Cycle Meetup 14:10:36 <mihalis68> thanks! 14:10:58 <shintaro> yes mihalis68. breakfast+morning coffee is the main sponsorship so far. 14:11:00 <mrhillsman> o/ 14:11:05 <mihalis68> good morning! 14:11:09 <shintaro> hi 14:11:38 <mihalis68> #mihalis68 to confer with colleagues over whether Bloomberg can pick up breakfast+coffee sponsorship for Tokyo 14:12:03 <mihalis68> I have not checked in with the planning ether pad. How are we doing for technical agenda planning? 14:12:24 <mihalis68> Have you firmly decided on two tracks called Telecom and Other, respectively? 14:12:55 <mihalis68> As I recall there's only 3 rooms, so it can't be very parallelised 14:13:14 <shintaro> right one large and two small 14:14:21 <shintaro> if we are going to do hackathon-like documentation session, we will be needing one room for that. 14:14:42 <mihalis68> #link https://wiki.openstack.org/wiki/Operations/Meetups/TYO-ops-meetup 14:15:08 <mihalis68> #link https://etherpad.openstack.org/p/TYO-ops-meetup-2018 14:15:41 <shintaro> no update on NFV track 14:16:03 <mihalis68> That seems to need some clarification. Is it "NFV" and "General" tracks and that's it? 14:16:15 <mihalis68> Themed track is vague if you've called it NFV previously 14:17:04 <shintaro> I was thinking to have two themed track, one is NFV and one is Enterprise. one day each 14:17:17 <mihalis68> it's your company's event as host, so I think you should just decide and represent the decision in that ether pad. Then It think we should try go get the community's attention 14:17:35 <mihalis68> hmm, interesting. In my imagination, tracks imply parallel 14:18:00 <mihalis68> if it's themed days I'd call them days, not tracks 14:18:15 <mihalis68> everyone else please weigh in here, it feels like typing at shintaro right now :| 14:19:10 <med_> I'd agree--tracks indicates parallelism, days is what I would call distinct focuses of each day 14:19:49 <shintaro> ok it would be themed days then. 14:19:53 <med_> and I think having that type of focus is fine--not at all concerned whether we do tracks or days. 14:19:56 <med_> sounds fine. 14:20:03 <mrhillsman> I'm indifferent...potatoe potahto 14:20:36 <mrhillsman> agreed with med_ 14:20:40 <mihalis68> think of it like software requirement stories : I am a telecom focused attendee and I wish to attend all the telecom-focused sessions 14:20:45 <mihalis68> is that person only attending 1 day? 14:21:59 <shintaro> if that person is not interested in any other topic, maybe yes, but general and enterprise sessions will cover many topics that may interest telcos as well, IMHO 14:23:13 <mihalis68> so it's two days with some mixed-audience sessions on each but in addition day 1 is focused on NFV? 14:23:37 <mrhillsman> not sure we have to get that specific 14:23:43 <shintaro> that's my idea 14:24:24 <mrhillsman> comment was meant to go further up 14:24:25 <mihalis68> I think when people look at such events, it's very important for them to understand clearly what it will be like 14:25:06 <mrhillsman> agreed 14:25:20 <mihalis68> let me see if I have it, shintaro. The event is 2 days of general operators related technical sessions. In addition, day 1 has an NFV focus 14:26:04 <shintaro> yes and day 2 have another track which I now think would be "Enterprise track" 14:26:29 <mihalis68> so day one is general+nfv, day 2 is general+enterprise? 14:26:40 <shintaro> yes 14:26:45 <mihalis68> ok! I like that 14:27:00 <mihalis68> perhaps that's what the google doc intended to convey but I didn't quite grasp it 14:27:13 <shintaro> we have another room if anyone want to do some work 14:27:17 <mihalis68> cool 14:27:26 <mihalis68> seems like all the business for tokyo. I 14:27:43 <shintaro> I will update the etherpad 14:27:49 <mihalis68> I'll look at the google doc and try to help clarify. I can't edit it from my desk unfortunately (google docs is very locked down here) 14:28:40 <mihalis68> once it's good I propose this: email the meetups team for final comments. Unless there's an objection then share the planning doc again with the operators mailing list and ask people to start submitting session ideas 14:29:19 <mihalis68> taking silence as assent. Please object strenuously if I'm wrong 14:29:32 <shintaro> do you mean share planning etherpad? 14:29:43 <shintaro> or google doc? 14:29:44 <med_> concur 14:29:46 <mihalis68> #topic future mid-cycles - to co-locate or not that is the question 14:30:13 <mihalis68> yes I meant the ether pad. google doc for precise agenda comes later 14:30:22 <shintaro> okay got it 14:30:30 <mihalis68> #link https://docs.google.com/document/d/1BgICOa-Mct9pKwjUEuYp_BSD1V_GdRLL_IM-BU0iPUw 14:31:00 <mihalis68> The link I just posted is a discussion document about co-location 14:31:19 <mihalis68> it needs more work to capture the different views 14:31:43 <mihalis68> As I understood it, originally the idea revolved around some inter-working between the operators and the developers 14:31:54 <mihalis68> however the PTG is 5 solid days already filled with content 14:32:12 <mihalis68> last week after the meeting, I learned that there's some support for a separate event that just happens to be in the same facility 14:32:24 <mihalis68> this I (to be honest) don't really grok 14:33:05 <med_> well, they probably get the rooms/whatevs at the already discounted rate. 14:33:17 <med_> btw, how did you learn this? Lauren? 14:33:19 <mihalis68> I understand the bit about leveraging foundation logistics expertise to only have one venue to acquire 14:33:29 <med_> kk 14:33:35 <mrhillsman> tie in support from foundation as ops meet up need 14:33:41 <mihalis68> mrhillsman was telling to stop worrying about the PTG part of it 14:33:51 <med_> nods. 14:33:52 <mrhillsman> may grok better then 14:34:04 <mihalis68> but it raises as many questions as it answers 14:34:18 <mihalis68> would it be two separate events in evenbrite? Two teams doing agendas? 14:34:18 <mrhillsman> that's perfectly ok 14:34:34 <med_> V W was concerned about lack of focus and lack of ... pureness if they were concurrent. 14:34:38 <mihalis68> would it still be "PTG" or would it be called something more inclusive 14:34:54 <mrhillsman> sure, should be different as it is now 14:35:01 <mriedem> pureness? 14:35:20 <mrhillsman> the only point is, can this reduce some costs 14:35:41 <med_> I don't want to "wake" him, but something about keeping Operators as a discipline distinct from Devs mriedem 14:35:55 <mihalis68> unleash the Krake... the VW! 14:35:56 <mriedem> i'm not sure why we should be silo'ed 14:35:56 <med_> but I'm poorly representing his POV. 14:36:23 <mriedem> just because you're in the same venue doesn't mean you have to talk to each other 14:36:32 <mrhillsman> or not :) 14:36:33 <mriedem> it just makes it easier if you have agenda items that overlap 14:36:43 <mihalis68> that's true, but does the venue have to have multiple rooms with AV capabilities? 14:36:45 <mriedem> or need operator (or dev) input to something 14:37:06 <mihalis68> the sessions where we get the ether pad up on a big screen seem helped by that 14:37:15 <mrhillsman> talk in the evening :) 14:38:35 <mrhillsman> surely there is A/V caps in multiple rooms at a hotel 14:38:45 <mriedem> there was at the place in denver for the last ptg 14:38:46 <mriedem> yes 14:39:00 <mrhillsman> ^ 14:39:04 <mihalis68> yeah but think about NYC. There was clearly one main room 14:39:06 <med_> many actually 14:39:17 <mriedem> who coordinated NYC 14:39:18 <mriedem> ? 14:39:20 <mihalis68> I did 14:39:27 <mriedem> the foundation scouts out the location for the ptg 14:39:40 <mriedem> they obviously didn't scout the train noise in denver, but... 14:39:47 <mrhillsman> hehe 14:39:49 <mriedem> the venue itself was ok for the actual work part 14:39:51 <med_> yep, they look for the place with the most train horns 14:39:57 <mihalis68> I think for the event to be a combined PTG and ops Meetup it would need to be redefined as that 14:40:05 <mihalis68> and then foundation would plan a combined event 14:40:12 <mihalis68> we'd be a "track" effectively 14:40:19 <mrhillsman> it's not combined my friend 14:40:31 <mihalis68> it is combined from their point of view 14:40:34 <med_> collocated is what you are thinking mrhillsman 14:40:38 <mihalis68> same dates same location 14:40:54 <mihalis68> they are doing logistics - people, rooms, catering 14:40:56 <mrhillsman> who's viewpoint? 14:41:00 <mihalis68> the proposal combines those things 14:41:00 <med_> I think that's a matter of marketing/splitting hairs. 14:41:05 <mrhillsman> ^ 14:41:11 <med_> the devs define the content, tracks in ptg 14:41:14 <med_> not the foundation 14:41:15 <mrhillsman> which is not our concern 14:42:05 <mihalis68> the definition of PTG is on the foundation website. The definition should be expanded if the venue is hosting an operators meet up at the same time as the PTG, surely? 14:42:20 <smcginnis> This might help as far as putting on the event. But I'm concerned on whether it would help having devs involved in more ops discussions. 14:42:26 <mrhillsman> we are only discussing costs and man/woman hours on making the event happen by including staff for ops logistics needs 14:42:39 <mihalis68> we are discussing the whole thing 14:42:50 <mihalis68> because as a proposal it hasn't been accepted by the community 14:42:56 <mrhillsman> let the foundation be concerned with foundation concerns 14:43:10 <mihalis68> hi smcginnis, by the way! 14:43:16 <smcginnis> mihalis68: o/ 14:43:17 * med_ notes that he would be able to get more out of his travel dollars if they were co-located and get more from the event.... 14:43:42 <smcginnis> med_: That would be a benefit for sure. 14:43:55 <mihalis68> smcginnis I think it wouldn't help that, since the devs are at an event dedicated to dev issues, by definition 14:44:06 <mihalis68> I don't see random bar encounters as a substiture 14:44:11 <mihalis68> substitute 14:44:15 <mriedem> it depends on the topic in the room at the time, 14:44:17 <smcginnis> mihalis68: Yeah, I agree. 14:44:25 <mriedem> in denver, the first two days were cross-project things 14:44:33 <mriedem> like fast forward upgrades, interop, API SIG, etc 14:44:44 <mriedem> wed-friday are traditional meetup-style silo'ed roms, 14:44:46 <mriedem> *rooms, 14:44:52 <mriedem> but we do overlap for scheduled topics, 14:44:54 <smcginnis> During the PTG I was double booked almost the entire week as far as discussions I wanted to be in. This wouldn't help me, but it may help others. 14:44:56 <mriedem> like nova/neutron, nova/cinder, etc 14:45:27 <mihalis68> maybe PTG is defined too narrowly in the foundation page. At risk of repeating myself, having read it closely, it's not for me 14:45:41 <mriedem> from my pov, i can't travel to ptg and summit *and* an opts midcycle, so colocating 2 of those would be helpful 14:45:53 <lbragstad> ++ 14:45:54 <mriedem> mihalis68: yes they'd have to redefine that probably 14:46:06 <smcginnis> mihalis68: As it is right now, yes, it is not really meant for ops. 14:46:10 <mihalis68> PTG says you should be either on a dev team, or a subject matter expert where the subject is one of the development activities 14:46:17 <mihalis68> this is what I am getting at 14:46:37 <lbragstad> having some operator representation during the PTG would be nice - otherwise we're waiting two months to share what we came up with at the PTG with operators for feedback 14:46:37 <mihalis68> we now seem to have a view that co-location would be for inter-working 14:46:39 <med_> ah, but if they schedule another event simultaneously, that changes. 14:47:06 <med_> it could be another event at the same place -OR- they could change the definition. I think either approach would be fine. 14:47:06 <mihalis68> I'd totally attend PTG if I felt remotely encouraged 14:47:10 <mriedem> lbragstad: well you wait 2 months for the summit, in australia, where no ops have budget to travel anyway :) 14:47:23 <lbragstad> or that 14:47:23 <mihalis68> at least those two days which seem set up for cooperation 14:47:48 <mrhillsman> lbragstad PTG should be based on previous Forum discussions 14:48:03 <mriedem> mrhillsman: not necessarily 14:48:14 <mriedem> the ptg replaced the midcycle and dev summit, 14:48:18 <mrhillsman> According to the change yes 14:48:20 <mriedem> so it's part planning for the upcoming release 14:48:27 <med_> I've been at both PTGs. Knowledgeable, contributing operators were very welcome. 14:48:36 <smcginnis> med_: ++ 14:48:37 <mriedem> and it's part talking about shit that is still broken and needs work, w/o 40 minute time boxed sessions 14:48:38 <mrhillsman> Looking at the lifecycle graphic and explanations 14:49:04 <mrhillsman> It is supposed to move towards that 14:49:48 <smcginnis> I think in reality we want to have some dev representation at ops events and some ops representation at dev events. The Summit is really what's conceived as the focus for full interaction at this point, but there should probably always be a little mix. 14:49:51 <mihalis68> well this is much easier for me to understand : first two days of PTG become officially slightly broader and we declare those to be effectively the mid-cycle meetup 14:49:56 <lbragstad> not sure if we were supposed to - but our project spent a significant amount of time redesigning some stuff that has cross-project implications... that's where we really wanted operator feedback (like 10,000 ft view of what we came up with) 14:50:53 <mihalis68> the question for those who in any way represent the dev teams is : is that acceptable to them? 14:51:00 <med_> mihalis68, if you want a 2d on that idea, I'm the 2d 14:51:17 <med_> thingee, ^ 14:51:21 <mrhillsman> the issue with that change is what VW was alluding to 14:51:28 <mihalis68> thanks med_ ... am I wrong or is this idea quite different from where we ended last week? 14:51:32 <mrhillsman> losing ops focus mihalis68 14:51:51 <med_> mihalis68, it's different from where we ended but closer to where we started a few weeks/months ago. 14:51:58 <mriedem> "first two days of PTG become officially slightly broader and we declare those to be effectively the mid-cycle meetup" - describes, to me, what used to be the old dev summit format 14:52:12 <med_> the folks endorsing this idea SEEM TO BE KEY DEV LEADERS and it seems like they approve. 14:52:13 <mriedem> except it's not mid-cycle at all, since it's at the beginning of the cycle 14:52:23 <mrhillsman> exactly 14:52:39 <mrhillsman> if we stop trying to put the two events together we can move forward 14:53:06 <med_> and understand that we (ops) won't have even kicked the tires on the just released bits, that feedback will have to come at the Forum in the summit still. 14:53:07 <mrhillsman> it comes down to ops midcycle asked foundation for logistical support 14:53:23 <mrhillsman> this is an option to ease foundation work, costs, etc etc 14:53:25 <mrhillsman> in doing that 14:53:27 <mihalis68> mrhillsman you've communicated your opinion on this, but there are others 14:53:39 <mrhillsman> if it will not, we should not 14:53:56 <mrhillsman> because we will lose the "pureness" of ops, as VW mentioned 14:54:13 <mihalis68> what I'm seeing above is a desire to get ops into PTG, at least on days 1 and 2 from several people 14:54:24 <mihalis68> not just at same hotel, but in same room, discussing shared challenges 14:54:36 <mrhillsman> yes, that makes sense, the PTG site you referenced covers this no? 14:55:12 <mihalis68> no PTG *as defined on openstack foundation website* seems to specifically welcome just dev team and software SMEs 14:55:12 <smcginnis> I think collocation could be useful as long as it's not intended as combination. It at least gives the opportunity for cross-collaboration. And being around each other for evening drinks can definitely have benefits too, even if the daytime sessions have too many conflicts. 14:55:42 <smcginnis> mihalis68: I consider you all SME's. :) 14:56:00 <mihalis68> the S is not any specific chunk of software 14:56:01 <mriedem> i think he means SME on a particular project 14:56:05 <mihalis68> yeah 14:56:09 <mriedem> like cburgess on nova 14:56:09 <med_> fwiw, entire teams (cinder/nova) frequently join together for an hour or two (same any time during the week) as well as individual stakeholders/experts cross pollinating on an as needed/as owned basis 14:56:21 <med_> ie, the etherpad says: Grab Jay Pipes for this one 14:56:29 <med_> or grab miihallis68 when we talk about this 14:56:30 <smcginnis> Yes, but still subject matter experts in general on a lot of things. 14:57:03 <mihalis68> I run an openstack team and in the past I built all our clusters, but I wouldn't consider myself an expert on any specific topic in the list 14:57:07 <mrhillsman> how many operators attend ptg, i think we mentioned 5 or six 14:57:21 <mihalis68> we sent nobody who actually does things here. 14:57:29 <mrhillsman> if the content of the overlap days is still ptg specific 14:57:34 <mrhillsman> how does that change? 14:57:35 <mihalis68> Literally if any people represented bloomberg they do not participate in on-call support at any level 14:57:52 <mriedem> for nova, we generally get 2 ops 14:57:57 <mriedem> med_: and cbrugess 14:58:00 <med_> but you could have--I think there was a tradeoff/self-selection 14:58:01 <mriedem> *cburgess 14:58:40 * med_ nods, was there all week, but it was also "local" 14:58:41 <mrhillsman> my point i guess is, if there is no change in content right now, and no change in content if we go this route, how is there a change in attendance 14:58:45 <mihalis68> I could have attended, but it seemed not for me. My team all came to same conclusion (I asked them to consider going) 14:58:51 <med_> but traveld to atlanta for same reasons 14:59:00 <med_> kk 14:59:02 <mihalis68> we're gonna run out of time here. 14:59:09 <mrhillsman> right mihalis68 so why push combining the content/schedule/days 14:59:15 <mihalis68> on the agenda I reminded people of the converted Operators Guide 14:59:17 <mriedem> we do get better ops attendance at the forum for sure 14:59:41 <mihalis68> mrhillsman it was suggested as a possible good thing by PTG attendees during this very meeting 15:00:12 <mihalis68> e.g. sean 15:00:20 <mrhillsman> i think that was an evolution of the focus on combining events 15:00:39 <mihalis68> no, I don't think so at all 15:00:43 <mrhillsman> vs staying on the focus of ops midcycle needing help on logistics 15:00:51 <mihalis68> there is one viewpoint that says get ops in for days 1 and 2 15:01:30 <mrhillsman> i think last meeting we spent quite a bit of time trying to keep the conversation on the latter 15:01:39 <mihalis68> out of time 15:01:41 <mihalis68> #endmeeting