19:04:06 <zehicle> #startmeeting Refstack 19:04:07 <openstack> Meeting started Mon Jan 5 19:04:06 2015 UTC and is due to finish in 60 minutes. The chair is zehicle. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:04:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:04:10 <openstack> The meeting name has been set to 'refstack' 19:04:20 <catherine_d|1> o/ 19:04:25 <sslypushenko__> Hi to all! o/ 19:04:43 <catherine_d|1> Happy New Year to all! 19:05:14 <sslypushenko__> You to! 19:05:15 * sarob lurking 19:05:36 <hogepodge> o/ 19:06:29 <zehicle> we've got a lot to talk about since there's a f2f next week 19:07:05 <zehicle> agenda items: f2f agenda and logistics, API spec & updates 19:07:09 <catherine_d|1> yep .. Perhaps we should plan on the agenda 19:07:19 <hogepodge> +1 to agenda 19:07:32 <zehicle> gerrit backlog, capabilities lists 19:07:56 <zehicle> I'm not going to be there, who wants to lead the discussion about agenda? starting w/ start/end times 19:08:35 <catherine_d|1> We need Rocky ... 19:08:39 <hogepodge> The scheduled dates are January 12-14 19:09:37 * zehicle texting Rocky to see if she's online 19:10:14 <zehicle> #topic face to face meeting 19:10:24 <zehicle> sorry, not a good moderator :P 19:10:34 <hogepodge> That's three days. From what I gather Huawei is providing space and lunch through Rocky. 19:10:41 <hogepodge> Location address is 2330 Central Expressway, Santa Clara, CA 95050 19:11:06 <zehicle> Start/end times? 19:12:07 <hogepodge> I was assuming 9-5, but travel and work times may differ. My flight leaves San Jose at 7:00 pm Wednesday, so I won't be able to get a full day in. However, it would make sense to split our time up into six blocks 19:12:18 <hogepodge> morning/afternoon each day. 19:13:15 <hogepodge> Anyone else have any thoughts on that? 19:13:46 <zehicle> so 10am to 6pm? 19:13:47 <catherine_d|1> That is good for me ... 19:14:01 <zehicle> that means there are 6 blocks of time to schedule 19:15:08 <hogepodge> I'd like at least two devoted to hacking and dev work. Try to advance things in person with the end goal of having a fully working client/server. 19:15:48 <hogepodge> At least one needs to be planning future work. API additions and plans. 19:15:57 <zehicle> I'd put those in the middle 19:16:15 <zehicle> Maybe Tues AM & Wed AM? 19:16:47 <hogepodge> I think that's a good idea. 19:16:52 <catherine_d|1> Are we going to have working session that actually will work on code to get a few API up and running? 19:17:12 <catherine_d|1> or just planning session? 19:17:18 <zehicle> do we have an etherpad for this yet? 19:17:22 <hogepodge> catherine_d|1 yes, I'd like to have that. I don't want to leave without a working system that can collect results. 19:17:34 <catherine_d|1> hogepodge: +1 19:17:49 <catherine_d|1> at least have a couple working ... 19:17:59 <hogepodge> https://etherpad.openstack.org/p/refstack-january-2015-midcycle 19:18:11 <catherine_d|1> to set the framework at least 19:18:35 <rockyg> O/ 19:19:35 <catherine_d|1> Hello rockyg: 19:19:50 <hogepodge> I'd also like a session for UUID hacking. Move the implementation of the spec forward (and spend the week getting the qa team to review the spec in advance) 19:19:56 <catherine_d|1> is 10:00 to 6:00 sounds good? 19:20:00 <hogepodge> (by spend the week I mean this week) 19:20:33 <rockyg> Hi.might get dropped.On I 5 19:20:37 <hogepodge> Except for Wednesday that's fine by me. I actually prefer an earlier start, but I realize that others may have travel concerns from around the region. 19:21:01 <hogepodge> Of course, rockyg is the ultimate authority on space availability :-) 19:22:09 <rockyg> Lots of space.but traffic is a bear here. 19:22:33 <zehicle> I've been updating the page.... but since I'm not going to be in person, it needs review 19:22:51 <hogepodge> I'd also like a session on project organization. Try to set up a framework for regularizing the meetings, setting agendas ahead of time, and working on goals and project timelines. 19:22:59 <zehicle> earlier will be better for SergeyLukjanov and I 19:23:05 <zehicle> sorry sslypushenko__ and I 19:23:13 <rockyg> We could schedule docs/help sprints early and give time to coders to get to the show. 19:23:14 <hogepodge> Maybe Monday morning for that? 19:23:39 <rockyg> S/show/site 19:25:04 <rockyg> Works for me 19:25:32 <zehicle> ok, adjusting 19:26:25 <rockyg> Don't know if this was covered....could probably get a session with some infra folks local.Monty and maybe j Blair 19:26:29 <zehicle> davidlenwell, said that we'd have the API working by then 19:26:47 <zehicle> getting infra and/or QA engaged would be great 19:27:23 <zehicle> anything formal in the evening? 19:28:13 <rockyg> Let's let David Lenwell and me work the Infras 19:28:18 <catherine_d|1> On the Tuesday coding ... is that for API or UUID? 19:28:47 <zehicle> API is top prority 19:28:58 <catherine_d|1> zehicle: ++1 19:28:59 <zehicle> so, UUID only if the API work is moving 19:29:48 <hogepodge> yes. We already have some uuid code from sslypushenko in progress 19:29:57 <zehicle> other discussion for agenda? 19:30:06 <hogepodge> Dinner on Tuesday night? 19:30:24 <hogepodge> Monday night works too. 19:30:55 <rockyg> Let's see what day we can get infra 19:31:29 <zehicle> I added places for RSVPs so you can get a head count 19:31:38 <zehicle> #link https://etherpad.openstack.org/p/refstack-january-2015-midcycle 19:32:01 <rockyg> Thanks 19:32:36 <zehicle> next topic? 19:33:06 <zehicle> #topic API spec & updates 19:34:07 <zehicle> David and I talked on Friday about this. I suggested that we could decouple how we store the results from the API for the very short term 19:34:30 <zehicle> so some of the DB schema items could be deferred to unblock actually getting the API in use 19:34:53 <hogepodge> Sure. 19:34:54 <zehicle> that does NOT mean they are not import, I'm trying to work out an MVP that let's us start external testing of the refstack client 19:35:06 <sslypushenko__> I'm planning to propose new DB schema with will be universal 19:35:10 <zehicle> double negative - storing results is important 19:35:33 <hogepodge> I'd like us to consider building on the sql tools that the tempest team have built out for reporting and possibly storage. 19:36:09 <hogepodge> qa and infra have a solid interest in storing and managing results based on their needs. It feels right to build on their work. 19:36:19 <sslypushenko__> I think I will do it my tomorrow. So anybody can review it 19:36:51 <hogepodge> sslypushenko__ are you planning on going nosql? 19:36:55 <sslypushenko__> No 19:37:00 <hogepodge> +1 :-) 19:37:04 <rockyg> Cool 19:37:20 <sslypushenko__> It will slow down our development process 19:37:32 <zehicle> hogepodge, that could be interesting. I'd want to hear more about their use cases 19:38:07 <sslypushenko__> Just an extension for David's proposal 19:39:27 <zehicle> where do we stand w/ the API right now? 19:41:02 <sslypushenko__> I'm planning to land DB schema and connect DB to API prototype 19:41:48 <zehicle> ok 19:41:54 <sslypushenko__> It is real to get demo version to the next Monday 19:42:05 <zehicle> seems like we have API spec and code in process. no blocks on reviews 19:42:42 <zehicle> davidlenwell, said that he would be getting something up there too. 19:42:50 <sslypushenko__> It will be good if this patch will be landed as soon as possible https://review.openstack.org/#/c/144602/ 19:42:55 <zehicle> so, we should expect that API will be usable for f2f meeting 19:43:30 <sslypushenko__> <zehicle> Hope so) 19:44:32 <zehicle> #topic UUID 19:44:44 <zehicle> we need reviews of the UUID spec 19:45:03 <zehicle> I am, reluctantly, giving up on human readable UUIDs 19:45:28 <zehicle> so I'm not blocking. We will need to have code that does translations from LONGUUIDs into things that people can understand 19:45:52 <hogepodge> zehicle I'm not opposed to human readable uuids. I just don't know how to design it. 19:46:01 <rockyg> Do we have any qa reviews yet? 19:46:05 <zehicle> hogepodge, the QA & Infra teams are 19:46:14 <zehicle> they don't want to deal w/ collision issues 19:46:18 <sslypushenko__> <hogepodge> +1 19:46:27 <zehicle> so, it's a non-starter 19:46:42 <zehicle> rockyg, I think they would be more eager if we showed our team was aligned on it 19:47:13 * zehicle reserves the right to say I told you so when people complain about the capability JSON files 19:47:25 <catherine_d|1> Did we have an UUID spec written after Paris? 19:47:42 <zehicle> #link https://review.openstack.org/#/c/144329/2/specs/meta-data-and-uuid-for-tests.rst 19:48:48 <zehicle> please review and comment 19:49:04 <zehicle> other gerrit/review topics before I move on? 19:49:23 <hogepodge> I dropped the link to the qa team too and asked for some reviews so we can iterate on it. 19:49:40 <zehicle> hogepodge, thanks. we need their support to move this ahead. 19:50:01 <zehicle> #topic defcore capabilities files 19:50:31 <zehicle> before Julia left, she submitted updates to both Icehouse & Juno capabilities files 19:50:51 <zehicle> these are PRELIMINARY for discussion and I'd like to move them forward 19:51:04 <zehicle> but wanted catherine_d|1 to review and +1 19:51:26 <zehicle> hogepodge, and I were talking last week and we'd like to see if DefCore can work BOTH lists in parallel 19:51:26 <catherine_d|1> zehicle: I will review them ... 19:51:46 <zehicle> does not impact refstack directly, but wanted to keep you all in the loop 19:52:04 <rockyg> Zehicle. +1 19:52:24 <zehicle> we're making contingency plans for DefCore so vendors can self-certify without Refstack 19:52:38 <zehicle> ideally, the API progress makes that a non-issue 19:52:49 <hogepodge> #link https://review.openstack.org/#/c/136314/ 19:52:55 <hogepodge> #link https://review.openstack.org/#/c/131187/ 19:52:58 <zehicle> either way, we really need to have those files for community review in early March 19:53:44 <zehicle> questions or comments? 19:54:12 <zehicle> since leadership on DefCore may change after mid-January, I'm trying to get this setup in a good place. 19:54:32 * zehicle means that he may not get re-elected to the Board and would not be leading DefCore 19:54:46 <sarob> we need the json files out there so operators at the various companies that will be 19:55:05 <sarob> testing and reporting to mgmt on refstack impact to their distro 19:55:20 <zehicle> sarob, +1. hogepodge is going to help run the scoring/feedback sessions starting ASAP 19:55:25 <sarob> so getting a version out there is important 19:55:30 <sarob> even if flawed 19:55:50 <rockyg> So maybe we should also try to get a session at the ops meet up 19:55:51 <zehicle> according to Van L, there are places that need to be addressed 19:56:07 <sarob> rockyg most def 19:56:30 <sarob> im aiming to have it as one of the main topics for the pm midcycle 19:56:31 <zehicle> rockyg, +`1. hogepodge can you coordiate w/ Tom F? 19:56:37 <rockyg> Mid March 19:56:40 <hogepodge> zehicle yes. 19:56:46 <rockyg> I think I'll be there 19:56:55 <sarob> i will as well 19:56:57 <zehicle> sarob, that would be excellent 19:57:17 <zehicle> # other topics? 19:57:22 <zehicle> #topic other items 19:58:04 * zehicle reminds everyone that EVERY VOTE will count for the Bylaws changes. We need to get a quorum to pass anything! 19:58:31 <sarob> roger roger 19:59:02 <rockyg> When will vote mail come out 19:59:05 <zehicle> let's make sure to say on top of reviews leading into the F2F. 19:59:09 <zehicle> rockyg, 1/12 19:59:27 <rockyg> Thanks 19:59:44 <hogepodge> Oh, add that to the agenda. Time for us to vote during the meetup. 19:59:49 <zehicle> Gold member elections should be this week. can you confirm sarob ? 20:00:08 <sarob> umm 20:00:14 <sarob> i think next week 20:00:20 <zehicle> done 20:00:27 <zehicle> has to be before the general elections 20:00:59 <zehicle> "January 7: Gold Member Director Selector Election (1 day)" 20:01:05 * redrobot checks the time... ;) 20:01:17 <hogepodge> Operators meetup. Mar 9-10, 2015, Philadelphia, USA 20:01:19 <zehicle> #endmeetign 20:01:25 <zehicle> #endmeeting