17:00:50 <thinrichs> #startmeeting CongressTeamMeeting 17:00:50 <openstack> Meeting started Tue Jul 14 17:00:50 2015 UTC and is due to finish in 60 minutes. The chair is thinrichs. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:51 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:54 <openstack> The meeting name has been set to 'congressteammeeting' 17:01:39 <thinrichs> I see we have some folks from Asia today—thanks for making the effort! 17:02:02 <thinrichs> What time is it for all of you? 17:02:14 <RuiChen> hi thinrichs 17:02:25 <RuiChen> 1:00 AM 17:02:37 <masahito> 2 am 17:02:48 <alexsyip> Thanks for joining at such a late hour! 17:03:00 <RuiChen> but it's ok for me 17:03:14 <thinrichs> Since it's so late let's get to the stuff that matters to you two first, so you can drop off if you like. 17:03:45 <thinrichs> Last week we talked briefly about possibly moving the meeting time to 5-6p Pacific (which is in the middle of the day in Asia). 17:04:06 <thinrichs> alexsyip: does that (ever) work for you? I know it's later than you usually work. 17:04:18 <alexsyip> That is ok with me. 17:04:52 <thinrichs> RuiChen, masahito: if it were 1 hour earlier would that work, or is that too early? 17:05:16 <thinrichs> So 4-5p Pacific. 17:05:24 <masahito> thinrichs: It works for me. 17:05:46 <thinrichs> Which I think is 11p UTC-midnight UTC 17:05:46 <RuiChen> works for me too 17:06:09 <thinrichs> alexsyip: any idea if 4-5p would work for pballand? 17:06:23 <alexsyip> No, I don’t know what his schedule is like. 17:06:34 <alexsyip> He works from home on some days, so it could be fine for him. 17:07:03 <alexsyip> I say go for it. 17:07:29 <thinrichs> Seems like a reasonable adjustment, to save some people from staying up late to join. 17:07:50 <thinrichs> pballand is out of town for a couple more weeks I think. 17:08:07 <thinrichs> So let's move it to 4-5p Pacific at least until he gets back. 17:08:27 <thinrichs> I'll make sure we can find a room that's free and send out email. 17:08:34 <thinrichs> Oops. forgot to start the topic. 17:08:37 <thinrichs> #topic Meeting time 17:09:03 <thinrichs> #action thinrichs will find a room for meeting 4-5p Pacific on Tuesdays and will email out details to the ML. 17:09:53 <RuiChen> thanks for adjusting the meeting time :) 17:10:02 <thinrichs> RuiChen: happy to. 17:10:14 <thinrichs> The rest of what I have on the agenda is straightforward. 17:10:33 <masahito> thinrichs: thank you for changing the time ;) 17:10:38 <thinrichs> Status update on the mid-cycle sprint 17:10:46 <thinrichs> masahito: you're welcome. 17:10:56 <thinrichs> 1. Status update on mid-cycle sprint 17:11:02 <thinrichs> 2. UI demo discussion 17:11:04 <thinrichs> 3. Milestones 17:11:06 <thinrichs> 4. Python 3 17:11:10 <thinrichs> 5. Talks for summit 17:11:34 <thinrichs> Anything else? 17:11:41 <thinrichs> 6. status updates (if there's time) 17:11:54 <RuiChen> i have a bug https://bugs.launchpad.net/congress/+bug/1473924 17:11:54 <openstack> Launchpad bug 1473924 in congress "datasource-row-list API raise 400 exception when thread switching occur in update_from_datasource()" [Undecided,New] - Assigned to Rui Chen (kiwik-chenrui) 17:12:09 <RuiChen> and i want to refactro datasource driver 17:12:36 <thinrichs> That's a good bug that has shown up several times. 17:12:50 <thinrichs> We definitely want to make sure datasource drivers update self.state atomically. 17:12:58 <thinrichs> Some refactoring might help with that. 17:13:02 <RuiChen> and abstract common logic like updating self.state to super class. 17:13:24 <thinrichs> RuiChen: sounds reasonable. Maybe have update_state *return* the new dict and have the base class overload it. 17:13:49 <RuiChen> yeah, i need to draft a spec or bp? 17:14:01 <alexsyip> You can overwrite the existing content of self.state. 17:14:18 <alexsyip> They only drawback is that you’ll be holding two copies of all the state while you’re building the new copy. 17:14:26 <alexsyip> I’m not sure that’s avoidable though. 17:14:33 <thinrichs> RuiChen: don't worry about a BP/spec for this. 17:14:56 <thinrichs> RuiChen: I'd do the refactoring and implement 1 datasource driver and push to review. 17:15:08 <thinrichs> Then we can all give you comments before you touch all the 15 datasources. 17:15:19 <thinrichs> alexsyip: good points there. 17:15:41 <RuiChen> yeah, good suggestion, thanks 17:16:36 <thinrichs> RuiChen: one other thought.. 17:16:57 <thinrichs> If we're pulling a ton of data, 17:17:08 <thinrichs> and we knew that only some of it changed, 17:17:22 <thinrichs> it might be useful to only retranslate the portion that changed. 17:17:59 <thinrichs> I don't think this is an issue right now. 17:18:05 <thinrichs> But keep that in the back of your mind. 17:18:32 <thinrichs> I don't think there's a datasource driver that would do that right now. 17:18:56 <thinrichs> Anything else to add to the agenda? 17:19:18 <RuiChen> maybe we should add congess spec to specs.openstack.org? 17:20:47 <thinrichs> RuiChen: I don't know how to do that off the top of my head. 17:20:56 <thinrichs> And I don't have the cycles to look into it. 17:21:10 <thinrichs> We also aren't big enough to rely on the spec process the way other projects do. 17:21:43 <thinrichs> But if you want to look into it that'd be great. 17:22:01 <RuiChen> ok, i can try 17:22:42 <thinrichs> RuiChen: I'd say don't spend too long. Do a little googling; look at project-config; if nothing turns up, let me know. 17:23:08 <thinrichs> Let's move on to the mid-cycle sprint update. 17:23:11 <thinrichs> #topic mid-cycle sprint 17:23:22 <RuiChen> thinrichs, get it 17:23:32 <thinrichs> Here's the wiki with info 17:23:34 <thinrichs> #link https://wiki.openstack.org/wiki/Sprints#Liberty_sprints 17:23:45 <thinrichs> alexsyip: any progress on getting a room? 17:23:53 <alexsyip> Yup, we got a room at vmware. 17:24:13 <alexsyip> I will also get food when we have a better headcount. 17:24:16 <thinrichs> alexsyip: Want to add the room details to the wiki? 17:24:38 <thinrichs> alexsyip: Maybe also some instructions about where to park and where to sign in. 17:25:19 <thinrichs> I suppose we should be those details on the eventbrite page as well. Or maybe just the eventbrite page so there's only one place. 17:26:41 <thinrichs> Anything else we should discuss about the sprint? 17:27:12 <alexsyip> I don’t think I can edit eventbrite. 17:27:18 <alexsyip> The room is “BASH" 17:27:23 <alexsyip> in hilltop - A 17:27:53 <thinrichs> #action thinrichs will add room details for the midcycle summit to eventbrite 17:28:09 <thinrichs> I'll take care of eventbrite 17:29:32 <thinrichs> Next topic: milestones 17:29:35 <thinrichs> #topic Milestones 17:29:59 <thinrichs> We still want to have people sign up for blueprints and choose the milestone by which they can get those blueprints implemented. 17:30:05 <thinrichs> #link https://blueprints.launchpad.net/congress 17:30:48 <thinrichs> masahito signed up for 2 and chose a milestone. Thanks! 17:31:13 <masahito> Oh. I forgot sending you a milestone though. 17:31:17 <thinrichs> alexsyip: any thought about the milestone for basic-ha? 17:31:22 <veena> thinrichs: can anybody sign for any of those blueprints? 17:31:28 <thinrichs> veena: yep 17:31:35 <veena> thinrichs: okay 17:32:09 <thinrichs> veena: I may have a good task for you, that would help us out a bunch and would be a number of small commits. 17:32:20 <thinrichs> veena: Python3 support 17:32:27 <thinrichs> I'll come back to that in a minute. 17:32:47 <thinrichs> masahito: I have you down for 1 milestone and figured you were waiting to get that done until choosing the next milestone. 17:32:48 <veena> thinrichs: Sure 17:33:25 <alexsyip> I estimate there’s about 2 weeks of work for basic HA. 17:33:55 <alexsyip> Having tempest tests running would make things better. 17:34:20 <thinrichs> alexsyip: I'm finally making progress on the tempest tests 17:34:30 <alexsyip> oh good. 17:34:43 <alexsyip> I reinstalled devstack yesterday and had some test failures. 17:34:54 <alexsyip> in test_neutronv2 17:35:04 <thinrichs> I've spent a *bunch* of time digging through tempest. 17:35:30 <thinrichs> alexsyip: shall I put liberty-2 down for the basic-ha then? 17:36:03 <alexsyip> yeah ok 17:36:23 <thinrichs> Done 17:36:54 <thinrichs> alexsyip: I know someone (masahito maybe) fixed a bunch of tempest failures. 17:37:03 <thinrichs> Maybe there are just some new ones. 17:37:12 <thinrichs> You were talking about neutronv2 tempest tests, right? 17:37:27 <thinrichs> Not the unit tests. 17:37:31 <alexsyip> yes 17:37:48 <masahito> I'm not familier with tempest test. 17:38:01 <thinrichs> masahito: sorry—must have been someone else. 17:38:19 <thinrichs> I'll keep plugging away trying to get tempest tests working in a virtualenv then. 17:38:29 <thinrichs> We really do want those tests gating our changes. 17:38:44 <thinrichs> Time to move on. 17:38:46 <thinrichs> #topic Python3 17:39:02 <thinrichs> All of OpenStack is moving off of Python2, since Python2 is at end-of-life. 17:39:16 <thinrichs> The Python community has said they're not supporting it any longer. 17:39:40 <thinrichs> In fact we're behind on migrating over, compared to other projects. 17:39:50 <thinrichs> So if everyone can pitch in and try to get fixes in, that'd be good. 17:40:06 <veena> thinrichs: I can take that up 17:40:11 <thinrichs> Has anyone actually done this kind of porting before? 17:40:17 <thinrichs> I don't know the best way to go about it. 17:40:28 <thinrichs> There could be a bunch of work, and there could be just a little work. 17:40:34 <thinrichs> veena: thanks. 17:40:35 <veena> thinrichs: there's a tool 2to3 which might be of some help 17:41:03 <veena> thinrichs: Thanks 17:41:13 <thinrichs> I don't know if we're going full to Python3, or trying to stay backwards compatible with 2. 17:41:31 <thinrichs> I imagine we want to stay backwards compatible, unless there's a good reason not to. 17:41:44 <thinrichs> Which would mean relying on the six package. 17:41:51 <veena> thinrichs: Okay 17:42:40 <thinrichs> We have a python34 job running in the Gate. 17:42:40 <thinrichs> http://logs.openstack.org/10/199910/3/check/gate-congress-python34/cedccfe/console.html 17:43:14 <thinrichs> I haven't looked into what it does. I imagine it's just running all the tests with python3 instead of python2. 17:44:45 <thinrichs> An OpenStack guide on moving from python2 to python3: 17:44:46 <thinrichs> #link https://wiki.openstack.org/wiki/Python3 17:45:01 <thinrichs> So let's all try to help with this over the next few weeks. 17:45:15 <RuiChen> I think backwards compatible is important, many openstack customers are not ready for python3 AFAIK, 17:45:41 <RuiChen> yeah, I can help to review these patch. 17:45:53 <thinrichs> Interestingly, that OS page says we're dropping Python3.3 support in LIberty. 17:46:15 <thinrichs> I'm inclined to believe we should be backwards compatible. 17:47:01 <thinrichs> So let's move forward supporting Python2 and Python3. 17:47:17 <thinrichs> Questions, issues? 17:47:58 <veena> thinrichs: Okay. No questions as of now. 17:48:24 <thinrichs> veena: I think going through that wiki page is a good first cut at getting to Python3 compatibility. 17:48:39 <veena> thinrichs: Okay 17:48:42 <thinrichs> Next topic. 17:48:51 <thinrichs> #topic Summit talks 17:49:03 <thinrichs> Just a reminder that the talks for the summit are due tomorrow. 17:49:04 <thinrichs> #link • https://www.openstack.org/summit/tokyo-2015/call-for-speakers/ 17:49:32 <thinrichs> I'm happy to give feedback or edit abstracts. Just let me know. 17:51:03 <RuiChen> we will plan congress hands-on-lab in tokyo? 17:51:57 <thinrichs> RuiChen: yep 17:52:09 <thinrichs> alexsyip submitted a proposal 17:52:10 <RuiChen> great, I like it 17:52:15 <masahito> thinrichs: I've already submitted my abstract. Can you read it if I add you as a speaker? 17:52:30 <thinrichs> masahito: I imagine I can read it if you add me. 17:52:54 <thinrichs> RuiChen (and everyone else): let us know if you're interested in preparing/delivering the HOL. 17:52:59 <masahito> thinrichs: OK. I'll add you just after the meeting 17:53:26 <masahito> I can help the HOL as Japanese :) 17:53:57 <thinrichs> masahito: :) 17:54:12 <thinrichs> Just a few minutes left. 17:54:22 <thinrichs> Open discussio 17:54:28 <thinrichs> #topic open discussion 17:55:20 <masahito> I want to add a new 'service' table in nova datasource. Do I need to write bp? 17:55:34 <thinrichs> masahito: no blueprint for that. 17:55:48 <masahito> thinrichs: ok 17:55:49 <thinrichs> masahito: you can file a bug if you want 17:55:56 <thinrichs> masahito: but that's not required either 17:57:27 <RuiChen> everybody can take a look this spec https://review.openstack.org/#/c/198279/ ? 17:57:46 <RuiChen> api version list 17:58:59 <thinrichs> Let's everyone take a look at that, especially if you know how other projects return API versions. 17:59:06 <thinrichs> We're out of time for today. 17:59:16 <thinrichs> Thanks everyone! 17:59:23 <RuiChen> thank you 17:59:45 <veena> Thanks 17:59:53 <masahito> thank you 18:00:13 <thinrichs> #endmeeting