00:00:25 #startmeeting CongressTeamMeeting 00:00:25 Meeting started Thu Dec 10 00:00:25 2015 UTC and is due to finish in 60 minutes. The chair is thinrichs. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:00:26 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 00:00:29 The meeting name has been set to 'congressteammeeting' 00:02:03 Who is here today? 00:06:25 sorry, late. 00:06:52 masahito: hi 00:08:27 Anyone see this now? 00:13:26 The connection for the freenode isn't good. 00:13:48 test 00:14:30 thinrichs1: hi 00:14:48 ramineni: hi 00:14:55 masahito: you there? 00:15:28 #chair thinrichs1 00:15:32 thinrichs1: i'm here 00:15:52 another test 00:16:11 masahito_: I don't see you. Did you get this? 00:17:10 thinrichs1: yes 00:17:11 masahito_: do you see thinrichs1 second message saying he doesn’t see you? 00:17:30 ekcs: I saw both message 00:17:36 all: something seems wrong with IRC today. 00:17:45 I can't see masahito_ messages but he can see mine. 00:18:02 So if someone else could resend his messages so I can see them that'd be helpful. 00:18:22 Sorry for the late start everyone. 00:18:45 Here's my agenda for today 00:18:53 1. distributed arch 00:18:56 2. python3 00:19:07 3. bryan_att and install issues 00:19:10 3. mid-cycle meeting 00:19:19 5. misc 00:19:22 Anything else? 00:19:47 Let me know as we go then. 00:19:55 #topic distributed arch 00:20:02 pballand: Want to give a status update? 00:22:00 Am I disconnected again? 00:22:10 thinrichs1: no :) 00:22:30 thinrichs1: we can see the messages 00:22:39 ramineni: phew! 00:22:47 Let's come back to this topic then. 00:22:50 thinrichs1: me too. 00:22:50 #topic python3 00:23:21 ekcs: status update? 00:23:33 ekcs , also doesnt seem to be there in room? 00:23:38 I’m here. 00:24:10 Yeah—he seems to have dropped out too. 00:24:35 Maybe it's worth canceling this week and going over stuff via email. 00:24:38 ugh. I can see your messages 00:24:50 but you can’t seem to see mine. 00:25:00 ramineni: while you're here, want to give an update on the tempest tests (and devstack)? 00:25:02 I can see the messages from ekcs 00:25:12 thinrichs1: sure 00:25:34 thinrichs1: almost got the tempest tests running on major components :) 00:26:11 thinrichs1: the problem seems to be we are not waiting until data gets converged , exiting before only , so they are failing intermittently 00:26:23 ramineni: Nice! 00:26:51 thinrichs1: but we might need to make changes if tempest does any changes wrt to client 00:26:52 ramineni: I thought there was code in place to wait until convergence 00:27:12 ramineni: but the tests were definitely flakey 00:27:38 ramineni: were there some places where we were waiting correctly, and there were just places we missed? 00:27:47 thinrichs1: yes, it was there , but we are handling exception in between , so exiting instead of calling again 00:28:45 thinrichs:yes 00:35:04 ping? 00:35:09 pong 00:35:18 pballand: thinrichs1:hi 00:35:29 Should we try this again? 00:35:30 hmm - maybe things are back? 00:35:37 hi 00:35:39 Seems okay 00:35:46 ya 00:36:02 Let's give it a try. 00:36:12 thinrichs1: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 00:36:17 hello 00:36:32 Seems like it's remembering that the meeting is still going. 00:36:34 haha - going to be an interesting meeting log 00:36:41 pballand: :) 00:36:58 Let's see how much we can get done in 25 min 00:37:03 #topic distributed arch 00:37:06 pballand: status update? 00:37:32 sure - pushed the bare version of the dsenode - and tweaked for some minor pep errors 00:37:55 now working on the feature-rich version, since it seems masahito and ekcs have leads on the fake driver 00:39:01 once pep8 passes, I think the commit can/should go in as we discussed last week 00:39:27 and other future feature commits can sit in code review pending oslo messaging changes, or we can push them as well, depending 00:39:34 that’s it for me 00:39:55 Sounds good. Thanks. Looking forward to getting that into the tree. 00:40:23 ekcs, masahito: quick update on how you plan to fix the fake driver? 00:40:53 I found the cause of the oslo bug, so I'm writing the patch now and can push it in this week. 00:41:12 masahito: you rock! 00:41:14 masahito: great! 00:41:53 and I noticed that when we use fake driver we need little tweaks in our test. but it's not a lot :) 00:42:25 masahito: I'm looking forward to the details. 00:42:31 moving on… 00:42:34 #topic python3 00:42:46 ekcs: I saw you've made good progress. Want to report? 00:42:56 Patches were merged. Py34 should be passing. I think we should turn on py34 voting asap to prevent regression. 00:42:57 I'm also happy to help if anyone runs into issues passing py34 on their patches. 00:43:03 Next step: start using congress on python3. We'll probably encounter a few bugs over time. 00:43:50 ekcs: fantastic! That one has been on our list a looong time. Glad to have it done. 00:44:04 I'll try to submit a patch to make it voting. 00:44:19 Everyone else: until that merges, make sure python34 tests pass. 00:44:27 #action thinrichs will make python34 job voting 00:44:36 #topic mid-cycle meeting 00:44:45 We finalized dates. Here's the email. 00:44:52 #link http://lists.openstack.org/pipermail/openstack-dev/2015-December/081606.html 00:45:10 Jan 26-28 in the Bay Area 00:45:16 You can see a tentative agenda in there too. 00:45:21 Please respond with comments/questions. 00:45:25 On the mailing list. 00:45:44 #topic Tempest tests 00:45:54 ramineni: status update on the tempest tests? 00:46:18 thinrichs1: enabled most of them , i think murano one is pending 00:46:43 Is it safe to start asking people to write new tempest tests again? 00:46:55 thinrichs1: we might need to make changes over time , if tempest changes , as they are in huge migration process 00:47:09 thinrichs1: yes , that would be good 00:47:21 ramineni: good to know. So maybe only add tests that we think are pretty important, until they get their migration done 00:47:30 thinrichs1: I will keep an eye on tempest tests 00:47:51 thinrichs1: yes 00:48:01 They're gating again, so we'll all keep an eye on them. :) Thanks for volunteering 00:48:15 ramineni: you're definitely our tempest expert at this point 00:48:28 thinrichs1: :) 00:48:31 bryan_att: are you there? 00:48:40 #topic Installation 00:49:05 bryan_att posted to the openstack mailing list about his experiences getting Congress installed on a non-devstack openstack installation 00:49:14 #link http://lists.openstack.org/pipermail/openstack/2015-December/014760.html 00:49:25 He had some suggestions for us and I promised him time in the meeting. 00:49:46 But it seems he might not be around 00:50:00 We can try again next week. 00:50:23 #topic Project navigator 00:50:37 Just this week Congress was added to the OpenStack project navigator 00:50:38 #link 1. https://www.openstack.org/software/releases/liberty/components/congress 00:51:08 Here's the navigator link (the one above is just for Congress) 00:51:09 https://www.openstack.org/software/project-navigator 00:51:43 Here's another way to look at it. 00:51:46 #link https://www.openstack.org/software/ 00:52:17 One of the things we're not doing that many other projects are doing is including our APIs in the OpenStack documentation repo… 00:52:52 #link http://developer.openstack.org/api-ref.html 00:53:15 To include our APIs in that documentation, we need to encode them in XML and contribute them to the appropriate repo. 00:53:34 #link https://git.openstack.org/cgit/openstack/api-site/ 00:53:53 The rest of our docs are hosted on readthedocs... 00:54:00 https://congress.readthedocs.org/en/latest/ 00:54:06 That includes our API. 00:54:32 So one thing we'll need to decide is whether we want to move our API to the openstack XML repo or not. 00:55:07 I'd say we want it in 1 place (not 2), which means either we cut it out of the RST format it's currently in and rewrite it in XML, or leave everything as is. 00:55:43 That's the last thing for me. So in the last 5 minutes, we'll open it for discussion (and feel free to chime in about the docs). 00:55:47 #topic open discussion 00:56:17 Do we need have api reference in api-site as OpenStack official project? 00:56:56 masahito: no formal requirement that I know of. 00:57:00 If so, IMO we should move the RST doc to XML. 00:57:10 thinrichs1: but many other projects do use rst right as of now? 00:57:11 If not so, I prefer RST 00:57:55 I've written some reference for api-sit, but the structure of the XML is worse to maintain. 00:57:59 ramineni: most of the others use XML, though I think that's b/c the documentation team wrote theirs 00:58:27 thinrichs1: does it affect this “indicator”? “Is there an install guide for this project guide (at docs.openstack.org)?” on #link https://www.openstack.org/software/releases/liberty/components/congress 00:58:30 murano (another new project) is using rst as well, though whether they plan to use xml in the future I don't know. 00:58:48 thinrichs1: ok 00:58:49 ekcs: I think we should add an install guide, independent of the API. 00:59:15 ekcs: since the install guide does seem to be visible to our maturity 00:59:39 thinrichs1: got it. 00:59:53 I don't know what they're going to do in the long-run. 00:59:58 thinrichs1: i have one last thing https://review.openstack.org/#/c/250598/ 01:00:13 I'd rather not port to XML, but if that's what everyone will do in the long run, we probably should. 01:00:14 thinrichs1: we can discuss in channel , as time is up 01:00:20 ramineni: sounds good. 01:00:33 Thanks all! Sorry about the accelerated meeting and all the hiccups at the start. 01:00:35 #endmeeting