00:06:06 #startmeeting CongressTeamMeeting 00:06:07 Meeting started Thu Sep 15 00:06:06 2016 UTC and is due to finish in 60 minutes. The chair is ekcs. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:06:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 00:06:13 The meeting name has been set to 'congressteammeeting' 00:06:49 Anyway here are the agenda items I have. Anything to add? 00:07:05 1. RC1 00:07:10 2. Outstanding bugs 00:07:13 3. Status 00:08:21 #chair ekcs_ 00:08:22 Current chairs: ekcs ekcs_ 00:09:41 Ok so RC1 is due this week. I think thinrichs will release to today or tomorrow. Any pressing changes that should be merged before then? 00:10:31 do you know deadline of release note? 00:11:30 we've not written releasenote about new architecture. 00:11:44 Hmm good question. I’m not totally sure, but it shouldn’t be a problem to change release notes in intermediate RC’s right? 00:12:04 hmm maybe we should get a draft of that in. 00:12:11 I think so. 00:12:57 ok I can look into that. #action ekcs will look into release notes for RC1 00:13:24 If possible I’d like to get this merged for RC1. #link https://review.openstack.org/#/c/369750/3 00:13:40 but it hasn’t been reviewed yet so maybe that won’t happen. 00:13:54 it’s important because congress fails to start under python3 without it. 00:14:16 ok, I'll check it. 00:14:36 Thanks! anyway I guess we can move on to the next topic. 00:14:50 #topic outstanding bugs 00:15:58 We have quite a few new bugs. Thanks for all the testing effort! Are there any notable bugs we’d like to discuss? 00:16:58 I added more info to the bug entry for #link https://bugs.launchpad.net/congress/+bug/1620868 00:17:00 Launchpad bug 1620868 in congress "Stack trace returned to end-user" [Medium,In progress] - Assigned to Aimee Ukasick (aimeeu) 00:17:57 but perhaps not as important as other bugs because it's not likely to happen that often 00:18:19 Got it. Thanks for doing so much work on it! 00:18:36 so I think if I can't fix it, you all can help after Newton has been released 00:18:38 I’m not following closely. But it seems we’re making progress, though ultimately still mysterious? 00:19:34 yes - the problem is with the entire policy_engines/agnostic.py namespace "error" reporting, both CLI and API pass traceback to the user 00:19:50 resorted to using pdb command line debugger 00:20:22 Right. Hopefully we can figure it out before final RC, but it’s not highest priority since it doesn’t affect functionality. 00:20:33 clarifying this bug: https://bugs.launchpad.net/congress/+bug/1602837 00:20:34 Launchpad bug 1602837 in congress "Policy UI (Horizon): Unable to get policies list (devstack)" [High,In progress] - Assigned to Aimee Ukasick (aimeeu) 00:20:46 things work fine as long as we avoid keystone v3 right? 00:21:11 It doesn't work in keystone v2 either 00:21:38 thinrichs verified that on 31 Aug 00:22:36 workaround is to use the API or CLI; I'll take a look at it again tomorrow 00:22:57 I think it's been broken for quite a while 00:23:08 ok got it. that’s more serious then. Seems like something we definitely want fixed before final RC. I’ll see if I can help too. 00:24:07 So to prioritize our time and resources, I created a new release milestone newton-rc2 on launchpad. 00:24:37 I’ll be tagging all the bugs that we can aim to fix before final RC with that tag. 00:24:49 ok 00:25:10 It’d be great if everyone else can take a look to and see if some things should be tagged that I missed or vice versa. 00:25:37 I've also found bugs that are related to each other 00:25:59 #action ekcs will tag bugs for rc2 00:26:06 so I'll add notes to the entries in launchpad 00:26:26 aimeeu: perfect. anyhting you want to talk about now? 00:26:42 nope - that's it for me 00:26:53 ok great. let’s move on to statuses then? 00:27:25 #topic status 00:27:38 masahito: would you like to start us off on status? 00:27:50 sure 00:28:07 I'm working on the bug. https://bugs.launchpad.net/congress/+bug/1622865 00:28:09 Launchpad bug 1622865 in congress "control bus sends 3 hb message in one loop" [Medium,New] - Assigned to Masahito Muroi (muroi-masahito) 00:28:29 I think I can push a patch for it today. 00:28:57 and also testing dse2 system especially about heartbeat stuff. 00:29:03 that's from my side. 00:30:21 That’s great! Looked like a mysterious bug to me, but I’m glad to hear you’ve got a handle on it! 00:30:52 aimeeu: do you have anything to add on status? 00:31:06 Nothing more for me 00:31:15 got it. 00:31:37 Here’s thinrich’s status: After 15ish devstack installs, I've not successfully set up multiple devstacks to talk to each other. Had some trouble with my laptop crashing, which didn't help. 00:32:10 I can feel his frustration 00:32:34 Yea… I guess it’s expected to be hard though. doing multiple devstacks talking for the first time. 00:32:49 And for me. 00:32:53 After a lot of wrangling with infra issues, I've finally gotten a basic haht tox test working. Will now work on adding more comprehensive testing, which should not be hard in comparison. #link https://review.openstack.org/#/c/358927/ 00:33:32 Once I have that under control, I can give a lot more time to other testing and fixing bugs. 00:34:05 great! 00:34:28 I’ve already mentioned the python3 startup bug I’d like to get merged before RC1. And there is another bug fix in review. That’s all from me. 00:35:00 #topic open discussion 00:35:13 Anything else we should talk about? 00:36:42 masahito: since we have some time, I figure I’d bring this up and see if yo uhave any thoughts. 00:37:07 #link http://logs.openstack.org/32/369132/1/check/gate-congress-dsvm-api/d97ddb4/logs/stackviz/#/stdin 00:37:40 I’ve been seeing this doctor failure in tempest every once in a while. not frequent and always goes away on recheck. any ideas? 00:39:14 hmm... I didn't see it before. 00:39:47 Initializing event table looks fail. 00:40:10 I try to reproduce it in my local. 00:40:36 ok great! 00:41:09 btw, I didn't know the UI showing us gate log 00:41:38 Oh yea it’s great. I didn’t realize it was there until the top level test results went missing for a while. 00:41:44 so I was forced to find an alternative. 00:42:22 The other things I’d mention real quick is that on this patch: https://review.openstack.org/#/c/369750/3 00:42:43 The thing I’m not totally sure about is whether we need to support an eventlet server being stopped and started. 00:43:06 You guys know if there is actual case for that? 00:43:43 By delaying the node creation to the harness, the eventlet_server module no longer has access to the DseNode to stop the node in case the eventlet server is stopped. 00:44:26 But I don’t think it matters. Just throwing it out so if someone knows otherwise they can tell me or comment on the patch. 00:44:29 Thanks! 00:45:36 If we implement graceful-shutdown, we could need the feature. 00:46:24 Ah good to know. What does graceful-shutdown do? kill an eventlet server without exting the process? 00:47:13 And is there a case when the Congress API server would be called to gracefully shutdown without exiting the process? 00:47:25 I'm not sure what eventlet server is in charge of now. 00:47:49 ah.. eventlet server only works for API now? 00:48:03 In that case, it sounds no usecase. 00:48:37 eventlet server still does non-API stuff. But it’s only in the API case that I can’t stop the node. 00:49:12 in eventlet server there is regular Server and APIServer classes. 00:49:53 I guess we can discuss over gerrit. Thanks for the note on graceful shutdown though! 00:50:20 Anyway I don’t have anything else to discuss. Lots of testing and bug fixing still, but it’s exciting that we’re very close to newton release and finally releasing the distributed architecture that’s over a year of work. 00:50:41 anything else or should we end early? 00:50:53 I don't have anything else to add 00:51:40 ok then let’s end it. See you all next week! 00:51:48 #endmeeting