14:00:30 #startmeeting oslo 14:00:31 Meeting started Mon Oct 9 14:00:30 2017 UTC and is due to finish in 60 minutes. The chair is gcb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:34 The meeting name has been set to 'oslo' 14:00:47 courtesy ping for amotoki, amrith, bknudson, bnemec, crushil, dansmith, dhellmann 14:00:48 courtesy ping for dims, dougwig, e0ne, electrocucaracha, flaper87, garyk, gcb 14:00:48 courtesy ping for GheRivero, haypo, jd__, jecarey, johnsom, jungleboyj, kgiusti 14:00:48 courtesy ping for kragniz, lhx_, lifeless, lxsli, Nakato, ozamiatin, rbradfor 14:00:48 courtesy ping for redrobot, rpodolyaka, sergmelikyan, sileht, spamaps, sreshetnyak, sreshetnyak 14:00:49 courtesy ping for stevemar, therve, thinrichs, toabctl, viktors, zhiyan, zzzeek,zxy 14:00:57 hi 14:01:02 o/ 14:01:09 \o 14:01:19 o/ 14:01:26 o/ 14:02:17 welcome everyone, I'm just back from long vacation :-) 14:02:46 #topic Red flags for/from liaisons 14:03:01 http://status.openstack.org/openstack-health/#/?groupKey=build_name&resolutionKey=hour&searchProject=-with-oslo 14:03:01 @! 14:03:01 <_pewp_> jungleboyj (ه’́⌣’̀ه )／ 14:03:35 jungleboyj: who is _pewp_ ? 14:03:49 o/ 14:04:02 :-) It is our table flipping bot. 14:04:42 o/ 14:04:43 lol 14:05:07 @!h 14:05:07 <_pewp_> jungleboyj (/ .□.) ︵╰(゜Д゜)╯︵ /(.□. ) 14:05:43 periodic jobs for trove failed for a long time, https://review.openstack.org/#/c/507087/ is trying to fix trove gate issue 14:07:11 #topic Releases 14:07:43 #link https://review.openstack.org/510567 14:08:49 BTW recently, we're preping for the stable/newton EOL http://lists.openstack.org/pipermail/openstack-dev/2017-October/123088.html 14:09:09 we have one release for stable/newton now https://review.openstack.org/510462 14:10:20 #topic Zuul v3 migration 14:11:03 it seems the migration raised some issues for some oslo libraries like oslo.messaging 14:11:10 +1 14:11:41 I'm currently working on it: https://review.openstack.org/#/c/510197/ 14:12:50 kgiusti: thanks, I just began to learn zuul v3 from https://docs.openstack.org/infra/manual/zuulv3.html and https://etherpad.openstack.org/p/zuulv3-migration-faq 14:13:11 IMHO a big improvement 14:13:38 But we've (messaging) have slapped in a few hacks over the years that need to be cleaned up 14:13:51 * kgiusti gets his bucket and mop.... 14:14:53 kgiusti, yeah, we need more work if we have customize jobs 14:17:26 #topic Stuck Reviews 14:19:11 I'm trying to retire pylockfile in https://review.openstack.org/#/c/508652/, but need we end the gate jobs before merging it 14:19:48 background : http://lists.openstack.org/pipermail/openstack-dev/2017-October/122978.html 14:20:32 any other stuck reviews ? 14:23:02 okay, let's move on 14:23:20 #topic Open discussion 14:23:46 Looks like I cannot make sidney :( 14:24:06 gcb: so I can't do live from oslo... 14:24:52 kgiusti: I'm sorry to hear that, but it's okay, I will help cover the 20 minutes sessions 14:25:10 gcb: thank you 14:25:23 gcb, others : https://review.openstack.org/#/c/507098/ is ready (http/https check rules as stevedore extensions) 14:26:07 kgiusti: actually, My visa application was rejected, will try to apply it later :( 14:26:10 dims: ack 14:26:37 rejected? :O 14:28:39 ouch :( 14:31:39 yes, it doesn't matter , will try it later 14:32:14 o/ 14:32:25 dims: I think we'd better have a release note for the change, others LGTM 14:33:00 hi dhellmann 14:33:02 gcb : ack thanks 14:35:14 dims : +2 with the anticipation of documentation to come later 14:35:37 dhellmann : yep, will file one today 14:35:50 ++ 14:37:48 dhellmann: is it okay to release stable/newton now like https://review.openstack.org/510462? 14:38:27 gcb : I think so, but I'm not sure about the deadlines so we should coordinate with tonyb 14:40:19 dhellmann: okay, tonyb is not available on IRC now, will ping him later 14:41:17 any other topics ? 14:41:47 nothing from me 14:41:57 Sorry, I was in a meeting earlier, but it would be nice to get https://review.openstack.org/509919 in and released. 14:42:07 It's breaking a couple of other projects right now. 14:42:49 bnemec : +2A 14:42:58 dims: Thanks 14:43:09 bnemec : the test doesn't appear to verify that the 2 contexts are the same 14:43:42 dhellmann: I was just trying to confirm that all the keys from to_dict are valid parameters to the constructor. 14:43:47 bnemec: feel free to update https://review.openstack.org/#/c/510567/ when the patch was merged 14:43:57 bnemec : yeah, that's a good test, too 14:44:16 I can look at that as a followup if we don't already have a round-trip test. 14:44:43 bnemec, dims : the other fix would have been to use the existing from_dict() method, which is the supported way of doing this 14:44:56 and really, that's what we should do 14:45:13 dhellmann: I tried that too and it didn't work because these projects have custom attributes on their context classes. 14:45:25 then they need custom from_dict() methods 14:45:27 The built-in from_dict only handles the ones in the base class. 14:45:38 removing workflow ... 14:45:55 They have custom from_dict methods, but they were relying on the 1:1 key:parameter behavior we used to have. 14:46:04 ah 14:46:10 was that ever formally part of the API? 14:46:19 I thought the whole purpose of from_dict() was to reverse to_dict() 14:47:33 I guess I'm not clear on what API we're actually trying to support here. If we want __init__ to work with the output of to_dict() we don't need a from_dict() 14:48:00 I guess that's true. 14:48:12 Oh, user_identity was a special case too. 14:48:26 They added a dummy param on the sub-class constructor to handle that one. 14:48:45 we can have the base class from_dict() strip out values like that 14:48:48 https://github.com/openstack/designate/blob/46de766e513cfb97cbfc50b001734e02711517e4/designate/context.py#L42 14:49:21 I guess right now it's mostly merging values together, so maybe I'm wrong on what it's intended to be used for 14:50:09 So initially I was going to just remove the override: https://review.openstack.org/#/c/509685/ 14:50:11 does to_dict() return all of the values needed to recreate the object? 14:50:35 But that broke other things, and then I found out there were multiple projects affected and thought it would be nicer to just fix it in oslo.context. 14:50:38 sheesh, yeah, that needs to be removed 14:51:03 that looks like copypasta from somewhere else 14:52:21 I don't want to hold up the meeting, maybe we can move this back to openstack-oslo? 14:53:00 I have another meeting starting in a couple of minutes, but I think we want to be explicit about the decision about what API we're going to support here. The pragmatic approach may be to take bnemec's patch, but that implies the constructor can never be cleaned up so we need to be sure that's what we want. 14:53:15 and yeah, we should discuss in the regular channel or via email 14:53:28 okay, let's move back to openstack-oslo 14:53:28 dhellmann: Yeah, it's a good discussion. I've obviously gone back and forth on it already. :-) 14:53:46 I'll start a -dev thread so we can loop in the people from the other projects. 14:53:52 sounds good, thanks bnemec 14:54:12 thanks everyone, let's end the meeting :-) 14:54:15 #endmeeting