16:00:18 #startmeeting oslo 16:00:20 Meeting started Mon Jan 12 16:00:18 2015 UTC and is due to finish in 60 minutes. The chair is dhellmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:22 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:24 The meeting name has been set to 'oslo' 16:00:26 o/ 16:00:27 o/ 16:00:28 o/ 16:00:29 our agenda, as usual: 16:00:30 #link https://wiki.openstack.org/wiki/Meetings/Oslo 16:00:32 hi 16:00:42 courtesy ping for jd__, dims, bnemec, flaper87, harlowja, viktors, rpodolyaka1, zzzeek, sileht, kgiusti 16:00:46 courtesy ping for redrobot, jungleboyj, zhiyan, therve, amotoki, GheRivero, bknudson, ihrachyshka, jogo, dougwig, sreshetnyak, amrith 16:00:53 o/ 16:01:02 o/ again. :-) 16:01:17 heh, I have those lined up to paste so you don't have to answer twice :-) 16:01:18 o/ 16:01:24 dhellmann: Nice. 16:01:30 o/ 16:02:02 we have a lot to cover today, so I'm going to dive right in 16:02:03 #topic Review action items from previous meeting 16:02:13 #info dhellmann email dev list reminding oslo contributors to be ready to discuss schedule at the next meeting DONE 16:02:17 *foreshadowing* 16:02:26 #info dhellmann talk to gordc about releasing oslo.middleware DONE 16:02:29 #info dhellmann release oslo.concurrency DONE 16:02:55 #topic Red flags for/from liaisons 16:03:14 does anyone have any blocking issues? 16:03:30 I think we've at least identified all of the fixes for things broken by the namespace package changes, right? 16:03:34 things are looking good in keystone-land 16:03:50 o/ 16:03:58 * flaper87 loves dhellmann's courtesy pings 16:04:03 dhellmann: I have this issue: https://review.openstack.org/#/c/145642/ 16:04:06 I have some complaints about config generator that I haven't brought up yet. 16:04:08 cinder, heat, and ceilometer had some breakages 16:04:25 I am about to submit a patch to set olso.i18n to be <=1.1.0 for Juno. 16:04:37 dhellmann: That makes sense? 16:04:59 I don't think you can cap versions in juno since then can't upgrade 16:05:11 jungleboyj: yeah, that may be the best way to address stable branch issues. We'll see if it works. :-) 16:05:25 bknudson: ??? 16:05:27 bknudson: that used to be true, but I don't think it is any more. 16:05:46 must have changed the grenade test then 16:06:13 we keep trying to make it work, but I honestly don't know the current state 16:06:28 I will push it out there and see if anyone has concerns and if it blows up in grenade. 16:06:31 grenade does apparently update dependencies 16:07:08 jungleboyj: yeah, let's see what happens 16:07:21 * jungleboyj makes it so 16:07:22 how blocked is cinder development based on this? are all patches failing now? 16:07:42 All the stable/juno ones. We got master fixed up already. 16:07:55 ok, as long as the blockage is limited 16:08:25 how about any of the other projects? heat and ceilometer also had some issues, are those resolved? 16:08:39 therve, and jd__ ? 16:08:59 these are all related to using internal interfaces that change? 16:09:06 that does not ring a bell 16:09:10 yeah 16:09:19 jd__: ok, I thought I saw something about unit tests being broken 16:09:41 maybe it was already resolved, but as long as you're not blocked 16:10:04 if there's nothing else, we can move on to our kilo-2 status check 16:10:14 in keystone we had an issue with using timeutils but fixed that quickly. 16:10:41 bknudson: did you just change the mock? 16:11:21 dhellmann: yes, a mock was made: https://review.openstack.org/#/c/145989/ 16:11:32 bknudson: ok, cool 16:11:33 would be nice to see this mock in oslo_utils 16:12:00 class TimeFixture(fixtures.Fixture): 16:12:08 oh, wrong, we were using the mock before. 16:12:18 we switched to set_time_override() instead of the mock. 16:12:21 oh, yeah, we should put that in the lib 16:13:03 bknudson: do you want to work on a patch for that? 16:13:11 dhellmann: I'll put it on my list! 16:13:33 #action bknudson to work on patch to add fixture for timeutils in oslo.utils 16:13:37 thanks! 16:13:57 anything else from liaisons before we start our status review? 16:14:21 Not from me 16:14:29 No, I am good. 16:14:49 ok, moving on then 16:14:55 #topic kilo-2 status review 16:14:58 https://launchpad.net/oslo/+milestone/kilo-2 16:15:18 #topic Drop our use of namespace packages 16:15:18 #link https://blueprints.launchpad.net/oslo-incubator/+spec/drop-namespace-packages 16:15:29 we made good progress last week, in spite of breaking a few things 16:15:36 we have two more patches to land to update oslo.messaging and oslo.vmware 16:15:42 #link https://review.openstack.org/#/c/144794/2 16:15:42 #link https://review.openstack.org/145924 16:15:51 then we can continue releasing libraries as we go, and I think we'll be able to finish those by k-2 16:15:53 there's some reviews out there: https://review.openstack.org/#/q/status:open+topic:bp/drop-namespace-packages,n,z 16:16:18 that oslo.messaging patch is still not working with some of the imports in the other projects, so it needs more work, but I think the vmware changes are OK 16:16:48 bknudson: it's good to see the migration happening already 16:17:36 I was planning to release oslo.db after the meeting, but I have to run an errand so I won't be around if things break. Is someone else going to be around to keep an eye on things for an hour or two, or should I wait until tomorrow? 16:17:43 rpodolyaka1, zzzeek: ? 16:18:08 waiting isn't a problem, but releasing when we're all going offline is :-) 16:18:24 dhellmann: im around 16:18:31 dhellmann: I'll be around too 16:18:31 bnemec: oslo.serialization also looks ready to go 16:18:38 dhellmann: it might be nice to merge https://review.openstack.org/#/c/139738/ 16:18:49 dhellmann: have been trying to get viktors to drop his -1 16:18:55 zzzeek: ok, great, I'll do the release when we're done here 16:19:01 zzzeek: I'll take a look 16:19:06 * dhellmann is typing slowly today 16:19:27 we can wait, but we can also do another release later in the week with that change in it 16:19:40 I want to get the package stuff out there so we can find what it breaks before k-2 16:19:58 make sense? 16:20:10 dhellmann: You already released oslo.serialization: http://lists.openstack.org/pipermail/openstack-dev/2015-January/054049.html 16:20:21 bnemec: hmm, I forgot to update my notes 16:20:23 dhellmann: OK that’s fine, its a bug fix so should be out soon 16:21:08 zzzeek: yep, there's no reason to delay once we have the ns pkg version of the lib out there 16:21:32 I'm only spacing out the other releases because we anticipate things breaking, and I want to know we can fix issues by limiting them to one lib at a time 16:21:46 dhellmann: that’s good. so just release as it is 16:22:02 sounds good 16:22:06 let's move on then 16:22:08 there was some discussion of a hacking rule to prevent use of the deprecated package names, which I agree with ... there's a reviews in flight and merge conflict resolution to consider. 16:22:48 bknudson: Is did someone already write that? 16:22:51 bknudson: yeah, it's ok if projects want to maintain something like that I suppose. I don't want to have to wait on oslo releases until it's updated, though, so I'm not super excited about having it 16:23:18 I don't want to be in a situation where we can't release an oslo lib because a hacking rule would break a bunch of apps all at once 16:23:19 bknudson: That was on my list for today. :-) 16:23:25 jungleboyj: apparently dims__ has one in nova. 16:23:39 bknudson: Cool, I will look at what he did. 16:23:44 bknudson: yep just a regexp 16:23:49 not sure if it should be shared or copied. 16:23:57 dhellmann: So you would prefer this be a per project thing? 16:24:34 jungleboyj: it depends on how it's implemented. How does it know which libs need the rule applied, and which don't? 16:24:39 if it's like other hacking rules, we have to explicitly update. 16:25:10 dhellmann: Not sure, will look at what dims__ did. I was going to add it and then update it with each namespace I got moved over. 16:25:22 dims__: Do you have a review link? 16:25:36 that would work. I'm not sure how we could do that in the central hacking project, but i haven't given it any real thought 16:25:49 I have no problem with project-specific... we seem to have a lot of projects in keystone so it's annoying. 16:26:13 dhellmann: Good point, seems it would need to be project specific. 16:26:45 https://github.com/openstack/nova/blob/master/nova/hacking/checks.py#L452 16:26:55 jungleboyj: yeah 16:27:21 bknudson: Thanks! 16:27:54 how about oslo-incubator? 16:28:03 want a hacking check in there? (does if have them already?) 16:28:31 bknudson: I haven't started updating any of the code there, yet. I was counting on being able to delete more of it. :-) 16:29:14 we have several other blueprints to look at, so let's move on and come back to this one if we need at the end of the meeting 16:29:15 #topic graduate oslo.context 16:29:21 #link https://blueprints.launchpad.net/oslo-incubator/+spec/graduate-oslo-context 16:29:27 I think for this we just need to remove the old code from the incubator. 16:29:34 Is that blocked on the oslo.log release? 16:29:41 #link https://review.openstack.org/#/c/140149/ 16:29:43 dims: ? 16:31:05 ok, I'll check with dims__ later today 16:31:12 #topic adopt debtcollector library 16:31:16 #link https://blueprints.launchpad.net/oslo-incubator/+spec/adopt-debtcollector 16:31:26 harlowja_away: ? 16:32:37 ok, I'll bump that one to kilo-3 16:32:59 #topic graduate oslo.policy 16:33:05 #link https://blueprints.launchpad.net/oslo-incubator/+spec/graduate-policy 16:33:12 ayoung is working on this one 16:33:19 and I don't see him online at the moment 16:34:34 I think we're waiting on the actual import to happen, which is probably blocked on that governance change. That should happen today. 16:34:38 dhellmann: I caught him in openstack-keystone 16:34:51 There's a link in the bp to the repo set up to be imported, so we could start reviewing that 16:34:52 bknudson: thanks 16:34:58 #link https://github.com/rodrigods/oslo.policy 16:35:49 ayoung: we were just checking up on https://blueprints.launchpad.net/oslo-incubator/+spec/graduate-policy 16:36:07 dhellmann, knocking off items on the ticklist 16:36:47 ayoung: it seems like it will be tight to get the release in by k-2, but some of that depends on how much cleanup is needed before a release 16:37:34 dhellmann, why? It's pretty minimal 16:37:45 ayoung: review time? 16:38:00 dhellmann, K+2 is when? 16:38:04 how much do you anticipate changing once that repo is imported? 16:38:16 k-2 is 5 Feb, about 4 weeks from now 16:38:42 dhellmann, we can't get the library released by then? If not, we are broken 16:39:02 ayoung: come on, you know this isn't the only thing we're doing right now 16:39:04 To start? Nothing 16:39:18 I havea achange I want in, but if it misses, its not worth holding up the release 16:39:35 dhellmann, I realize, but this is just procedural 16:39:54 The code is not going to change, just the distro mechanism 16:40:14 I'm not even assuming that the other projects will switch over to using it 16:40:18 but let's get it out 16:40:36 RIght now we have the launchpad stuff basicall set: 16:40:54 need to probably put together a policy specific group , as it is all owned by me right now 16:40:57 who should that be? 16:41:43 #link https://launchpad.net/oslo/oslo.policy 16:42:16 dhellmann, the repo is in git under rodrigods 16:42:22 HA! 16:42:34 ayoung: you must have made him really mad. 16:43:01 dhellmann, don't you like me any more? 16:43:33 ayoung: I think for the lp project you just need to set it to be part of Oslo: http://docs.openstack.org/infra/manual/creators.html#put-your-new-project-in-the-correct-project-group 16:43:42 bnemec, done 16:43:59 bnemec, see the link? Part of oslo 16:44:14 bnemec, question is who should be the maintainers. 16:44:30 I'm not Oslo core, nor are any of the other Keystone core 16:44:34 I'm guessing we need a minimum of 3 16:44:48 as that is what we need to always have 2 to be able to code review 16:44:55 seems like the maintainers would start with whoever's on there now, + whoever else you want. 16:45:16 then can trim from there 16:45:22 so: bknudson "whoever's on there now" meaning what? 16:45:29 Rightnow it is just me 16:45:37 O 16:45:42 I'd say: 16:45:51 ayoung: whoever has core reviewer authority on policy in oslo-incubator. 16:45:51 Oslo cores are always added (part of the oslo-core group), and we can add anyone from Keystone that you want 16:46:12 I lost internet, reading scrollback 16:46:19 bnemec: thanks 16:46:26 #undo 16:46:27 Removing item from minutes: 16:46:30 bnemec, Ok, I'll make an oslo.policy.core group and start with the oslo core team plus.... 16:46:33 ayoung: http://docs.openstack.org/infra/manual/creators.html#set-up-supervisors-for-your-project 16:46:38 me, morganfainberg bknudson 16:46:41 ayoung: oslo-core should be added to the review group in gerrit 16:46:48 ayoung: and yourself, and then we can talk about who else needs to be added 16:46:55 I do want to make sure we're adding people interested in reviewing, and not just people we *want* to be interested in reviewing 16:47:06 Is there a difference? ;-) 16:47:12 ++ oslo core needs to be there. 16:47:35 dhellmann, it was origianlly going to be a Keystone project. I'd like a healthy dose of Keystone core in there, as we want people that understand the wide scope of policy 16:47:50 and people that have touched the policy mechanism in the past 16:47:58 ayoung: yep, that's fine, I just want to make sure we get some sort of confirmation from them as we add them 16:48:10 You can add me to that list. 16:48:12 OK...so new group? oslo-policy-core is the right name? 16:48:24 (Quick confirmation) 16:48:38 or oslo.policy.core? 16:48:48 ayoung: I spent many many hours working on http://docs.openstack.org/infra/manual/creators.html Please go read it. 16:49:12 dhellmann, so hyphens? 16:49:21 that is consistent with the other names, yes 16:49:21 dhellmann: ooh nice documentation!! 16:49:34 “This is a long document. It’s long because it has to be, not because we want it to be. If you follow it, everything will be fine." 16:50:22 dhellmann, I had been reading it, just the oslo.policy value got prepoulated. My real question is "Is there a group for any other oslo library that is different from oslo-core?" And, if so, is there a naming convention 16:50:25 I'd really like to have a script to do this stuff, but a lot of the pieces are tricky to automate 16:50:44 ayoung: yes, we have been creating oslo-foo-core and oslo-foo-release groups 16:50:50 Got it 16:50:56 ayoung: See https://review.openstack.org/#/admin/groups/?filter=oslo 16:50:57 especially for libs that have reviewers who are not oslo-core 16:51:11 maintainer is now oslo-policy-core. I'll seed it 16:51:30 wait, we're talking about gerrit right? 16:52:19 ayoung: launchpad and gerrit work differently 16:52:28 starting with launchpad first 16:52:45 https://launchpad.net/~oslo-policy-core 16:53:04 hadn't gotten as far as Gerrit yet. 16:53:30 ok, apparently some of the oslo-specific details got lost when I extracted that documentation from the wiki to move to the infra manual 16:53:46 the maintainer on lp should be https://launchpad.net/~oslo-drivers 16:53:59 and the driver should also be the same 16:54:27 dhellmann, Um. Not quite, as that rules me out 16:54:38 Oh, wait, drivers? 16:54:41 yeah 16:54:46 Isn't that like the galactic list? 16:54:52 we're going to have to talk about this later, ayoung, we're running out of time 16:55:10 dhellmann, that is fine. I was just working through the list 16:55:25 #topic notification dispatcher filter 16:55:28 #link https://blueprints.launchpad.net/oslo.messaging/+spec/notification-dispatcher-filter 16:55:32 dhellmann, if you update the doc and ping me, I'll take it from there 16:55:34 it looks like I blocked the patch for this until the namespace package changes merge into oslo.messaging 16:55:39 sileht? 16:55:55 dhellmann, I have just propose a patch 16:56:07 dhellmann, https://review.openstack.org/#/c/146547/ 16:56:08 sileht: ok, do you think this work will complete for k-2? 16:56:29 I hope to release a new version of oslo.messaging for k-2 16:57:05 ayoung: the vmware lib sets a precedent for having a separate set of drivers so go ahead with that plan 16:57:14 dhellmann, if my change works work you can squash it into your change and they we can release a new version 16:57:23 dhellmann, will do 16:57:35 sileht: ok, rebasing those namespace changes is pretty painful but I'll give it a try 16:57:48 oh, squash, I see 16:58:29 yeah, I see what you're doing 16:58:47 ok, sileht, if this works it should let us unblock the notification dispatcher 16:59:06 #topic application-agnostic logging parameters in format strings 16:59:06 #link https://blueprints.launchpad.net/oslo.log/+spec/app-agnostic-logging-parameters 16:59:06 I think we can do this work after oslo.log has graduated, but not for k-2 16:59:33 it's in the lib, with next-kilo, so either way I think it's fine 16:59:49 #topic semantic version support for pbr 16:59:49 #link https://blueprints.launchpad.net/pbr/+spec/pbr-semver 16:59:49 lifeless? 16:59:58 I'll check with him on this one later, we're about out of time 17:00:02 #topic open discussion 17:00:05 one minute or so 17:00:16 oops, not even 17:00:34 I think we're in pretty good shape for k-2. We have time to finish quite a bit of what we have scheduled for then. 17:00:36 thanks everyone! 17:00:41 thanks 17:01:08 #endmeeting