17:00:01 #startmeeting ironic 17:00:02 Meeting started Mon Aug 14 17:00:01 2017 UTC and is due to finish in 60 minutes. The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:03 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:05 The meeting name has been set to 'ironic' 17:00:07 o/ 17:00:08 o/ 17:00:10 o/ 17:00:10 o/ 17:00:15 o/ 17:00:23 o/ 17:00:24 o/ 17:00:25 o/ 17:00:27 o/ 17:00:34 o/ 17:00:50 hi everyone! and welcome :) 17:00:58 our agenda as usual can be found at 17:00:59 o/ 17:00:59 O/ 17:01:03 #link https://wiki.openstack.org/wiki/Meetings/Ironic 17:01:17 o/ 17:01:39 o/ 17:01:47 #topic Announcements / Reminder 17:01:56 \o 17:01:59 #info The PTL election is still on, please vote! 17:02:17 we're proud to have one of only two elections this time :) 17:02:32 \o/ 17:02:36 cuz we all love elections :D 17:02:41 yep! 17:02:49 #info Please continue proposing ideas for the PTG: https://etherpad.openstack.org/p/ironic-queens-ptg 17:03:17 #info ironic-ui 3.0.0 was released last week 17:03:36 #info dtantsur on PTO Thu-Fri (17th and 18th) 17:03:39 anything else? 17:03:47 deadlines coming up? 17:04:02 I'll likely have connectivity next week. 17:04:08 I'm traveling 17:04:14 well, the hard deadline is next week 17:04:20 I'd prefer to release everything this week 17:04:34 on Wed at the latest, because of my PTO 17:04:35 ok, so no more features. only major bug fixes? 17:05:01 right. we may put in a few vendor patches, but that's it 17:05:02 depending on our gate state too :( 17:05:11 yeah, and the gate is.. not great 17:05:33 wrt gate -- is whiteboard up to date on that? 17:05:50 nope 17:06:02 * vdrok updates 17:06:08 thanks vdrok 17:06:09 ok, please update and let us know if you'd like us to review or whatever. thx vdrok! 17:06:20 it's a new problem, just appeared today or over the weekend 17:06:41 anything else? 17:07:15 #topic Review subteam status reports (capped at ten minutes) 17:07:38 #link https://etherpad.openstack.org/p/IronicWhiteBoard this time please start with line 118 17:07:56 I've added a list of stuff I consider urgent 17:08:08 we managed to clean up most of it last week, but there is still more things 17:08:17 especially wrt nova and resource classes 17:08:55 #link https://review.openstack.org/#/c/473717/ separates us from calling BFV done for Pike 17:09:09 o/ 17:09:40 dtantsur: wrt the nova patch -- is that the lsat thing that will get nova resource classes working with ironic? 17:09:49 rloo: yes, hopefully 17:10:04 dtantsur: ok, so we need to get that into pike 17:10:06 but it raises quite a few concerns due to the way we publish our resources 17:10:25 and tl;dr of the current discussion, is that we're going to have a few known issues for people not using them 17:10:40 dtantsur: so we need more documentation to be added too 17:11:02 rloo: https://review.openstack.org/491773 covers most of it, there will be nova release notes too on this topic 17:11:15 it's all complicated by the fact that nova already has stable/pike 17:11:19 dtantsur: thx, will look 17:11:36 wrt L152 -- hponeview. seems like it is too late to do that? 17:12:00 I'd say so.. I told the folks that I don't like introducing a dependency without merging the patches to use it 17:13:01 to be clear, the stuff under L150. dtantsur, did you add those, or was that for people to add their patches? 17:13:21 rloo: I added it for people to extend without any promise that the stuff will be merged 17:13:30 I don't want the vendors to feel ignored 17:13:34 dtantsur: gotcha, thx. 17:14:58 sambetts: unfortunately, it does not seem like we can manage to get IPA API in :( 17:15:12 dtantsur: wrt driver composition -- is there anything more to do there? 17:15:14 For ilo hardware type doc , can we consider it part of the overall hardware type work? 17:15:35 wanyen: docs are easy, we can even backport them 17:15:53 rloo: dev docs should be written, never have time for it.. otherwise, only new hardware types 17:15:58 dtantsur: :( never mind... its so close it should land fast for the next release 17:16:04 yep 17:16:23 dtantsur: so UCS and drac 17:16:32 dtanstsur, ok. It's been listed as non-blocker so I was concerned that the code is in but the doc may not get in in time 17:16:34 but drac is wip so probably won't make it? 17:16:45 rloo: yeah. in drac we have to fix a bug first, rpioso has a patch up, but we'll see 17:17:12 dtantsur: so someone ping me about drac, i'm not going to look at it otherwise 17:17:51 rloo: Will do :) 17:17:54 rloo: you can review https://review.openstack.org/#/c/493421/, it seems ready, just the gate is not good 17:17:56 i am willing to look at sambetts ipa version today unless folks think it won't make it 17:18:34 i guess it means i will look if some other core will also 'commit'. otherwise, no use. 17:18:43 well, I think it won't make it, and I'd give vendor patches some love in the end.. 17:18:51 dtantsur: ok 17:19:16 rloo: ty 17:19:50 dtantsur, ty 17:19:59 dtantsur: ty 17:20:29 given the gate, etc, we should aim to get things done by end of tues, so dtantsur can do a release on wed 17:20:37 yes, this ^^^ 17:21:06 dtantsur: and you're back next week to do a final release if need be? 17:21:08 #info Please be ready to land the stuff by Tue, as we plan to release ironic pike on Wed CEST morning 17:21:15 rloo: yep, I'm only out for 2 days 17:21:22 my patch chain for clients refactoring is also going to miss Pike AFAIU, so please someone put -2 on the first one https://review.openstack.org/#/c/469627/ 17:21:30 we can do plenty of docs though, those jobs are pretty stable :) 17:21:32 thanks pas-ha, will do 17:21:53 pas-ha: :-( 17:22:18 rloo: no big problem, immediate bug with glance is worked around 17:22:34 yeah, let's make it one of our early Queens goals 17:22:41 vdrok, lol 17:22:44 +1 17:22:54 vdrok++ the docs are good, let's have moar of them 17:22:54 * rloo ready to move on 17:23:01 everyone ready? 17:23:17 #topic Deciding on priorities for the coming week 17:23:38 easy one - make a release :) 17:23:45 I think I'll just remove the priorities list, and leave "Release countdown" 17:23:50 everyone take two days off with dtantsur does 17:24:02 s/with/when/ 17:24:05 :) 17:24:05 hehe 17:24:10 +1 17:24:15 * milan actually does :D 17:24:59 dtantsur: should we add ucs & drac hw types to that list? 17:25:14 yeah, this is a good call 17:25:31 rloo: please :) 17:25:50 rloo, dtantsur: Yes, pleeeeez 17:26:30 * pas-ha envisions sambetts|afk and rpioso as cat from Shrek :) 17:26:41 lol 17:27:00 heh 17:27:19 dtantsur: if we have priorities, we should add nova/resource classes too 17:27:36 rloo: that's in the countdown 17:27:42 yes, this ^^ 17:27:55 just don't want to duplicate the lists 17:27:59 pas-ha: lol 17:28:12 ok, i guess i didn't understand why the hw types weren't added to the countdown at L104 17:28:28 so thought the priorities were to highlight stuff in the countdown :) 17:28:30 rloo: because I did not consider them blockers 17:28:39 dtantsur: ah. ok. 17:29:14 how's the list looking? 17:29:43 dtantsur: messy. i'll clean it up (remove crossed out stuff) after rama sends out the weekly email. 17:29:51 dtantsur: otherwise, looks doable. 17:29:59 ok :) 17:30:21 so, from release perspective, that's mostly docs (kudos to rloo for jumping on reviewing them), and ova stuff 17:30:21 dtantsur: oh, there is a question about virtualbmc. is it ready to release? 17:30:30 *nova 17:30:45 pas-ha: yeah, and a couple of hw types 17:30:55 rloo: I think so, just checking if anybody disagrees 17:31:09 actually, I plan to send release requests for vbmc and sushy-tools tomorrow 17:31:25 maybe for IPA and inspector too, if the folks finish reviewing the release notes 17:31:27 dtantsur: ok, am deleting the '?' then :) 17:31:31 ++ 17:31:35 bifrost is done? 17:31:42 dtantsur, just a query...sushy and sushy-tools are different? 17:31:42 pas-ha: TheJulia has requested a release 17:31:47 ok, cool 17:31:54 Nisha_Agarwal: yep, sushy-tools contains the emulator for the CI 17:32:03 it's dev stuff, not something we ship to people 17:32:13 dtantsur, ok 17:32:26 people involved in ironic-staging-drivers: please get ready for a release too 17:32:27 * Nisha_Agarwal was not aware of it 17:32:38 ok 17:32:39 I may do it next week, but probably better this Wed 17:32:51 dtantsur: should that be added to our list then? 17:33:00 rloo: it's not under our governance 17:33:08 dtantsur: right 17:33:26 and I think Molteniron does not have releases, right? 17:33:40 yeah, not yet anyway 17:33:44 dtantsur: might want to merge the hw types for staging drivers first https://review.openstack.org/#/c/466000/ 17:34:01 pas-ha: I tried reading it, but encountered ETOOMUCHBASH :) 17:34:06 I will retry until success 17:34:25 hehe, no problen if it does not land in Pike 17:34:30 anything else? 17:34:47 my next topic was supposed to be "Review of release readiness", but I feel like we've just done it 17:34:56 so 17:34:58 #topic Open discussion 17:35:29 dtantsur, so are we planning to release ironic this week? 17:35:44 Nisha_Agarwal: yes, probably Wednesday morning (CEST) 17:35:48 not final I suppose, kind of RC? 17:36:03 vdrok: well, yes and no 17:36:17 it's where we create stable/pike 17:36:36 and it may end up our final release, if we don't backport any patches until the final release time 17:37:08 yup, that's what I meant :) basically it is an RC, but not named as one 17:37:21 yeah, yeah, something like that 17:37:41 so we'll get 9.0.0, I think. and the final may be 9.0.1 if we fix something, or just this 9.0.0 17:38:02 dtansur, that does not leave much time for reviewing all the patches in teh countdown and non-blocker list 17:38:06 dtantsur: if we backport docs, it'll still be 9.0.0 right? 17:38:15 rloo: docs backports are fine, they are not features 17:38:28 dtantsur: so they don't up the version. 17:38:31 wanyen: I know. let's keep reviewing :) 17:38:39 wanyen: i highly doubt that all the non-blocker patches will make it 17:38:47 rloo: wait, sorry. any new release requires a version bump 17:38:59 rloo: if we backport docs, we'll get 9.0.1 too 17:38:59 dtantsur: ah, ok, so there will be a version bump even if only docs 17:39:12 patch version are cheap, we can have as many as we want 17:39:27 what the release team does not want to see is feature backports (ending up in 9.1.0) 17:39:32 wanyen: i meant the vendor-related ones. am going to try to get the non-blocker code patches in. 17:39:46 dtantsur: yup 17:40:08 actually, we have enough time. the only problem is that I'm buried in that nova change :( 17:40:34 dtantsur: so you aren't sure the nova change works? 17:40:50 rloo: it works, but there is some design contention 17:41:00 dtantsur: oh. :-( 17:41:07 i.e. we have a terrible way of reporting out resources. and that gives us headache 17:41:12 s/out/our/ 17:41:14 * rloo runs away 17:41:21 good for you ;) 17:41:36 dtansur, seems like there are still some uncertainty about nova rsc class. so Why it has to be in Pike? 17:41:37 * Nisha_Agarwal is ready to help if dtantsur need 17:42:03 Nisha_Agarwal: thanks a lot! I'm afraid, there is no much chance to jump in at this stage. 17:42:12 :) 17:42:16 I've spent 3 or 4 days talking to nova folks for hours 17:42:24 I know what to do, it's just... messy.. 17:42:29 yeah i saw the patch 17:43:12 dtantsur: so just changing one bit related to only RC is not an option there? 17:43:56 vdrok: it fixes the RC problem, but it still leaves open the issue with our inventory reporting. I'll probably split the patch in two. 17:44:08 lemme just wrap my head around it tomorrow with fresh (?) brain 17:44:27 yeah, but at least the RC one is easy, and we won't be in so much rush :) sure 17:44:38 this is my thought too 17:45:36 anything else? 17:46:35 okay, let's get back to work :) 17:46:38 thanks everyone! 17:46:41 thanks! 17:46:44 o/ 17:46:48 thanks! 17:46:52 #endmeeting