17:00:02 #startmeeting ironic 17:00:03 Meeting started Mon Oct 23 17:00:02 2017 UTC and is due to finish in 60 minutes. The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:06 The meeting name has been set to 'ironic' 17:00:10 o/ 17:00:22 o/ 17:00:24 o/ 17:00:25 #chair TheJulia 17:00:26 Current chairs: TheJulia dtantsur 17:00:29 o/ 17:00:32 o/ 17:00:37 #link https://wiki.openstack.org/wiki/Meetings/Ironic our agenda 17:01:07 o/ 17:01:15 o/ 17:01:31 o/ 17:01:40 o/ 17:01:45 o/ 17:02:06 hi all! 17:02:17 #topic Announcements / Reminder 17:02:25 #info Congratulations to TheJulia for being elected to the TC 17:02:33 * TheJulia blushes 17:02:43 yippee!!! 17:02:48 \o/ 17:02:48 dtantsur: Thanks! 17:02:54 Nice! :) 17:02:55 \o/ 17:02:59 o/ 17:03:02 o/ 17:03:19 congrats TheJulia!!! 17:04:10 #info A bunch of releases from master and stable branches to happen this week 17:04:19 \o/ 17:04:42 ironic-lib 2.11 was just released with etingof's "by_path" root device hint 17:04:47 more cool stuff to follow 17:05:01 mmmm, anything else to announce? 17:05:32 I'll be traveling next week starting Tuesday night, to sydney. If there is anything people need from me, or want to pile on my plate, please let me know by monday. 17:05:34 oh, note that releases are going out slowly, as the releases tooling has not fully recovered from zuul v3 move 17:05:58 dtantsur: maybe mention the email thread about traits. 17:05:59 TheJulia: you'll have to party for all of us who stay home :D 17:06:08 rloo: aha 17:06:22 #info Traits discussion is happening on the ML 17:06:25 lemme find links 17:06:34 wondering if the meeting has to happen this week if more folks are travelling next week 17:06:35 * TheJulia makes a wile e coyote sign that says "eek!" 17:06:40 #link http://lists.openstack.org/pipermail/openstack-dev/2017-October/123675.html 17:06:54 #link http://lists.openstack.org/pipermail/openstack-dev/2017-October/123893.html planning for a call 17:07:01 rloo: Thu is the winner currently 17:07:23 anything else? 17:07:30 dtantsur: when does voting end? 17:07:42 TheJulia: let's say EOD tomorrow 17:07:48 ok 17:07:50 dtantsur: So we will do a pike release or Ironic soon? 17:07:56 dtantsur: Another one, that is. 17:08:03 s/or Ironic/of Ironic/ 17:08:11 :p 17:08:12 if you're getting a 404 on the link in the thread, remove the `.` at the end 17:08:12 Pike? lemme check 17:08:49 o/ 17:08:59 mjturek, yes i got that 17:09:09 mjturek, thanks for pointing that out 17:09:25 jlvillal: yep, enough good stuff for another pike release 17:09:39 anything else? 17:09:44 dtantsur: Great. One of our customers wants a fix in there :) 17:10:02 btw, if you wonder how I do it, there is a great tool in the openstack/releases repo: ./tools/list_unreleased_changes.sh stable/pike openstack/ironic 17:10:17 I suspect we need to say thank you to dhellmann for it :) 17:10:19 Thanks. 17:10:21 I <3 those tools 17:10:36 thanks dhellmann! 17:11:22 okay, let's move on, I guess 17:11:31 #topic Review action items from previous meeting: http://eavesdrop.openstack.org/meetings/ironic/2017/ 17:11:46 #link http://eavesdrop.openstack.org/meetings/ironic/2017/ironic.2017-10-16-17.00.html 17:11:55 1. dtantsur to ask milan and sam about approving https://bugs.launchpad.net/ironic-inspector/+bug/1719812 17:11:56 Launchpad bug 1719812 in Ironic Inspector "[RFE] support API access policy definition and enforcement " [Wishlist,Fix released] - Assigned to Pavlo Shchelokovskyy (pshchelo) 17:12:09 the patch was merged in the meantime, sooooo :) good job, Dmitry! 17:12:28 :) 17:12:31 :-P 17:12:43 2. TheJulia to take lead the bug triaging this week 17:12:53 how did this go? I see the bug number decreasing 17:13:16 I fell down on triaging due to being sick last week, although it looks in fairly good shape this morning 17:13:24 might be partly due to me triaging the HIGH bugs last week, i 'removed' a few of them 17:13:48 nice, thanks rloo and TheJulia 17:14:37 #topic Review subteam status reports (capped at ten minutes) 17:15:01 #link https://etherpad.openstack.org/p/IronicWhiteBoard line 143 17:15:32 oh, announcement. zuulv3 is working. yayyyyyy! :D 17:16:04 \o/ 17:16:08 Woot! Does that include stable too? 17:16:15 jlvillal: not yet 17:16:24 vdrok has patches to move jobs to stable branches too 17:16:37 rloo: re "Old ironic CLI deprecation", we call it done for now, right? 17:16:44 dtantsur: yup 17:16:46 so we can remove the item before the next meeting 17:16:56 (after it gets recorded on the ML) 17:17:00 dtantsur: yup. that's my plan. or yours :) 17:17:08 whoever is faster :) 17:17:40 it feels like good progress on everything, except for rescue 17:17:47 we need to give it some attention, I guess 17:17:49 dtantsur: wrt API version negotiation: I'm good if the version=1 work goes into a subsequent release 17:18:04 rloo: I'll check how easy it is to do it 17:18:14 it may be one-two lines, then I'll roll it into the same patch 17:18:27 as it's also kind of a breaking change 17:18:31 dtantsur: the decision about what to do with 'ironic' CLI is more important i think 17:18:45 dtantsur: i tested, and setting version to 1 never worked if i tested correctly 17:18:55 oh, I'll double-check 17:19:18 mm, let's get back to "ironic" fate in the discussion section 17:21:17 so wrt Policy-in-code (L296). I just updated it a bit. but i think it is finished, yes? 17:21:24 seems so 17:21:35 jlvillal: we're ready for the tempest plugin switch, yeah? 17:21:51 so we can delete that task too 17:21:53 dtantsur: I think so. They only thing that would hold it up is the stable branches. 17:22:22 dtantsur: I kind of think we might want to wait until we have stable branches transitioned over. 17:22:25 What do you think? 17:22:35 yeah, maybe we should wait until after the Forum to let the zuul v3 transition settle down 17:22:50 The stable branches use the same tempest as master. 17:22:51 yes, wait for zuul v3 transition 17:22:56 That does kind of sound like a good idea 17:23:08 which reminds me -- i'd like to see volunteers to work on the zuulv3 migrations 17:23:25 pas-ha has an etherpad up. names for each of the projects would be good 17:23:43 * dtantsur assigns milan as he is slacking off today 17:23:48 etherpad tracking zuulv3 work: https://etherpad.openstack.org/p/ironic-zuulv3-intree-tracking 17:24:30 I can try taking the bifrost jobs on, although I'm not sure I'll be able to get to it this week. 17:24:55 I can take sushy 17:25:30 I can also try ironic-lib, if nobody wants. but it is blocked on IPA, I suspect 17:25:46 vdrok: you may want to take ironic-lib too 17:25:57 it uses IPA from source, so they're kinda related 17:26:32 * dtantsur suspects pas-ha will do ironic-staging-drivers 17:27:00 anyway, folks, are you done with statuses? we can continue assign names async 17:27:42 * rloo done with statuses 17:28:06 I'm ready to move on 17:28:19 +1 17:28:25 #topic Deciding on priorities for the coming week 17:28:56 "Switch to none auth" is blocked, so let's kick it out? 17:29:23 +1 soft kick 17:30:18 is nova feature spec freeze over? i didn't see any recent reviews from them for the migrate/resize spec 17:30:31 rloo: no idea, but should be 17:30:41 I could not gran anyone to review it.. 17:30:42 * rloo sad 17:30:44 s/gran/grab/ 17:31:01 folks, can we add the ref arch bit please? it's been without reviews for... time 17:31:13 dtantsur: sure 17:31:22 +1 17:31:37 I'm keeping the RESIZE spec, will sync with Matt on their freeze status 17:31:43 dtantsur: oh, if we're planning an ironic release, did you get a chance to see if any HIGH bug fixes should be reviewed? 17:32:11 rloo: not yet 17:32:46 dtantsur: i just did a quick look, don't see anything close to ready. but that doesn't mean there isn't something we really want fixed soon (but i suspect there aren't any) 17:32:59 at least nothing on my radar 17:33:09 we should finish up your rolling upgrades series though 17:33:23 dtantsur: yup, we shouldn't release w/o that dbsync change 17:33:32 yeah, I'm aware of that 17:33:58 okay, how is the list looking? 17:34:18 also, vendors, do you have anything for this week? I only see a patch from oneview 17:34:26 dtantsur: oh, i guess we should put your 'latest' patch there 17:34:34 yep, let's do 17:34:57 dtantsur: We're all set for now. Thanks for asking :) 17:35:19 good! 6 items + oneview + subteams. how is it looking? 17:35:39 +1 17:36:02 rloo made me think whether we should have a similar list for bugs 17:36:11 like, pick 2-3 bugs to fix every week 17:36:25 dtantsur, a good idea i would say 17:36:28 +1 17:36:32 maybe the next person leading the triage could make a list for us 17:36:43 speaking of which, 17:36:47 #topic Appointing a bug triaging lead for the coming week 17:36:55 anybody? 17:37:09 I doub't I'll touch code this week with summit prep, so I can 17:37:16 I'm also not deathly ill this week :) 17:37:23 \o/ 17:37:57 (phew) 17:37:58 TheJulia: let's try to pick a few bugs that don't have a solution yet, and throw them on the priority list for next week 17:38:11 not sure whether it will work, but worth trying 17:38:17 wdyt? 17:38:20 I like it 17:38:50 oh, 'don't have a solution yet'. you mean, no one is working on it? 17:39:14 rloo: well.. I was thinking about it, but we can relax this requirement to just "not fixed yet" 17:39:30 but I'm really worried about bugs nobody is working on 17:39:35 our priorities have all been patch reviews 17:39:37 the bugs that have a patch will eventually be fixed 17:39:47 yes, and I'm suggesting an experiment to change it :) 17:40:05 to change what is included in our priorities? 17:40:12 yes 17:40:24 I do wonder what disaster Cisco CI had. Sounds like an interesting story... 17:40:36 I encourage non-core people to pick such bugs, and I will provide help, if needed 17:40:36 you mean, list 1 or more bugs in that priorities list, that do not have patches submitted for them but that are triaged? 17:40:45 rloo: correct 17:41:09 dtantsur: and then see if anyone assigns/works on it? then what. remove from the list the following week? 17:41:21 rloo: I think that is what dtantsur is hoping 17:41:26 yes 17:41:31 raise visibility a bit more above just saying "go look in launchpad" 17:41:40 as to "the following week", I don't know :) I'd like to try and see 17:41:42 Which helps drive the human interaction part :) 17:41:59 probably not after one week, maybe after two-three weeks of inactivity. or up to the bug triager 17:42:03 TheJulia++++ 17:42:11 ok, you two seem to know what you're planning on doing. maybe i am the only one that wasn't sure :) 17:42:24 moar opinions on this? 17:42:37 i would also say that bugs that have patches do NOT get reviews sometimes. 17:42:51 this is also true, but we have our priorities list for that 17:43:05 we don't put bugs there too often, but it can be changed 17:43:17 we just need somebody to pick them - the weekly triager can be The Person 17:43:39 ok, just wanted to point it out. so now you're suggesting that the weekly triager do two new things, right? 17:44:07 essentially. not just clean up the list, but pick a few bugs that you think deserve to be fixed asap 17:44:25 "Hey everyone, what do you think about this?" 17:44:31 those would be critical or high bugs i would think but anyway 17:44:42 We kind of already do it in a sense, in irc at times, but not with the best broadcasting or recording 17:44:46 probably, we have enough high bugs 17:44:50 yep 17:45:00 wow, 15 minutes left only Oo 17:45:07 let's move on? 17:45:08 Yeah, we really need to move on 17:45:18 #topic Old CLI API version (dtantsur): the spec says to change it, should we? 17:45:23 this is what rloo asked 17:45:34 we're changing the default API version for "openstack baremetal" 17:45:43 should we do the same for "ironic", given that it's deprecated? 17:45:55 it was mentioned in the spec and the code now has a warning that it will be changed in the future. 17:46:05 dtantsur: We came to that consensus with a different context than what the api-sig kind of reached consensus on recently 17:46:30 TheJulia: what do you mean by that? what consensus? 17:47:33 tl;dr That an api user should explicitly make their calls with version awareness. 17:48:18 If we do the same for the api client, as in set latest, then we run the risk of exposing breaking changes unintentionally. 17:48:40 are you talking about the python library? 17:48:45 I am 17:48:51 because the spec only touches the CLIs 17:49:05 i thought we were leaving the python API alone, it does whatever it did 17:49:06 oh, I thought we also covered the api client 17:49:12 okay, nevermind 17:49:14 ignore me! 17:49:22 TheJulia: that's a good topic to discuss, but that's the next step 17:49:23 I'm fine with ripping the "ironic" cli out of the spec 17:49:37 given we are deprecating it 17:49:49 I think the options are 1. remove the warning about the versoin change, 2. change the versoin 17:49:58 meh, /me cannot type "version" 17:50:04 * TheJulia is for 1 17:50:04 am i correct in thinking that TheJulia, dtantsur and I are the only ones that are ready to discuss? if so, i suggest we take it off line. there are other things in the agenda and 10 min left 17:50:24 we three can work it out in irc later 17:50:41 ++++ 17:50:42 I'm fine with that 17:50:55 #topic Mogan (TheJulia): Anyone working with this team? Anyone have visibility into Mogan's wants/needs as they relate to Ironic? 17:51:27 * johnthetubaguy is very curious about mogan too 17:51:36 I was not involved in mogan outside of a few ML discussions 17:51:50 there is a mogan session at summit. TheJulia, would you please report back? :) 17:51:51 So, as some of you may or may not be aware. Mogan has requested to be entered into governance. Naturally, ironic and nova contributors have varying feelings about mogan, but I think it is time for us to try and improve communication. 17:52:05 That being said, Is anyone presently involved with mogan? 17:52:18 I know mriedem has links with those folks, if that helps make connections 17:52:20 Also, is anyone aware of specific needs they need from us? 17:52:38 ++ 17:52:47 i think the person that proposed the tagging for ironic, is involved with mogan. he hasn't done much on that, so am guessing they don't need for mogan. 17:52:48 johnthetubaguy: I think that might be helpful. Is mriedem going to be at the forum? 17:52:50 we're aware of how nova is going to consume our stuff (well, we make it happen) 17:52:51 Personally I like the idea of another option for people to use Ironic. But no strong feelings on it. 17:52:57 TheJulia: I believe he is 17:53:03 but I don't know how and whether mogan consumes what we build 17:53:27 yeah, mogan uses our API. if i understand it. 17:53:36 best to find out at forum, ask for more info from them. 17:53:37 yeah, ironic is one of their drivers 17:53:56 Okay, I'll ping him, and report back after the forum, and hopefully be able to give us a better understanding/context. 17:53:58 but I mean, things like BIOS, deploy templates, traits - are they going to consume all this? do they like our design? 17:54:06 dtantsur: exactly 17:54:12 thanks TheJulia, I really hope we'll know more on the Forum 17:54:15 dtantsur: who wouldn't love our design? :) 17:54:26 * dtantsur is sorry for not going, but.. $LIFE_HAPPENS 17:54:31 dtantsur: plus, as soon as we have any concept of hardware ownership, we need to consider mutlitenecy... 17:54:31 the cut and paste a lot from Nova to kick start mogan, so its very similar right now: https://github.com/openstack/mogan/blob/master/mogan/baremetal/ironic/driver.py 17:54:50 Anyway, Thanks everyone! 17:54:55 i suspect/hope that if ironic wasn't going in the 'right' direction, mogan would let us know 17:55:01 dtantsur: life, indeed, does happen :) 17:55:12 rloo: I hope that too, but I want to make sure nothing prevents them 17:55:22 moving on? 17:55:26 what dtantsur said 17:55:28 yes 17:55:30 #topic RFE review 17:55:38 5 Minute warning 17:55:39 #link https://bugs.launchpad.net/ironic/+bug/1725211 introduce "inspect wait" provision state 17:55:41 Launchpad bug 1725211 in Ironic Inspector "[RFE] introduce "inspect wait" provision state" [Wishlist,Confirmed] 17:55:42 this is short 17:55:50 I'm not even sure why we don't have "inspect wait" 17:55:58 but it was raised on the inspector call that we may need it 17:56:07 All for this rfe 17:56:18 my usual concern is whether we should do microversion magic to hide this state.. 17:56:19 it makes sense. i guess the question is do we approve now, or ask for a spec? 17:56:30 what did we do with enroll? 17:56:37 I don't care about as spec as long as my question is answered 17:56:38 microversion magic is my guess 17:56:45 rloo: with "available" rather 17:56:52 None <-> available 17:57:00 let me think about it... will get back to the spec later today. 17:57:05 i mean the rfe :) 17:57:15 thanks! if in doubt, we can ask for a spec 17:57:24 dtantsur: will do 17:57:48 oh, two more items in agenda, 3 min left... 17:57:53 #topic Open discussion 17:58:10 #link https://review.openstack.org/458867 significant update to our user guide 17:58:32 #info Rolling upgrade feature (namnh): 1. dev-patch: https://review.openstack.org/#/c/419439/ (reviewing) 2. Request "support-rolling-upgrade" tag (not yet) 17:58:45 not sure I understand the last item, but I carefully copy-pasted it from agenda :) 17:59:12 i started reviewing the user guide patch, there's a lot of good stuff there, would be good for people that are familiar with the details/steps of ironic provisioning, to look at that. cuz i don't want to look at code to compare if it is accurate :) 17:59:14 I mean, we need a patch set to get "rolling-upgrade" tag from TC 17:59:22 namnh: ah, right! 17:59:41 rloo: you may have the honor of proposing the tag :) just please double-check the requirements 17:59:49 One minute warning 17:59:49 dtantsur: i don't think the #1 patch is needed in order to get the tag. that is developer doc. 17:59:58 I don't think either 18:00:01 dtantsur: ok, i'll put in on my plate. thx namnh for remindind us! 18:00:13 sure, thanks 18:00:14 yeah, thanks namnh for reminder and vdrok for the docs patch 18:00:17 and thanks all for the meeting! 18:00:22 #endmeeting