17:00:11 <dtantsur> #startmeeting ironic
17:00:12 <openstack> Meeting started Mon Jul 24 17:00:11 2017 UTC and is due to finish in 60 minutes.  The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:15 <openstack> The meeting name has been set to 'ironic'
17:00:23 <mjturek> o/
17:00:25 <dtantsur> who's here for the ironic meeting? :)
17:00:34 <rloo> o/
17:00:35 <jlvillal> o/
17:00:37 <kaifeng> o/
17:00:38 <stendulker> o/
17:00:38 <crushil> o/
17:00:44 <jlvillal> Me! Me! Me!
17:00:50 <rama_y> o/
17:00:54 <vdrok> o/
17:00:54 <milan> o/
17:01:02 <dtantsur> hi all, great to see you :)
17:01:07 <dtantsur> our agenda is as usual at
17:01:08 <milan> jlvillal, lol
17:01:13 <dtantsur> #link https://wiki.openstack.org/wiki/Meetings/Ironic
17:01:40 <TheJulia> o/
17:01:42 <baha> o/
17:02:04 <dtantsur> #topic Announcements / Reminder
17:02:13 <dtantsur> #info Final releases for ironicclient and ironic-inspector-client this week
17:02:31 <dtantsur> please pay particular attention to reviews in these two projects
17:02:39 <dtantsur> I think the inspector client looks good
17:02:54 <fellypefca> o/
17:02:55 <dtantsur> ironicclient has quite a few outstanding patches (mostly around testing though)
17:03:11 <rloo> dtantsur: please put links to any patches that need reviews, maybe in this weeks' priorities section ^^
17:03:16 <xavierr> o/
17:03:26 <dtantsur> will do when we get to it
17:03:35 <dtantsur> next,
17:03:35 <rloo> dtantsur: any patches that *need* to get in, not all :)
17:03:42 <dtantsur> right, I'm not sure there are many
17:03:55 <etingof> o/
17:03:57 <dtantsur> but I also encourage the folks to review the list and decide for themselves
17:04:07 <dtantsur> and shout loudly if we've forgotten something
17:04:08 <dtantsur> :)
17:04:24 <dtantsur> #info Soft feature freeze proposal: http://lists.openstack.org/pipermail/openstack-dev/2017-July/119883.html
17:04:32 <dtantsur> this is something I was talking about on the last meeting
17:04:44 <dtantsur> so far I got no responses to this thread, I guess people are good with it ;)
17:05:04 <rpioso> o/
17:05:23 <dtantsur> finally,
17:05:25 <dtantsur> #info IPA builder repo was started, please add to your review dashboards: https://github.com/openstack/ironic-python-agent-builder
17:05:37 <vdrok> yup, soft freeze date seems fine
17:06:06 <rloo> dtantsur: does that builder repo participate in the release schedule?
17:06:15 <dtantsur> rloo: nope, at least not now
17:06:25 <rloo> dtantsur: good :)
17:06:27 <dtantsur> I will probably cut releases from it, but we may have to skip Pike
17:06:35 <dtantsur> it's really all nice and empty :)
17:06:44 <dtantsur> any other announcements, reminders or comments?
17:06:58 <rloo> nova feature freeze is this week, but we all know that
17:07:13 <vdrok> attach-detach got in btw
17:07:18 <dtantsur> \o/
17:07:29 <sambetts> \o/
17:07:30 <dtantsur> good news indeed! let's try to get BFV in as well
17:07:30 <rloo> and our gates should be working...
17:07:46 <dtantsur> yeah, we had big problems last week, but should be better now
17:08:12 <TheJulia> Reviews on the BFV nova patch would be appreciated, and hopefully we should have any minor details sorted out in the next day or two.
17:08:17 <dtantsur> #topic Review subteam status reports (capped at ten minutes)
17:08:30 <dtantsur> #link https://etherpad.openstack.org/p/IronicWhiteBoard line 125
17:08:32 <rloo> oh. next week is Queens PTL self-nomination. All you wanna-bee PTLs, start preparing :)
17:08:41 <dtantsur> ouch :)
17:08:51 * TheJulia raises an eyebrow
17:09:17 <rloo> the Queen PTL. that is funny
17:09:21 <dtantsur> yeah, being PTL is easy and fun, everyone should do it :-P
17:09:33 <dtantsur> anyway, /me dives into the statuses for now
17:10:06 <ricardoas> o/
17:10:16 <rloo> is rpioso here?
17:10:18 <dtantsur> sambetts: do we consider "Post-deploy VIF attach/detach" done now?
17:10:27 <rloo> rpioso: what's status of drac hardware types?
17:10:53 <dtantsur> ВЩТУ indeed (<-- regular joke about wrong keyboard layout) :D
17:10:59 <vdrok> :D
17:11:05 <vdrok> or rather ЖВ
17:11:32 <sambetts> dtantsur: I believe so at least for the in-tree neutron driver and nova
17:11:36 <dtantsur> re physical awareness: there are a lot of things in flight
17:11:49 <dtantsur> let's please come up with a minimum list of patches to support this feature?
17:12:39 <dtantsur> sambetts: ^^^
17:12:57 <sambetts> I think the actual feature itself has landed, the remaining patches are a big refactor to clean up the code
17:13:04 <rloo> sambetts: are the networking-baremetal parts for routed networks support good to go, or are they still blocked by physical network awareness?
17:13:21 <dtantsur> sambetts: I think tempest and upgrades are kinda mandatory, no?
17:13:51 <vdrok> yeah, upgrades should get in too I think
17:14:21 <rloo> i reviewed the upgrades code; i suspect they will land this week.
17:14:23 <sambetts> oh yeah, missed that patch in the list, tempest tests and rolling upgrades should definatly be priority
17:14:43 <sambetts> rloo: I should be able to start working on them again now the physical network field is exposed in the API
17:14:57 <dtantsur> it is exposed
17:14:59 <rloo> sambetts: great, please let us know when they are ready to be reviewed!
17:15:37 <sambetts> will do :) there is quite a lot of brain required to get the plumb right so bare with me :)
17:15:44 <sambetts> plumbing*
17:18:22 <dtantsur> are folks done with the statuses, keeping in mind we're going to come back to them in a few minutes?
17:18:26 * rloo good with statuses although skipped the non-high priorities
17:18:33 <rloo> 10 min mark :)
17:18:46 <dtantsur> #topic Deciding on priorities for the coming week
17:18:59 <dtantsur> so, we're still in the docs refactoring business, esp. ironic-inspector
17:19:17 <sambetts> dtantsur: I think the ironic-inspector patches are all good to go now
17:19:26 <milan> sambetts ++
17:19:30 <dtantsur> cool!
17:19:49 <dtantsur> also, some progress with BFV (cheers!), rolling upgrades and physnet awareness
17:19:51 <TheJulia> Bifrost also, but that takes a lower priority in my book since it is a massive re-org of the docs.
17:20:50 * milan wanted to capture the reviews input on the dnsmasq pxe filter patch before forgetting the discussion, hence the docs delay
17:20:51 <rloo> should we remove bifrost, molteniron from docs list then?
17:20:57 <dtantsur> mjturek, TheJulia: what's else critical for BFV? I think you have a couple of small but important patches
17:21:10 <dtantsur> rloo: molteniroic is small, let's keep it
17:21:30 <dtantsur> fine with removing bifrost for now. anyway I get weekly reminders about our docs :)
17:21:34 <rloo> dtantsur: ok
17:22:15 <TheJulia> dtantsur: Critical, the patch mjturek is working on right now, the tempest tests, the first nova patch, and I _think_ we are good until next cycle for generic drivers.
17:22:45 <TheJulia> oh, and the project-config patch for the experimental ci job
17:22:58 <rloo> sambetts: out of curiosity, any idea when phase 2 of doc migration will happen (to move api ref)?
17:23:03 <dtantsur> TheJulia: I think we list all but the first one. could you please add it to the priorities?
17:23:17 <TheJulia> dtantsur: on the priority list?
17:23:21 <mjturek> yep - there are some other small patches on the ironic side, but nothing critical I belive
17:23:39 <rloo> Oh. we forgot to mention that we did another client release last week
17:24:15 <TheJulia> dtantsur: added
17:24:28 <dtantsur> thanks!
17:24:55 <TheJulia> rloo: Thanks, last week was hectic :)
17:25:34 <sambetts> rloo: according to the doc-migration etherpad sometime next cycle, but I've asked in #openstack-doc for a confirmation
17:25:40 <rloo> TheJulia: :)
17:25:44 <dtantsur> so, how is the list looking?
17:26:07 <rloo> sambetts: thx. once you get confirmation, maybe we can add a comment in that patch, and remove it from our whiteboard.
17:26:14 <rloo> dtantsur: busy. good though :)
17:26:24 <sambetts> rloo: which patch?
17:26:31 <rloo> sambetts: the api-ref patch
17:26:32 <dtantsur> yeah, good to me as well. any comments/additions?
17:26:35 <rloo> sambetts: L99
17:26:55 <sambetts> rloo: yup makes sense
17:27:03 <dtantsur> ready to move on? we have quite some work ahead :)
17:27:07 <TheJulia> dtantsur: Is anyone, as far as your aware, working on a configuration reference doc for ironic?
17:27:22 <dtantsur> not that I'm aware of. and ditto for ironic-inspector. thanks for the reminder
17:27:28 <TheJulia> :)
17:27:29 <rloo> TheJulia: there used to be a config ref for ironic. does it just need to be mvoed over?
17:27:52 <TheJulia> rloo: if it is an html auto rendered magical reference, yes
17:27:54 <rloo> TheJulia: it was generated from the ironic configs
17:28:05 * dtantsur looks at the clock
17:28:21 <rloo> TheJulia: i'll take a look and see if i can find it
17:28:25 <sambetts> me and rloo both have ironic doc clean up patches in flight, I think we need some opinions on: https://review.openstack.org/#/c/485820 and https://review.openstack.org/#/c/486649/
17:28:29 <dtantsur> thanks rloo :) added to line 103
17:28:32 <TheJulia> rloo: thanks!
17:28:43 <dtantsur> #topic continue/finish pruning Ironic priorities for Pike
17:28:47 <rpioso> rloo: I updated the status of the drac hardware type driver.
17:28:54 <dtantsur> thanks rpioso!
17:29:09 <rpioso> dtantsur: :)
17:29:13 <dtantsur> I think we last stopped right before the IPA API wotk (line 242)
17:29:20 * rloo only wants to spend minimal time on docs now since we can get those changes in up to the last minute.
17:29:27 <rloo> thx rpioso!
17:29:52 <TheJulia> rloo: +1, minus the configuration reference stuff, since we are going to get emails to the mailing list about a lack of it until it appears :(
17:29:52 <dtantsur> it seems that sambetts is still decided to get it in
17:30:05 <rpioso> rloo: np
17:30:13 <dtantsur> TheJulia: I'm already getting these emails :)
17:30:31 <dtantsur> so, keeping IPA API so far?
17:30:44 <sambetts> I rebased and pushed updated patches for the IPA versioning logic so they *should* be ready to review
17:31:02 <rloo> i think that is technical debt; if it isn't hard, we should do it
17:31:28 <dtantsur> yep
17:31:39 <TheJulia> sounds good to me
17:31:56 <dtantsur> python 3 is something we discussed already, right?
17:32:07 <TheJulia> yes
17:32:14 <dtantsur> node tags
17:32:26 <rloo> nisha doesn't seem to be here; she was going to report status today of python 3
17:32:35 <dtantsur> sigh
17:32:41 <dtantsur> well, it should not be a lot of work
17:32:55 <rloo> 'should not' :)
17:32:57 <TheJulia> I think we landed one of the two outstanding patches last week for python3
17:33:14 <dtantsur> vdrok: do you still want and can dedicate time to node tags at this stage?
17:33:17 <rloo> we can leave that as optional priority. yes?
17:33:25 <rloo> oh, i was referring to python 3.5
17:33:29 <dtantsur> rloo: +1 to keeping python 3
17:33:34 <dtantsur> as it's an openstack-wide priority
17:33:51 <dtantsur> as to node tags.. I'd limit the number of API additions, to be honest
17:33:56 <vdrok> dtantsur: I think so, yes. at least we've clarified some things with rloo last week
17:33:59 <dtantsur> especially since we cannot get it in ironicclient
17:34:05 <vdrok> dtantsur: fine with me
17:34:21 <rloo> oh yes, good point. no more changes to client.
17:34:32 <TheJulia> re: node tagging: I feel the same, I'm hesitant to land api patches at this point as we are going to release the client this week.
17:34:34 * dtantsur votes for moving it to Queens
17:34:39 <vdrok> it's not something we desperately need, it's also technical debt
17:34:50 <dtantsur> not technical debt, rather planning debt
17:34:50 <TheJulia> +1 to user visible to queens
17:35:00 <dtantsur> objections?
17:35:28 <dtantsur> done :)
17:35:33 <dtantsur> rescue mode
17:35:35 <TheJulia> dtantsur: that was fast :)
17:35:46 <dtantsur> TheJulia: it's not hard to revert, right? :)
17:35:52 <TheJulia> dtantsur: nope
17:35:54 <rloo> rescue mode needs nova patch. so it can't make it in PIke
17:36:00 <TheJulia> re: rescue, next cycle
17:36:07 <rloo> do we want to make it a priority to do the rescue-ironic patches now?
17:36:20 <dtantsur> well, I'm afraid we have enough priorities
17:36:29 <rloo> ok, so rescue in Queens then.
17:36:30 <TheJulia> re: tags: vdrok: I'd love to gain an understanding of how it is technical debt, but that is not for this meeting at the moment.
17:36:34 <anupn> i think so would be good to add in priority, gate are passing now i believe
17:36:37 <stendulker> rloo: Few reviews on 'rescue' would help, as there wer no reviews on it
17:36:51 <rloo> stendulker: i agree. more reviewers would help too.
17:37:29 <dtantsur> anupn, stendulker: I understand it's sad, and I know it's not your fault. However, we're where we are now, and we cannot stretch ourselves more :(
17:37:31 <vdrok> TheJulia: in a sense that we have half of the code landed for quite a while :) but yeah, maybe rather planning debt
17:38:16 <anupn> dtantsur: o/
17:38:44 <dtantsur> next?
17:38:56 * TheJulia is good with next
17:39:04 <dtantsur> Split away the tempest plugin
17:39:07 <rloo> are tempest plugins one of Queen's goals? (or maybe they were talking about it and it isn't)
17:39:17 <dtantsur> potential Queens goal
17:39:31 <dtantsur> I'm not sure the goals are finalized (I am supposed to know it, sigh)
17:39:47 <TheJulia> I think we need to move to queens at this point
17:39:48 <rloo> well, where are we at with those? i thought we were waiting for patches to be submitted
17:39:48 <dtantsur> anyway, I don't really feel like potentially de-stabilizing the gate right before the release :(
17:40:01 <rloo> yeah, move to Queens then.
17:40:03 <dtantsur> rloo: we have a huuuuge chain of patches, and it has to grow more
17:40:08 <dtantsur> jlvillal: what do you feel about this?
17:40:36 <jlvillal> dtantsur: I'm leaning towards waiting on it. Not a pressing priority. And I've let it hit the back-burner
17:40:52 <jlvillal> dtantsur: Much rather focus on our big priorities like BFV and others
17:40:52 <rloo> OMG, I thought we had merged solio's tempest patches
17:40:57 <dtantsur> nope..
17:41:11 <dtantsur> objections to moving it to Queens?
17:41:33 <jlvillal> Not from me. I think it is a good early cycle goal.
17:41:42 <TheJulia> jlvillal: ditto
17:41:43 <rloo> jlvillal: +1
17:41:52 <dtantsur> done
17:41:59 <dtantsur> now, deploy steps
17:42:10 <dtantsur> this is at the spec stage
17:42:16 <TheJulia> Queens imho
17:42:17 <rloo> that needs the spec to be approved which won't happen in the near future, so Pike
17:42:21 <rloo> Queens i mean
17:42:24 <dtantsur> right
17:42:38 <rloo> mind your Ps and Qs
17:42:41 <dtantsur> actually, I wanted to ask if the ansible deploy method will cover the same functionality
17:42:45 <vdrok> dtantsur: does the ansible driver cover that maybe?
17:42:47 <vdrok> yup
17:42:53 <dtantsur> PTG topic?
17:42:59 <rloo> +1
17:43:04 <TheJulia> +1
17:43:11 <sambetts_> +1
17:43:12 <dtantsur> ok, this goes to Queens for now, and let's talk about it again
17:43:19 <rloo> although... ok, let's discuss :)
17:43:43 <dtantsur> Supported power states API
17:43:51 <dtantsur> this is the left-over of the soft power/NMI work
17:44:00 <rloo> seems like someone is updating the patch now, so i think we can try to get it in
17:44:15 <rloo> oh, does it need client too?
17:44:17 <TheJulia> Even then, there is the question of should we have an api change
17:44:19 <dtantsur> does not look like a lot of work, but that's an API change
17:44:33 <dtantsur> I mean, we *can* have an API change, it's not going to be exposed in the client
17:44:36 <TheJulia> I say lower priority, a stretch goal for this cycle
17:44:43 <TheJulia> and then land client stuff in the next cycle
17:44:46 <dtantsur> it's already a low priority, yeah
17:44:46 <rloo> so i'm fine because it is small API change. we can land that and the client code. just not release client til Queens.
17:45:05 <dtantsur> okay, sounds good
17:45:07 * TheJulia senses consensu
17:45:10 <TheJulia> consensus
17:45:17 <dtantsur> Available clean steps API
17:45:23 <dtantsur> the mic goes to rloo :)
17:45:43 <rloo> heh, i started to look at it. then got distracted with other things. rolling upgrades is *almost* done, so I can take it up again. If we want?
17:45:58 <dtantsur> I would absolutely love to get it in, even if not exposed through the client
17:46:12 <dtantsur> figuring out available clean steps is not an awesome tasks right now
17:46:29 <dtantsur> other opinions?
17:46:29 <rloo> dtantsur: ok then. will update that in next 1-2 weeks.
17:47:00 <vdrok> yeah, worth getting in I think, thanks rloo
17:47:38 <dtantsur> the last but not the least: etags
17:47:50 <rloo> Queens I think. Patches haven't been rebased for a long time.
17:47:51 <dtantsur> everything needs rebasing; extensive client change expected
17:48:03 <TheJulia> Queens then
17:48:03 <dtantsur> no pas-ha to oppose us :)
17:48:26 <dtantsur> Queens?
17:48:26 <rloo> i htink pas-ha is still on vacation
17:48:28 <vdrok> yeah, pas-ha only gets back in a week, so..
17:48:46 <sambetts_> Should be early queen's priority imo because it touches everything right?
17:48:53 <dtantsur> kind of, yeah
17:49:01 <dtantsur> heh, we already have a backlog for Queens :)
17:49:08 * rloo is concerned about everything we'd like to get into early Queens...
17:49:17 <dtantsur> ok, this was the last item. please take another look at the whole list
17:49:48 <TheJulia> For discussion at the ptg, but perhaps we make it a queen thing targeting the middle to end of the cycle.
17:50:02 <dtantsur> also, should I already remove Post-deploy VIF attach/detach already?
17:50:18 <rloo> dtantsur: not until after rama_y sends out her email :)
17:50:23 <dtantsur> right
17:50:28 <vdrok> oh, yeah, I forgot
17:51:19 <dtantsur> okay, other comments? ready to move on?
17:51:32 <rloo> +1 moving on
17:51:38 <dtantsur> I will send out a email summing up our decisions
17:51:46 <dtantsur> #topic IRC business hours
17:51:51 <dtantsur> so, this is mine
17:52:10 <dtantsur> I see the folks dedicating some time slots where they can actively help people on IRC
17:52:14 <dtantsur> especially newcomers and beginners
17:52:31 <dtantsur> this looks like a good idea. I constantly see folks trying to get help at 2am my time
17:52:36 <dtantsur> without any luck
17:53:00 <TheJulia> I like the idea, but I'm not sure that is going to make the 2am requests for help disappear
17:53:03 <dtantsur> what if we have one slot in Europe evening/US morning and the other in US East evening/APAC morning?
17:53:15 <dtantsur> TheJulia: well, at least people will know *when* they can find someone
17:53:25 <TheJulia> perhaps both "recommended hours" and "if your question is not answered quickly, please hang out, wait." kind of approach?
17:53:45 <dtantsur> the idea of office hours is that we know that at least some people will be there
17:53:47 <TheJulia> dtantsur: true
17:53:52 <rloo> +1 for TheJulia. I am not convinced about the 'official' hours. cuz none of us can really say we will always be there.
17:54:20 <dtantsur> well, for me chances are high I'm online 9am - 5pm UTC
17:54:34 <rloo> 'office hours' seems to make it official. that someone *will* be there. and if not, then someone needs to put up a sign saying 'back in 10 min' or whatever.
17:55:05 <dtantsur> true, it's not cool if absolutely nobody will be online due to priorities, PTO, etc
17:55:06 <rloo> how do the other projects? deal with it. i mean, how do they define office hours, etc.
17:55:50 <dtantsur> I can look more into it, cannot find right now
17:55:51 <rloo> do people look at irc topic? maybe add a note there about hours when folks might be around?
17:56:09 <rloo> or a smart bot that replies if no one else replies? :)
17:56:21 <TheJulia> what about a notification keyword?
17:56:22 <dtantsur> hah, make lucas revive PixieBoots :)
17:56:32 <rloo> I'm sorry, no one seems to be around now to answer your question, you might have better luck trying at ...
17:56:42 <TheJulia> some of us do generally just help people at odd hours if they are prodded :)
17:56:59 <kaifeng> how about whiteboard
17:57:03 <sambetts_> I really like AFK nics we could have a bot that looks and recognises core nics and Afk nics
17:57:08 <rloo> this must be a similar issue in other projects. maybe (yes) time to ask for what 'good practise' would be :)
17:57:22 <dtantsur> sambetts_: my bouncer periodically reboots and reclaims the nick without afk..
17:57:36 <dtantsur> rloo: good call. I can take this task
17:57:39 <sambetts_> Ah :/
17:57:47 <TheJulia> I am only really AFK if I'm asleep or on a jet.... or just ignoring irc
17:57:54 <rloo> thx dtantsur. (I always love it when the PTL looks after us.)
17:58:04 <dtantsur> ^_^
17:58:35 <dtantsur> on a similar topic: some projects experiment with hosting these meetings in their own channels
17:58:49 <dtantsur> I know tripleo did it and like it. I'm not sure how different it is in practice though..
17:59:10 <dtantsur> anyway, one minute alert!
17:59:14 <rloo> why change if nothing is broke.
17:59:20 <dtantsur> true
17:59:28 <sambetts_> I like it personally because it means we don't have to deal with time conflicts and people not finding our meeting
17:59:43 <TheJulia> I like the distinct channel from a context switching standpoint
17:59:43 <dtantsur> yeah, that was the goal
17:59:59 <dtantsur> okay, this was just to make you think about it while I think about office hours ;)
18:00:01 <dtantsur> thanks all!
18:00:05 <dtantsur> #endmeeting ironic