15:01:39 #startmeeting ironic 15:01:40 o/ 15:01:40 Meeting started Mon Dec 3 15:01:39 2018 UTC and is due to finish in 60 minutes. The chair is TheJulia. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:41 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:43 The meeting name has been set to 'ironic' 15:01:45 o/ 15:01:45 oh, that time already 15:01:48 \o 15:01:48 o/ 15:01:50 o/ 15:01:50 o/ 15:01:52 o/ 15:01:58 \o 15:02:00 o/ 15:02:02 yay my first official ironic meeting o/ 15:02:15 I kind of lost track of time since I've been going through outstanding patches and making lists 15:02:29 No, not making lists as in the holiday kind of lists. 15:02:38 * TheJulia loads the agenda 15:02:46 o/ 15:03:01 Our agenda can be found on the wiki as always. 15:03:04 #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting 15:03:14 o/ 15:03:19 o/ 15:03:21 #topic Announcements/Reminders 15:03:55 #Info Following up on the clean/deploy steps action item to schedule a meeting. A doodle was posted. Please submit a time as the available times are coming up soon. 15:04:03 #link https://doodle.com/poll/yan4wyvztf7mpq46 15:05:36 One other announcement that I have. I've proposed kaifeng for ironic-core. The post can be found on the mailing list. Based on the replies I think congratulations is in order. 15:05:58 #info kaifeng has been proposed for ironic-core 15:06:02 ++ 15:06:11 Does anyone have anything they would like to announce? 15:06:57 kaifeng++ congrats! 15:07:12 congrats kaifeng 15:07:27 Out of curiosity, does anyone have vacations coming up that the community should be aware of or that we should focus on your feature code sooner rather than later? 15:07:28 Cograts Kaifeng!! 15:07:28 kaifeng++ thanks for past and future reviews! 15:08:19 TheJulia: I'll be out most of january, exact dates TBD :) 15:08:24 kaifeng: Well deserved :) 15:08:31 jroll: ack ack :) 15:08:46 can be summoned if needed though 15:08:47 I don't know what I should say, but I really grateful for all of you :) 15:09:02 Okay well, if there is nothing else to announce, I think we should proceed directly to subteam status updates. 15:09:13 kaifeng: :) 15:09:22 TheJulia: I'm out since 17 Dec and till Jan 15:09:40 kaifeng: smiling helps :) 15:09:48 dtantsur: ack 15:09:56 #topic subteam status updates 15:09:59 #link https://etherpad.openstack.org/p/IronicWhiteBoard 15:10:14 Starting around line 225 at present 15:12:13 mjturek is not here :( 15:12:14 TheJulia: do we know who is working on the ugprade checker, the TODO (L244)? 15:12:24 hamzy_: o/ 15:12:43 rloo: I believe that was going to fall to me if nobody does anything else on it 15:13:00 TheJulia: so you'd like a volunteer? I'll mention that. 15:13:00 I was going to ask you if you could take a look because I think we have the code and just need to call our sanity checking logic 15:13:21 TheJulia: heh, are we in a hurry for that? I can do it, next year? 15:13:34 just to highlight two big API specs that landed recently: 15:13:34 rloo: next year should be fine 15:13:37 #link http://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/expose-conductor-info.html 15:13:41 TheJulia: okey dokey 15:13:42 #link http://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/allocation-api.html 15:13:59 I haven't seen many reviews on them except for ppl who approved them, so I think I could post the link here 15:14:42 dtantsur: thanks! While not goals, They are asks we've been getting so \o/ 15:14:53 dtantsur: do you want to begin tracking specific patches? 15:15:34 the conductor API is worth tracking, the patches are up 15:15:45 I haven't got any patches for the allocation API, it's on my TODO list for this week 15:15:47 stendulker: Have you heard from shekar regarding the dhcp-less l3 virtual media boot work? 15:16:05 TheJulia: No 15:16:09 :( 15:16:33 TheJulia: But in PTG he was saying the code is ready and he would raise review once his CI is up 15:16:35 #link https://review.openstack.org/#/q/topic:1724474 conductor API patches 15:16:36 TheJulia: wrt IPXE/PXE split (L268). Is it done after that 621416 lands? 15:16:49 etingof: hshiina_: have either of you heard from mkrai? 15:17:06 TheJulia, I do not think so 15:17:09 rloo: I feel like done done really unless we feel we need a ton more docs. :\ 15:17:30 we always feel we need a ton more docs 15:17:36 TheJulia, i don't, either 15:17:49 dtantsur: true-ish... :( 15:17:59 TheJulia: ok, so done w/o that pr. after this meeting, let's delete that then. similar for enhanced checksum support. And thx!!! 15:18:23 Should we add a rolling "things to write docs for" list? 15:18:27 or "more docs" 15:18:37 since we've done the bare minimum, we could do better... 15:19:04 not sure, we don't have a good track record of 'just' writing docs after the fact. 15:19:10 yeah 15:19:27 the most assured way of docs is to not say a feature is complete until the docs are done. 15:19:30 yep 15:19:34 Just so everyone is aware, the neutron events spec merged 15:19:36 indeed 15:19:47 we also need people who are NOT ironic developers to provide consistent feedback 15:19:55 dtantsur: ++ 15:19:56 ++++ for neutron events. (it was on my list to look at this week so even more +++) 15:20:08 #link http://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/synchronize-events-with-neutron.html neutron events spec 15:20:12 rloo: you might want to check-in with vdrok re code 15:20:25 TheJulia: ok 15:20:49 dtantsur: re federation, I feel like it might be good to have another call soon, but maybe after the holidays to further discuss 15:20:57 ++ 15:21:01 after? 15:21:04 yep 15:21:07 okay 15:21:15 and.. I don't think we have a lot of interest in ironic-level federation 15:21:23 so it may be a short call :) 15:21:46 I would mostly concur with that, at the same time I think the folks who feel they need very strong federation will be appearing after the new year to discuss it 15:21:58 * TheJulia hopes 15:22:47 I'm good with status updates. I seem to have a number of emails to send :( 15:23:25 * dtantsur is good as well 15:24:11 On to priorities for the next week? 15:24:18 for whatever reason, i thought some nova spec got merged, that we were interested in 15:24:25 conductor grouping 15:24:43 rloo: I already deleted it from the priority list since it merged early last week 15:24:48 jroll: thx that was it 15:24:54 :) 15:25:04 TheJulia: oh, do we want to leave it on for the code? 15:25:05 Since we're drifting towards priorities :) 15:25:08 (i was looking at the 'Nova deployment state callback...) 15:25:09 #topic Priorities for the coming week 15:25:18 jroll: if code is posted, by all means add it back :) 15:25:23 ok 15:25:27 s/if/when :) 15:25:30 :) 15:25:48 #link https://etherpad.openstack.org/p/IronicWhiteBoard 15:25:57 Starting around line 107 15:26:04 First, time to remove merged things 15:26:18 mgoddard: any plans to update the deploy templates spec this week? 15:26:52 TheJulia: I think you can also remove stuff that only needs to get through the CI (like protected instances) 15:27:11 dtantsur: it helps me track in the mean time :\ 15:27:18 since CI is a mess right now 15:27:18 ah, ok 15:27:34 I've seen messier CI's :) 15:27:56 whoever is adding the neutron backports, I've already got a link for it 15:28:11 line 160 15:29:30 I feel like the CI happier kind of stuff needs higher priorities this week 15:30:21 yeah 15:30:22 We've been seeing a lot of intermittent node speed related failures, timing kind of stuff. :( 15:30:32 that is a *lot* of priority for one week 15:30:47 a lot of it already has +2s 15:31:18 ok 15:32:28 and the ipv6 one is mostly a big release note 15:32:31 :( 15:32:57 I'm good with the list for this week. I went ahead and added a note to the ones that are from last week that are approved that are just being tracked at the moment. 15:33:03 I'll delete them once they merge 15:33:32 can we delete the crossed off nova conductor_group support? (L176) 15:33:44 jroll: ? 15:33:53 we can add back when there's code. 15:33:57 ++ 15:34:01 it definitely is not a weekly priority :P 15:34:03 kill it 15:34:15 :) 15:35:14 Shall we move on to discussion? 15:35:19 we got sushy-tools twice in the list I think 15:35:33 oh, yup 15:35:35 * TheJulia fixes 15:37:07 * TheJulia hears crickets 15:37:23 #topic Discussion 15:37:45 We have one discussion item for today. 15:37:56 Do we continue with coreos ci testing? 15:38:43 my primary reason for keeping it, is that it's the only ramdisk that looks like something folks could run in production 15:38:59 How long has it been broken? 15:39:04 does it still look like that though? 15:39:43 derekh: Not broken as much as they are slightly unreliable 15:39:51 TheJulia: ack, ok 15:39:52 dtantsur: I think so, why not? 15:39:53 In terms of how long they take to unpack 15:40:11 we haven't updated it in a long while, we don't gate against it 15:40:28 derekh: tl;dr if we hit a slow node and it takes 8+ minutes to unpack ipa, then the deployment fails 15:40:54 gothca 15:40:54 They did work on November 26th 15:41:32 I guess I'm wondering if it makes sense for us to keep them working. Every time we try to update to a newer coreos it becomes a bit of a headache as well because the images just seem to grow :( 15:42:09 right, so in my ideal world, we'd have something smaller but production-ready that we test against, probably from DIB 15:42:38 I do concur with jroll, it is the only thing that we have that looks like something someone would run in production. Bifrost does do a Debian dib build, but that is sometimes as unreliable with package downloads as anything else. 15:43:02 we have RDO images, like master ones in https://images.rdoproject.org/master/rdo_trunk/current-tripleo/ 15:43:09 though they're pretty bloated 15:43:25 I think debian-minimal is ~226MB and we could remove some cruft most likely 15:43:47 * TheJulia feels like more thought is required 15:44:07 coreos master is 464MB 15:44:15 wow 15:44:19 not really here, but I think the buildroot idea for IPA was a good one 15:44:42 464MB means.... 2.5-3GB minimum memory footprint?!? 15:44:51 464M WOW it's bigger than RDO images 15:45:11 and RDO images contain a lot of additional stuff (like python-hardware for discovery of everything) 15:45:33 wow coreos that big? 15:45:40 I'd be open to doing CI on the RDO images 15:45:47 I would as well. 15:46:26 Anyone want to volunteer to make the ironic/devstack/lib/ironic changes and try to spin up a job in ironic-python-agent ? 15:46:27 RDO image is 284M .. tar? (Is it not compressed?) 15:47:06 TheJulia: I'll take a look if nobody else is interested 15:47:24 hjensas: it is compressed, but not by much 15:47:32 291.7MB is the ramdisk image size 15:47:46 That is still ~1.5GB of ram 15:47:56 yeah, better 2 15:48:11 derekh: That would be awesome 15:48:23 * TheJulia feels that discussion on this is done 15:48:32 #topic Open Discussion 15:48:43 Does anyone have anything they would like to discuss this morning? 15:49:21 * jroll does not 15:49:26 thank you derekh! 15:49:46 jroll: no prob 15:50:38 Well then, I hope everyone has a wonderful week if we have nothing else to discuss today 15:50:59 * dtantsur has nothing 15:51:43 crickethanks 15:52:04 o/ 15:52:06 #endmeeting