17:00:03 #startmeeting ironic 17:00:04 Meeting started Mon Jul 10 17:00:03 2017 UTC and is due to finish in 60 minutes. The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:06 o/ 17:00:08 The meeting name has been set to 'ironic' 17:00:15 o/ 17:00:16 who is here for the most ironic of the ironic meetings? 17:00:28 o/ 17:00:30 o/ 17:00:35 o/ 17:00:35 o/ 17:00:43 o/ 17:00:43 o/ 17:01:06 hi everyone! and welcome :) 17:01:13 our agenda as usual can be found at: 17:01:15 o/ 17:01:19 #link https://wiki.openstack.org/wiki/Meetings/Ironic 17:01:30 o/ 17:01:31 o/ 17:02:09 #topic Announcements / Reminder 17:02:21 #info Release request for ironicclient 1.14 with BFV API submitted (thanks TheJulia) 17:02:37 ++ 17:02:46 this is amazing progress. you folks seem to review things faster, when I'm on PTO. I need to go on PTO more often ;) 17:02:54 No!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 17:02:57 :) 17:02:59 Please no! 17:03:00 haha 17:03:09 :) 17:03:21 dtantsur: it is really the other way around, we review faster when you are here 17:03:25 #info Non-client libraries deadline next week. This affects ironic-lib and sushy. 17:03:46 I know there are certain patches for sushy in flight 17:04:00 wrt ^^, does anyone have a list of patches that we 'need' to get landed? 17:04:33 nothing really "needed" to my best knowledge 17:04:50 one patch supporting redfish session is quite important, but I guess it also needs an ironic RFE 17:05:04 #link https://review.openstack.org/#/c/471942/ 17:05:13 dtantsur: ok. also, how (or maybe it doesn't) does the neutron plugin fit, is it considered a non-client library? 17:05:27 good question, I'm not sure 17:05:59 we haven't had a single release for it yet 17:06:24 :\ 17:06:25 dtantsur: we need to find out, decide quickly whether we want that in pike or not then. 17:06:57 sambetts, vdrok ^^ 17:07:10 networking-ovn has type "other" 17:07:20 I guess we're similar, so this should not affect us 17:07:44 dtantsur: ok, that makes sense 17:08:11 dtantsur: sorry, wrt the sushy session, it is marked high importance. is sushy driver usable w/o this change? 17:08:41 dtantsur: well, we can take it offline. 17:08:58 rloo: it is, but only on hw not using these session 17:09:05 using plain HTTP auth 17:09:21 any other announcements or reminders? 17:09:40 PTG Hotel block closes on 8/20 17:09:49 oh, yeah. PTG IS COMING!! 17:10:14 the good time to ask for travel approval is yesterday 17:10:19 and just another reminder, nova feature freeze is week of jul 24 17:10:32 rloo: this is pike-3, right? 17:10:41 dtantsur: yup, pike-3 17:11:04 ok, if the ironicclient release is done soon (it should), we'll be able to push on the g-r and nova changes 17:11:12 TheJulia: link for the g-r change handy? 17:11:17 o/ 17:11:26 https://review.openstack.org/#/c/482146/ 17:11:40 #link https://review.openstack.org/#/c/482146/ 17:11:54 dtantsur: wrt travel support, round one was july 2; round 2 is aug 6: https://www.openstack.org/ptg#tab_travel 17:12:24 rloo: right, I was referring to asking for approval from employers though 17:12:32 dtantsur: ha ha 17:12:55 anyway :) good points. anything else to add? 17:13:20 * jlvillal still waiting to find out if he will get approved. Has submitted request awhile ago. 17:13:57 #topic Review subteam status reports (capped at ten minutes) 17:14:12 #link https://etherpad.openstack.org/p/IronicWhiteBoard starting with line 107 17:14:15 Is it possible to attend PTG remotely via call? 17:14:42 wanyen: it may be, but it's not a question to us here, but rather to the foundation. I doubt, however, that it would be a pleasant experience :( 17:15:06 better than not be able to attend at all 17:15:09 e.g. it will require to use microphones for everyone. I'm not sure how practical it is. we did not use them the last time.. 17:16:22 TheJulia: any reason not to approve https://review.openstack.org/#/c/463908/ and https://review.openstack.org/#/c/463972/ ? 17:16:40 dtantsur: None 17:16:54 then I encourage you to do so ;) 17:16:57 dtantsur: I didn't do it myself since I rebased them last and had to do it locally 17:17:22 I can do it, but with 3x +2 it's not critical IMO 17:17:37 okay, the diff shows I didn't change content 17:17:40 * TheJulia pushes the buttons 17:18:02 And pushed. 17:18:06 * TheJulia updates etherpads 17:18:09 \o/ 17:18:25 at least the ironic part is moving quite well 17:18:38 folks, in case you missed it, i am planning on removing the 'feature parity between two CLIs' subteam report after this week's meeting, unless people want it to stay 17:18:57 rloo: I guess it's right. we've achieved the parity. 17:19:08 I'd like to also remove the Redfish topic 17:19:46 folks, I'd like to highlight https://review.openstack.org/#/c/442153/. the pike-3 is also a client deadline, and this has to go in in Pike for us to change the version in Queens 17:19:56 #link https://review.openstack.org/#/c/442153/ 17:20:05 dtantsur: ++ wrt redfish 17:20:36 dtantsur: i think 442153 can be one of this week's priorities 17:20:43 I hope so 17:20:50 dtantsur: "should be" easy (w/o looking at it) 17:21:07 it's not really hard indeed 17:21:22 Yeah, looks simple 17:21:34 dtantsur: also, wrt your spec on classic driver deprecation -- do we need to discuss anything? julia had some comments 17:21:52 I'm just out of PTO, I need to refresh on it first... 17:22:00 what was the concern, the timing? 17:22:09 dtantsur: https://review.openstack.org/#/c/464046/. 17:22:12 I don't think we require any serious actions for Pike, do we? 17:22:36 I think it was just communication/expectation setting at this point 17:22:41 I'm good otherwise, I think 17:22:58 dtantsur: oh, i also had a question forgot about it. what if vendors don't provide hardware types that are analagous to classic. 17:23:08 * TheJulia thinks it has been a week since she looked 17:23:11 dtantsur: i think we know but it isn't documented in the spec. that's all. 17:23:24 rloo: well, the spec implicitly gives the answer. one day the classic drivers are removed, no matter if they have analogs or not. 17:23:59 dtantsur: yeah, i know it implicitly. i just want to make sure it is 'clear enough' for the hardware vendor folks out there, that's all. 17:24:10 I suspect we can solve it on the spec review itself. or we can discuss during the open discussion, wdyt? 17:24:25 dtantsur: if people think it is obvious or whatever, then i think we can +A it. 17:24:44 dtantsur: yup, was just trying to fasttrack it since the spec has been around for awhile. 17:25:05 if it's not critical for merging the spec, I can follow up with this 17:25:18 as long as TheJulia's concerns are also / can also be resolved 17:25:33 We can always follow-up imho 17:25:51 dtantsur: ok, i've brought it up, hardware vendors should be reading the minutes of this, so it is documented and i am fine to move on 17:25:54 sambetts or someone: we probably need a volunteer to nag nova folks about https://review.openstack.org/#/c/419975/ 17:26:03 * TheJulia pulls it up 17:26:08 thanks rloo! 17:26:25 meanwhile, anything else on the statuses? we're past 10 minutes (well, as always) 17:27:38 w/r/t the future of classic drivers, I just changed my vote to +2 17:27:45 thanks! 17:28:07 #topic Deciding on priorities for the coming week 17:28:08 * rloo ready to move on 17:28:31 we should get the rest of bfv client patches merged 17:28:33 dtantsur: I can do some pestering 17:28:33 so, the docs re-org is in progress, and is of very high priority. I guess we keep it 17:28:40 thanks sambetts! 17:29:05 rloo: right. TheJulia, what's the next for BFV? 17:29:10 speaking of the docs, I've got the IPA repo open on my other screen right now. 17:29:56 Client largely, aside from that the nova change. I suspect they are going to want us to make sure our tempest test is ready to rock and roll as well 17:30:14 TheJulia: then let's take the devstack and tempest 17:30:29 or do they need the OSC commands to be finished? 17:30:55 We need to get the OSC commands out the door and released if memory serves before we can take on devstack 17:31:49 oh, so another release 17:32:06 well, we'll need a pike-3 release anyway 17:32:08 so 17:32:22 please take a look at the list. we have quite a few items there, but many are simple 17:33:03 dunno how many of the rolling upgrades patches we want. rloo? 17:33:23 dtantsur: what i see there is enough 17:33:28 dtantsur: i mean, in this week's priorities 17:33:30 not sure how much more effort we need to put into refactoring the docs, the bulk of the work moving the content around is done, but the landing page needs revamping 17:33:36 dtantsur: i'll be happy if one gets reviewed :) 17:34:30 sambetts: I'd start with merely moving them. I just got an email about ironic-lib, virtualbmc, sushy, sushy-tools and molteniron work not started yet 17:34:33 sambetts: i think what we need to do is make sure the high level things are good. like contributor/*, separating developer docs from other contributor docs. 17:34:47 .. and bifrost and ironic-inspector-client 17:35:11 dtantsur: I can start looking at the sub-projects 17:35:24 dtantsur: oh, since ironic-lib & sushy freeze is next week, we should get those done this week 17:35:33 thanks sambetts! I think milan is looking at ironic-inspector-client now 17:35:36 sambetts: You might want to double check with the list since someone signed up to take care of some of the other projects 17:35:38 rloo: right 17:35:40 dtantsur, ack 17:35:53 dtantsur, sushy? 17:36:04 TheJulia: oh great, I'll get on review patrol for those, I think they're all linked of the whiteboard now 17:36:05 milan: wanna do it? ;) 17:36:12 heh :D 17:36:33 yeah I can add it to the doto list 17:36:54 sambetts: Only some of them had been submitted from what I've seen, fwiw. I'll be in the air for a good chunk of time tomorrow, so I'll likely have time to re-shuffle all of bifrost's docs around then. 17:37:32 milan: put your name down at https://etherpad.openstack.org/p/doc-migration-tracking 17:37:38 milan: L203 17:37:54 rloo, thx 17:37:58 * milan adding himself 17:38:34 that leaves sushy-tools, molteniron & virtualbmc 17:38:52 * dtantsur wonders why we even have docs for sushy-tools... 17:39:11 sushy-tools and virtualbmc only have dev docs, should be easy 17:39:18 anyone want to pick them? etingof? ;) 17:39:22 any volunteers for those? ^^ 17:39:30 * milan unsure about tuanlua in the projects but I guess it's OK if I patch 17:39:32 I can do it 17:39:58 * etingof is too slow 17:40:04 thx sambetts! 17:40:05 * dtantsur leaves an opportunity for non-cores to jump in 17:40:13 etingof: not too slow, you can do it if you'd like 17:40:13 etingof: feel free to take it 17:40:18 :D 17:40:29 sambetts: if you could guide etingof with reviews and advice instead, that would be awesome 17:40:31 :) 17:40:37 absolutly :) 17:40:45 ok, I can try to do that thing 17:41:17 molteniron.. who is driving it nowadays? 17:41:29 thx etingof! 17:41:37 unsure, jlvillal? 17:41:44 I'm drawing a blank on the name 17:41:54 isn't molteniron from ibm folks? 17:41:55 krtaylor_: ? 17:42:02 It is from IBM folk 17:42:08 yeah, one of kurt's people i think 17:42:37 anyway, can take that question off line? :) 17:42:38 sambetts: Not me. krtaylor_ was molteniron 17:42:43 well, I guess we can just migrate it and see. assuming we have core rights on it :) 17:42:52 * krtaylor_ pops up 17:42:57 mmm, are we done with the priorities? 17:43:05 I guess we can discuss the docs a bit later 17:43:06 * krtaylor_ reads scrollback 17:43:49 #topic REST API to list volume connectors & volume targets returns node_uuid, but port & portgroup doesn't 17:43:54 rloo: the mic is yours 17:44:22 so in ironic's REST API, if you do a port-list or portgroup-list, you do not get node_uuid. only if you ask for that field explicitly, or detail=True 17:44:38 i just noticed that for the new volume connector * volume target resources, you will get the node_uuid. 17:44:54 i love consistency, so wondered if we are ok with this inconsistency 17:44:56 * TheJulia thinks this is a good thing, fwiw 17:45:19 I always disliked that we don't return node_uuid for port listing FWIW 17:45:24 I would <3 to see it on ports, personally 17:45:30 I'd rather fix the ports 17:45:35 i think TheJulia thinks it is a bug that it isn't avail for ports & port groups? 17:45:47 dtantsur: only ports, or portgroups too? 17:45:50 well, I'd not call it a bug. rather an easy RFE 17:45:52 rloo: both 17:45:54 i'm fine either way :-) 17:46:07 ok, done. thx! (I mean, done, we've decided.) 17:46:28 i'll take care of the paperwork later :) 17:46:35 any objections to returning node_uuid on port(|group)s listing with detail=False? 17:46:53 3.. 17:46:55 2.. 17:46:57 1.. 17:47:00 done, thanks rloo! :D 17:47:07 :p 17:47:09 dtantsur: maybe some #decided thing? 17:47:36 #agreed We need to update port and portgroup listing API to include node_uuid by default (not only with detail=True) 17:47:49 #topic Open discussion 17:48:06 13 minutes for any bikesheding you like! 17:48:08 dtantsur: we should, soon, next 1-2 weeks? review our pike priorities 17:48:19 dtantsur: and focus ... 17:48:28 rloo: I would do it today, but I'm kinda busy this week 17:48:40 dtantsur: next monday meeting? 17:48:42 so yeah, totally. certain things are obviously moving too slowly 17:48:58 a voice meeting next week? or just as part of the next IRC meeting? 17:49:18 i think part of next irc meeting. what do others think? 17:49:20 dtantsur: I've implemented the flag based solution for agent_pxe_oneview driver we were discussing https://review.openstack.org/#/c/482209 17:49:26 For people who want a break. https://review.openstack.org/447310 ( hacking check to require autospec=something) and a WIP patch of mine: https://review.openstack.org/481316 for keyword argument enforcing. 17:49:32 Very low priority though!!!! 17:49:32 dtantsur: waiting for our UFCG OneView CI :) 17:49:46 please review https://review.openstack.org/#/c/422572/ 17:49:56 wanyen: Why? What is it? 17:49:59 dtantsur: no nays so i'll put on agenda for next week :-) 17:50:09 xavierr: ack, thanks 17:50:24 rloo: sure! thanks for the idea 17:51:23 Hi, would like to bring up the dmidecode patch for some reviews: https://review.openstack.org/#/c/435573/ 17:51:24 rloo: On the port/portgroup thing. Is there an owner for that? Or is it up for grabs? Will you create a bug/RFE? 17:51:44 jlvillal: i will create an RFE 17:51:49 rloo: Thanks. 17:51:59 sorry it is nothing to do with your topics. It's a patch hanging there for a long time and needs review 17:52:09 rama_y's patch has one +2 (mine). As a note. 17:52:11 rloo: this is going to be a low-hanging-fruit, so I guess we'll find someone 17:52:38 dtantsur: yup, i'm not worried about finding someone for that. 17:53:11 so many patches, so little time and few reviewers... 17:53:13 wanyen, rama_y, I'll try to, but I'm half-busy this week 17:53:30 dtantsur: I wonder the spec about fault, how is that going 17:53:43 Thanks dtantsur! 17:53:49 dtantsur, thanks! iLo team really like to get this one in Pike. 17:53:50 kaifeng: specific faults? it's stuck IIRC 17:54:10 wanyen: don't know how much it would help, but if your patches of interest showed that your folks reviewed them, it would help. 17:54:18 folks, I have an unusual request 17:54:20 wanyen: oh, i guess it would help :) 17:54:21 Same, I'm also traveling part of this week. I kind of wonder if it would help if we got a list of patches from driver teams that they need reviewed to make it into pike. 17:54:41 dtantsur: the spec about ironic detect & recover from some fault 17:54:42 Since, we can't read minds and all 17:54:46 I'm holding an internan deep dive on ironic tomorrow. could someone please take a look at my slides (a lot of them) in case I'm doing something stupid? http://dtantsur.github.io/talks/pike-ironic-deploy-deep-dive/#/ 17:54:47 dtantsur: Hawaiian shirt Friday? 17:54:57 Ah different unusual request 17:54:58 jlvillal++ LOL 17:55:01 rloo: yes. we all reviewed them alreay. We got a +2 last week but due to mege conflict, we rebased it and need more rvew to get 2 +2. 17:55:10 jlvillal: Shouldn't that be daily ;) 17:55:16 I spent the last 2.5 days writing it, and in the end I was tired 17:55:17 :) 17:55:37 wanyen: i just looked at that link you provided, unfortunately, w/o at first glance seeing that there are reviews on it... 17:55:38 if anyone has any comments, please send me them by tomorrow morning 17:55:49 wanyen: all i see is that jenkins is happy 17:55:56 dtantsur: I'll take a look today. 17:56:03 wanyen: i don't have time to spend 5-10 minutes just trying to figure out hwat the state of a patch is. 17:56:07 thanks jlvillal! 17:56:28 dtantsur: I have to learn to press right arrow and then when I see down arrow press it... 17:56:32 ^^ that goes to everyone's patches. we need to try to communicate/help as much as possible since again, so many patches, so little time... etc. maybe i will make up a song... 17:56:38 jlvillal: just press space all the time 17:56:40 :) 17:56:43 dtantsur: At first I did right arrow... and finished very quickly :) 17:56:50 LOL 17:56:57 rloo: ++ 17:57:18 rloo, we just re-based the patch. It's been patch 21. 17:57:35 rloo: I recommend to the tune of Too Much Time On My Hands by Styx ;) 17:57:37 jlvillal: I agree.. I did the same and see the presentation got over in no time 17:57:40 any rebase requires re-review, unfortunately 17:57:54 wanyen, i *only* looked at the top "fold"/part of that url. i am not going to scroll down and spend any more seconds on it. 17:58:27 shobhar: :) 17:58:50 jlvillal, shobhar, I've been using reveal.js for a while, and only learned about using space yesterday :D 17:58:58 also, i realize that folks may be new/not feel comfortable reviewing 'core' features cuz of whatever, but wrt driver patches, we're all in the same boat there, it would be great if driver teams could help review *other* driver team's patches. 17:59:23 * dtantsur points that reviewing core features is a great way to learn the project 17:59:37 One minute warning 17:59:51 yeah, I guess we can continue in the channel (and I have to run) 17:59:56 thanks everyone! 18:00:03 #endmeeting