17:00:23 #startmeeting ironic 17:00:23 Hey 17:00:23 Meeting started Mon Feb 27 17:00:23 2017 UTC and is due to finish in 60 minutes. The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:25 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:26 o/ 17:00:26 o/ 17:00:27 The meeting name has been set to 'ironic' 17:00:28 hello :) 17:00:29 o/ 17:00:30 o/ 17:00:32 o/ 17:00:33 \o/ 17:00:34 o/ 17:00:38 o/ 17:00:38 o/ 17:00:38 #chair jroll TheJulia 17:00:39 Current chairs: TheJulia dtantsur jroll 17:00:42 \o 17:00:49 \o 17:00:50 I may disappear due to horrible internet, please back me up :) 17:00:53 oh my, I'm special 17:00:59 o/ 17:01:03 o/ 17:01:05 o/ 17:01:07 o/ 17:01:18 #topic Announcements / Reminders 17:01:19 o/ 17:01:26 We've had an awesome PTG, thanks all for attending! 17:01:44 o/ 17:01:44 \o/ 17:01:48 I plan on writing a summary to the ML, but personal summaries from other folks are also welcome. 17:01:52 o/ 17:01:53 big thanks to mturek? for taking great notes! 17:01:57 #action dtantsur to write a PTG summary 17:02:04 :-D np rloo 17:02:09 oh yeah, mjturek++ 17:02:18 sorry I couldn't make the last day 17:02:35 mjturek, there was no much to write down, don't worry :) 17:02:41 :) 17:03:06 any other announcements / reminders / saying thanks you? :) 17:03:18 it was an awesome ptg, thank you all 17:03:22 and +100000 for mjturek 17:04:03 ok, moving on 17:04:11 #topic Review subteam status reports 17:04:26 #link https://etherpad.openstack.org/p/IronicWhiteBoard line 76 17:05:05 I don't expect a lot of updates, except for things discussed on the PTG, but please check what we have 17:05:10 * jroll would be surprised to see updates 17:05:28 * jlvillal is full of surprises ;) 17:05:43 Or at least three lines of updates 17:06:01 jlvillal & vasyl, great work getting multinode grenade working (or close to it) 17:06:04 more than nothing, you see :) 17:06:08 ++ 17:06:31 rloo: Thanks :) I am seeing the light at the end of the tunnel. Have to figure out the non-OSIC cluster failures still 17:06:48 +1 for vsaienk0 help! 17:06:52 someone needs to update the driver comp section... 17:06:54 ++ nice work 17:07:18 rloo, it's more or less up-to-date apparently 17:07:24 yeah, it's pretty close 17:07:39 once I go over notes from PTG I'm happy to update with pike plans and such 17:07:48 ditto 17:08:04 jroll: ok, i'll update with what you just said 17:08:10 heh 17:08:19 * dtantsur gives folks a few more minutes to read/update 17:08:30 :) 17:09:35 dtantsur: what was decided about the soft power off stuff. I forgot. 17:09:50 dtantsur: was it going to be a separate spec? 17:10:07 just tracked separately on the whiteboard, I think 17:10:19 the spec is still valid 17:10:37 dtantsur: ok, i'll wait til you do the priorities patch... 17:11:01 yes, please give me a day or two :) it's a hard recovery 17:11:18 dtantsur: wrt the bugs... didn't folks decide on a process for doing bug triage? 17:11:38 dtantsur: something to do with volunteers? 17:11:58 a process is to ask for volunteers every week to drive the bug triaging 17:12:11 which is part of the next topic in the agenda 17:12:22 * dtantsur thinks of moving to it 17:12:58 done with the updates? 17:13:05 * rloo done 17:13:17 #topic Deciding on priorities for the coming week 17:13:27 this was not an explicit topic before, but we anyway did it every time 17:13:35 priority one: put head back on straight 17:13:44 one and only, I would say :) 17:13:48 o/ 17:14:06 let's try to finish with the priorities by the next meeting, if possible 17:14:21 there are two subitems proposed for this meeting topic 17:14:30 +1, and get the subteam list updated 17:15:06 "Stuck specs for current and potential priorities" which is about useful stuff that does not get reviewed. we didn't finally agree on this item, but I added it anyway, mostly to continue thinking about it 17:15:31 "Appointing a bug liaison for the next week" this is what rloo was asking for (I know people don't like liaison in this context, just change to a better word) 17:15:51 now let's if it works :) is there anyway who wants to keep an eye on the bug list this week? 17:15:56 dtantsur: was this for volunteer to do bug triaging? 17:16:32 yep 17:17:31 I can do it this week 17:17:36 apart from today 17:17:39 \o/ thanks vdrok 17:17:45 Thank you vdrok 17:17:56 #action vdrok to drive bug triaging this week 17:18:13 side note: this does not mean that I stop doing it, but now it gets moar eyes 17:18:16 nice, thanks vdrok 17:18:27 another side note? the volunteer does NOT have to be core? 17:18:33 no, not necessary 17:18:35 should I add my name on the whiteboard? 17:18:41 vdrok, yes please 17:18:56 rloo, just a good understanding of the project is enough IMO 17:19:18 dtantsur: ++ 17:19:33 dtantsur: to be clear, are you going to continue to update the bug stats in the whiteboard? 17:19:47 yes, I will still do it 17:19:59 dtantsur: thx 17:20:36 okay, I guess we're good to move on 17:20:49 #topic OSIC Ops adopt-a-project 17:20:51 well 17:20:55 I don't see a name, who was it? 17:21:01 so we're good with the weeks' priorities? 17:21:11 (this is me asking mrhillsman to bring this topic, btw) 17:21:15 jroll, the priorities: get some priorities :) 17:21:21 heh 17:21:29 * jroll doesn't see that in the list 17:21:32 mrhillsman: around? 17:21:49 dtantsur: i'm a bit confused. is the 'deciding on prioriites..' stuff in the agenda a one-time (today) thing, or is that a new thing that we will look at weekly? 17:22:08 "review specs" priority item was removed; was that intentional? 17:22:13 rloo, weekly. that's what we've done for more than a cycle, just without an explicit item 17:22:31 Makes sense to me. 17:22:34 mariojv, it was. while it's always important, it's not a priority item right this second. 17:22:34 dtantsur: it doesn't make sense to have the 'appointing a bug liaison' under that? 17:22:43 ack, thx 17:22:55 rloo, this is planning for next week, but dunno 17:23:09 dtantsur: planning isn't priorities 17:23:32 rloo, how badly is this inconsistency going to affect everyone's work? 17:24:00 dtantsur: dunno, i can't answer for everyone. just seems ... odd, so i'm mentioning it. 17:24:29 I thought the end goal was a weekly bug triage change (with less than weekly while we ramp up) 17:25:10 it's weird but we can deal with it later 17:25:11 not necessary weekly, but yes. asking for volunteers every week 17:25:22 o/ 17:25:22 let's move it to the channel or Just Do It (tm) please :) 17:25:24 last call for mrhillsman for the topic we're currently on 17:25:24 sorry jroll 17:25:25 could we put down some more priorities for this week? eg, review tags? 17:25:25 mrhillsman, o/ 17:25:27 ah there we go 17:25:37 #link https://etherpad.openstack.org/p/ops-adopt-a-project-pike 17:25:51 we have rounded out what we think from an ops view but it is open 17:26:13 rloo: let's go back to it in open discussion in a few 17:26:16 it boils down to us basically wanting to establish loop between ironic and ops team in osic to help as we can 17:27:17 mrhillsman: mind explaining a bit more on what y'all are doing (and/or how we can help?) 17:27:19 mrhillsman: what does '2. supported by 7 or more SDKs' mean? 17:27:25 we have time for pike to determine the work so again, nothing is set in stone 17:27:33 ++ to rloo's question, it caught my attention too 17:27:42 rloo: i grabbed that from the project navigator 17:27:54 yeah it has this item 17:28:01 https://wiki.openstack.org/wiki/SDKs 17:28:05 i assume ^ 17:28:05 the "maturity" tag 17:28:11 mrhillsman: i don't know if ironic is supported by 7 or more sdks 17:28:12 #link https://www.openstack.org/software/releases/ocata/components/ironic 17:28:17 for what the project navigator is 17:28:39 it seems like only by 2 - PythonOpenStackSDK and OSC 17:28:46 oh wow great face on there 17:29:09 jroll: you mean cool dtantsur, right? 17:29:21 >.> 17:29:31 so we will ideally install ironic 17:29:36 according to the documentation 17:29:37 lol 17:29:48 if we find gaps/issues/whatever, file bugs/patches 17:30:10 woot 17:30:18 we will try to scale it as much as we can, again, bugs/issues/whatever 17:30:18 sweet 17:30:28 we could use help on testing rolling upgrades at some point :) 17:30:36 if there are things ironic team would like us to test, we are welcome to it 17:30:38 mrhillsman, this is fantastic! what kind of help you need from us if any? 17:30:57 * jlvillal likes :) 17:31:06 so on that etherpad, we just right now want to know if you had ops team that could focus on just your project 17:31:09 yeah, any help with upgrades and docs is particularly welcome 17:31:10 what would you expect from them 17:31:31 if just upgrades and docs, that is ok, we will try to do as much or as little as asked 17:31:32 I agree, docs will be a big one 17:31:39 probably multitenant networking 17:31:50 feel free to just dump stuff on there 17:31:57 mrhillsman: I'm also going to start writing a sort of reference architecture, that includes kvm+ironic in the same region, I might drag you into that 17:32:04 as it... isn't trivial 17:32:06 we have like 2-3 weeks plus the ops midcycle to gather intel and work out a plan 17:32:12 mrhillsman: do you also help with openstack gate/CI? I mean improving ironic tests in the gate? 17:32:18 sounds good jroll 17:32:36 jroll, just wanted to mention the reference architecture :) 17:32:38 jroll: I love the idea 17:32:39 * dtantsur types too slowly 17:32:59 rloo nothing has been decided, we are gathering info right now and will define our actual work in a couple weeks after the ops midcycle 17:33:08 and we hope to get others to participate as well 17:33:14 to increase adoption 17:33:16 and feedback 17:33:19 to your team 17:33:20 cool 17:33:39 #action ironic team to put some operations requests in https://etherpad.openstack.org/p/ops-adopt-a-project-pike 17:33:41 sound good? 17:33:41 mrhillsman: that'd be great 17:33:45 * dtantsur needs to add ironic to his pet SDK side-project :D 17:33:48 perfect jroll 17:34:20 mrhillsman: to be clear, where in that etherpad do you want us to add stuff? under the 'Ironic' part? 17:34:22 i would say keeps the ops midcycle in mind as well in case you can think of anything relevant to that audience 17:34:24 adding etherpad 17:34:37 sure, that works 17:34:45 we can make it all pretty and organized later 17:35:19 this is awesome, thanks mrhillsman 17:35:39 yes, thank you! 17:35:39 thank you all for entertaining us :) 17:35:52 looking forward to working together 17:36:00 that was all i had 17:36:03 thanks again jroll 17:36:18 any questions for mrhillsman before we move on? 17:36:33 no problem :) 17:37:25 ok, we don't have RFE prepared for review, I guess, so 17:37:37 #topic Open discussion 17:37:47 bring whatever you have :) 17:38:22 * jroll has nothing 17:38:32 unless we want to come back to weekly priorities 17:38:40 For those of us who weren't at PTG. 17:38:51 Curious on the movement of tempest plugin to its own repo. 17:38:51 I'll repeat my comment from the channel about priorities: we don't have Pike priorities, and I'm not convinced people have enough time to really jump on anything this week 17:38:55 Is that something we want to do? 17:39:02 jlvillal, the same state, but yes, we want 17:39:05 * TheJulia has nothing, nor brain cells to say much else 17:39:12 dtantsur: Thanks. 17:39:17 dtantsur: ah right, I did see that 17:39:22 jlvillal, they still target Queens, and the same people still disagree with that 17:39:23 There was an email from QA requesting to keep manage.py from tempest.scenario 17:39:24 http://lists.openstack.org/pipermail/openstack-dev/2017-February/112938.html 17:39:35 s/manage.py/manager.py 17:39:53 oh 17:39:56 oh ffs 17:40:06 it seems manager.py is going under refactoring/changes, and qa wants to minimize job gate breaks 17:40:09 that's the worst way to deal with their problems there 17:40:19 dtantsur: by the way, it is only monday, is something going on that folks won't have time, or do you mean that the folks that went to the ptg need to get back into work mode, etc? 17:40:43 rloo, just the latter 17:41:04 dtantsur: gotcha! 17:41:09 :) 17:41:25 so, I will take create an ironic bug to summarize the QA request 17:42:00 I guess we don't have many options here, do we? 17:42:08 the tempest thing? 17:42:11 Not really 17:42:26 yep 17:42:35 we've been using an unstable interface, apparently 17:42:43 I don't see how this helps openstack as a whole 17:42:52 we're just going to end with a bunch of copy pasted code around 17:42:59 but this isn't the right venue for that argument 17:43:12 * dtantsur thinks of reviving oslo-incubator 17:43:29 * jroll hopes that we can maneuver to just use the stable API and stop importing manager.py 17:43:34 isn't it just copy/paste until they 'formalize' the new interfaces or whatever? 17:43:49 anyway, thx solio for bringing it up etc 17:43:50 jroll: yeah, the suggestion is that every project keeps a copy of manager.py around 17:43:58 rloo, it is 17:44:11 'until' never happens 17:44:13 ever 17:44:20 looking at the source code, we may use quite a lot of it.. 17:44:27 * jroll stops talking, goes to be angry elsewhere 17:44:32 jroll: you're so pessimistic! ha ha. 17:44:42 I have a quick question, wrt to my patch https://review.openstack.org/#/c/403910/ do we want to allow logical names in PATCH call for port as well? 17:45:04 rloo: I have 3 year old TODOs if you want me to prove my point :) 17:45:17 jroll: i hear you ;) 17:45:43 aNuposic, if you mean node names, likely yes 17:46:00 dtantsur, yes node names 17:46:04 cool 17:46:26 as to the tempest issue, I don't see a lot of choice. I welcome someone to research the possibility of using only stable interfaces, but it does not look easy at first glance 17:46:45 yeah, I'm just yelling to yell, it's monday 17:46:54 I guess we'll have to do this "temporary" moving manager.py to our code (with potentially getting rid of everything we don't use) 17:47:32 I suspect if tempest team finds it hard to refactor, we will too, but we can try 17:48:12 well, we don't have to make every project happy in this case :) 17:48:31 so e.g. we may strip bits obviously belonging to other projects (e.g. base classes we don't inherit) 17:48:38 soliosg: Do you want to work on the tempest issue? 17:48:39 right 17:49:11 jlvillal: I can take a look to see what it'll require 17:49:20 soliosg: Thanks! 17:49:34 #action soliosg to look into problem stated at http://lists.openstack.org/pipermail/openstack-dev/2017-February/112938.html 17:49:56 thanks for mentioning it, much appreciated! 17:50:05 * dtantsur is still not even close to finishing his email backlog 17:50:24 10 minutes 17:50:34 * rloo wonders what is wrong with dtantsur's delete key 17:50:46 heh 17:50:48 hehe 17:51:08 Or a pythons script using an imap client.... >.> 17:51:24 s/pythons/python/ 17:51:40 moar pythons :) 17:51:57 dtantsur: the day I found the "mark folder as read" key in Thunderbird was a great day indeed 17:52:08 mgould, if it was so easy for me :D 17:52:16 anyway, does anyone has anything else (too many "any" in one sentence)? 17:52:54 crickets 17:52:59 ^ 17:53:07 TheJulia: machine-learning system assigning a probability that each email is worth reading 17:53:17 mgould: "return 0" 17:53:22 yeah 17:53:26 mgould: what troll said 17:53:28 err 17:53:29 WOW 17:53:31 troll LOL 17:53:32 jroll 17:53:33 * jroll ragequits 17:53:37 * TheJulia kicks autocorrect 17:53:38 :-) 17:53:44 I think we can wrap up on this note 17:53:46 more crickets 17:53:52 thanks everyone 17:53:53 jroll: sorry about that, autocorrect seems to hate me today 17:54:02 TheJulia: you know I don't care :) 17:54:02 thx dtantsur 17:54:06 either that or it hates jroll 17:54:13 * jroll points out he has the nick 'tjroll' registered 17:54:17 thanks all 17:54:18 :) 17:54:21 o/ 17:54:21 88 17:54:26 #endmeeting