15:00:01 #startmeeting ironic 15:00:02 Meeting started Mon Jun 11 15:00:01 2018 UTC and is due to finish in 60 minutes. The chair is dtantsur. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:06 The meeting name has been set to 'ironic' 15:00:12 o/ 15:00:13 * dtantsur waves 15:00:14 o/ 15:00:16 o/ 15:00:17 o/ 15:00:19 \o 15:00:24 o/ 15:00:27 o/ 15:00:33 #link https://wiki.openstack.org/wiki/Meetings/Ironic our agenda for today 15:00:37 looks pretty empty :) 15:00:40 o/ 15:01:01 #Announcements / Reminder 15:01:16 #info Some releases to be requested this week 15:01:24 I don't have a full list, will check what is needed 15:01:31 TheJulia: anything else to announce? 15:01:39 #chair TheJulia 15:01:40 Current chairs: TheJulia dtantsur 15:02:13 * dtantsur assumes Julia is fighting the visa monster right as we speak 15:02:17 Not that I have on my mind at the moment, aside from intent to try and release some things this week 15:02:27 cool :) anyone anything else? 15:02:50 S/monster/hydra/ 15:02:55 true 15:02:59 #topic Review action items from previous meeting 15:03:09 #link http://eavesdrop.openstack.org/meetings/ironic/2018/ironic.2018-06-04-15.00.html 15:03:16 no action items apparently? 15:03:23 we did good! 15:03:28 indeed! 15:03:29 can i get reviews on https://review.openstack.org/#/c/568616/ again? i changed one deprecation bug 15:03:30 patch 568616 - ironic-tempest-plugin - Add manual clean step ironic standalone test 15:03:30 yeah, was a short meeting 15:03:45 yolanda: sure, after the meeting 15:03:49 #topic Review subteam status reports 15:03:52 doh sorry 15:04:03 #link https://etherpad.openstack.org/p/IronicWhiteBoard starting with line 164 15:04:50 TheJulia: when the hydra is defeated, please elaborate on line 206 there 15:04:59 "Operators are requesting network specific documentation" 15:05:13 * TheJulia sharpens sword 15:05:19 heh 15:06:59 TheJulia: can we remove the "Migrate to Storyboard" topic? while it's far from being perfect, the migration is done, we even have some basic bug stats 15:07:00 wrt api versioning with nova. L230? Dmitry asked a question in patch 557850. Do we need to discuss? 15:07:00 Derek Higgins proposed openstack/ironic master: WIP: Testing a basic zuulv3 job https://review.openstack.org/573244 15:07:01 https://review.openstack.org/#/c/557850/ - python-ironicclient - Support per-call version: set_provision_state 15:07:36 dtantsur: works for me 15:07:37 rloo: good idea 15:07:51 dtantsur: i didn't realize it was done. just update it saying it is done, (vs 'in progress') 15:08:14 I’ve not seen the the question on the patch 15:08:28 > After working with openstacsdk for some time and talking to Monty about all this stuff, I'm no longer sure this is the direction we should take. Can someone remind me why we cannot just deduce the correct microversion from the input? 15:08:53 let's bring this on the open discussion, which is quite soon? 15:09:00 sure 15:09:04 dtantsur: wrt bugs then, L167. is that as of today? 15:09:29 rloo: yep 15:09:36 I was out for some time, so the diff is with May 15:09:50 * out last Monday, I guess 15:09:56 ++ to open discussion 15:10:00 241 untriaged out of 283? 15:10:01 dtantsur: ok thx. i was ignoring those cuz i didn't realize they were actually updated! 15:10:11 jroll: well, we have to mark them as triaged again.. 15:10:17 I guess this was not moved from LP 15:10:23 oh. 15:10:26 :| 15:10:29 yep :( 15:10:36 that seems... wrong. 241 out of 283 untriaged? 15:10:56 rloo: it may be a bug in my tool, but it also may be that we don't have the "triaged" tag on things we triaged back in LP 15:10:57 Also we had consensus that we needed to wave through them. It is a lot better than six underdogs untreated 15:10:58 TheJulia triaged a lot of them. i find it hard to believe we only triaged 42 of them. 15:11:22 * jroll tries to figure out what autocorrect did that 15:11:26 s/that/there/ 15:11:35 rloo: some were closed/already fixed 15:11:51 Yeah, autocorrect hates me 15:11:52 * rloo sighs 15:12:10 Merged openstack/python-ironicclient master: Power fault recovery: client support https://review.openstack.org/556774 15:12:15 I can double-check the tool next time 15:12:23 (or anyone can, it's not much code) 15:12:37 #link https://github.com/dtantsur/ironic-bug-report 15:12:58 question about the NOTE at L268... what does it mean? someone needs to migrate the rest of the stuff into storyboard? 15:13:21 I think the plan was to create a board per large effort 15:13:24 kind of a trello board 15:13:31 which is the way storyboard is supposed to be used 15:13:45 (we use it still the old way, it's one of the sources of our problems with it) 15:14:04 ok. not going to ask for an action item on that :) 15:14:09 rloo: it was just tracking/visibility wise 15:14:49 sambetts|afk: when you have a chance, would you please update the Cisco UCS status (L388ish). 15:15:01 What dtantsur said 15:15:15 * etingof can take a look at dtantsur's ironic-bug-report tool 15:15:27 etingof++ 15:16:25 * rloo done with status 15:16:31 15 more people until I reach the hydra... 15:16:45 can we remove "Queens high priorities"? 15:16:47 * rloo roots for TheJulia 15:16:54 dtantsur: yes 15:16:55 or is it there for historical purposes? 15:16:58 okay 15:17:03 dtantsur: nope. not until they get moved to storyboard 15:17:11 * etingof starts to think that TheJulia is trying to get Russian visa 15:17:12 dtantsur: that's what that NOTE was about, right? 15:17:21 rloo: Queens priorities? no, the NOTE is for current things 15:17:26 I kept it originally to make sure we wrapped the queens stuff up 15:17:31 it's line 300 15:17:57 dtantsur: i thought it was for 'everything after this'. i mean, isn't that what storyboard is for? to track this stuff? 15:18:08 It was separate 15:18:12 rloo: I assumed that these things are mostly finished? 15:18:21 Rescue has a nova part, but nova (IIRC) does not use SB 15:18:25 I thought they were delineated with another note 15:19:01 shall we discuss each of these now? we could. 15:19:40 okay. Routed networks - done. Rescue - Nova+Tempest left. Zuul V3 is duplicated with line 276. 15:20:10 let's delete the routed networks and zuul v3 at least then 15:20:11 we could probably leave rescue with these two patches, but everything else can be removed, I guess 15:20:12 i just looked at rescue patch (last ironic patch for rescue?) https://review.openstack.org/#/c/528699/. it links to launchpad, not storyboard. presumably we have a storyboard link for it, so it is being tracked so we can remove the rescue section? 15:20:13 patch 528699 - ironic-tempest-plugin - Rescue mode tempest tests 15:21:39 rloo: because nova has not migrated 15:22:06 Oh, our tempest, yeah the tag should get updated 15:22:24 We can’t do anything until the nova portion lands 15:22:46 right 15:23:19 give me a sec, i'm adding a comment to storyboard for resuce, to include the stuff in whiteboard. then i'll delete from whiteboard. 15:24:36 ok, i'm going to delete the rescue mode stuff. anyone disagree? (you have 3 secs) 15:24:50 No disagreement 15:24:58 +1 15:25:03 gone 15:25:14 what about the py 3.5 stuff? 15:25:21 how are we tracking that in storyboard? 15:25:43 or err maybe we should do this after/outside this meeting 15:27:14 yeah, let's maybe clean up separately.. 15:27:20 I think openstack goals are tracked on SB now 15:27:24 Py35 should have community wide tracking afaik 15:27:24 globally 15:27:30 yep 15:28:05 7 humans to the hydra 15:28:14 so all that stuff wrt TODO, submit py3 job for IPA. is that in storyboard? or have we done it? 15:28:27 do we create a subtask to track that? 15:28:36 wow, TheJulia's line is moving fast! 15:28:38 yes, we need a subtask on the main story 15:28:47 I can do it, if needed 15:28:52 someone want to do it? we can have an action item! 15:28:57 thx dtantsur! 15:29:18 #action dtantsur to make sure a task is present for IPA support of python 3 15:29:24 anything else on statuses? 15:29:46 inspector & wsgi stuff -- dunno what to do with that. can i leave it for you dtantsur to decide to delete or not from whiteboard? 15:30:01 I will do the same - find the story and leave it there 15:30:18 thx dtantsur! 15:30:22 moving on? 15:30:26 ++ 15:30:58 #topic Deciding on priorities for the coming week 15:31:06 okay, some progress this week 15:31:35 are the BIOS links ordered in priority? 15:31:57 * dtantsur puts OSC first 15:33:21 power fault is done, wanna put the neutron events spec o nthe list? 15:33:51 oh, so SB is not used for Pike-era goals 15:34:12 ah, neutron even spec needs updates. vdrok are you on it? 15:34:42 I am fine with what we have, i don't think we need to add more 15:34:51 yeah, I guess it's good 15:34:54 any opinions? 15:35:47 * dtantsur hears crickets and assumes they agree 15:36:17 #topic Open Discussion 15:36:41 so, ironicclient. blame mordred, he quite convinced me that we should stop exposing innocent souls to microversions. 15:37:06 so i was thinking maybe we should defer this until TheJulia is 100% focussed on this. 15:37:07 esp. for set_provision_state where we actually know which state (+ args) corresponds to which version 15:37:20 fine, just want to raise it to more folks 15:37:38 #link https://review.openstack.org/570895 is what I'm trying with openstacksdk 15:37:39 patch 570895 - openstacksdk - [WIP] Add set_node_provision_state to baremetal Proxy 15:38:37 dtantsur: for now, shoudl we put a -2 or something so that client patch doesn't land? or does it matter whether it lands or not? 15:39:03 dtantsur: yeah, kind of, slowly progressing with local testing 15:39:07 I have mixed feelings. We should have done this back in Kilo, and blocking it again.... 15:39:25 thanks vdrok. we may make it a priority next week then. 15:39:35 dtantsur: is there anything wrong with landing it? 15:40:05 rloo: wrong... well, I think we will commit to a much lower-level interface that I'd like to. 15:40:10 i will come clean and say i actually don't know the details of this stuff. 15:40:12 on the other hand, I cannot commit to change it all over ironicclient 15:40:33 maybe let's leave it as it is (and then eventually bury it in favor of openstacksdk) 15:40:36 could we use this lower-level interface for now, until we hae something to replace it? 15:40:40 do we advertise python-ironicclient's python sdk to be high level? 15:40:40 * leave it as it is = land the patch 15:40:57 jroll: we don't either way, I think 15:40:58 yeah that's what I'm leaning toward dtantsur, let openstacksdk be the higher level interface 15:41:21 that's an option. if we agree on that, I'm fine with proceeding with this patch. 15:41:39 * jroll is okay with that, but would like TheJulia input 15:41:41 we just have to be ready, after some time when openstacksdk reaches 1.0, to start telling people to use it 15:42:13 * dtantsur should probably stop using commas at all, at least until he gets a clue how to do in in English 15:42:18 that makes sense to me. wonder what other projects think, cuz i think this is a good rule of thumb, and would like it if other projects did a similar thing. 15:42:32 dtantsur: your commas were good there fwiw 15:42:38 commas are even worse in Russian 15:42:42 hah, guessed then! :D 15:42:43 dtantsur: the commas there were good 15:42:49 :) 15:43:00 (one rule, if the stuff in the commas can be removed and the rest is a complete sentence) 15:43:04 * dtantsur tries to summon mordred again 15:43:16 * mordred runs in screaming and waving his arms around 15:43:19 lol 15:43:24 * rloo runs away 15:43:37 * etingof is scared 15:43:37 mordred: do you know if there is any kind of a general agreement to switch over to openstacksdk eventually? 15:43:39 has the screaming stopped? :) 15:43:50 hehe 15:43:52 Ugh 15:44:12 dtantsur: no - we haven't _generally_ agreed to that yet - it's just the thing I keep saying over and over hoping it'll gain traction 15:44:12 * dtantsur hopes that Ugh is related to TheJulia defeating the hydra 15:44:15 If we’re not going to expose it, I’ll go down back to re-writing our entire usage in nova... 15:44:28 And we should revert my last patch to python-ironicclient 15:44:34 I think we have a little more work to do before we can fully make that push 15:44:43 TheJulia: not really. if we do what I think we should, we will just have a try..except block. 15:44:52 But... there is no way that will likely ever land because we did awful things had been one 15:44:56 mordred: like finishing microversions ;) 15:44:58 but I like what you said earlier - don't try to change ironicclient- and at a point in the future we'll just migrate people over 15:45:02 dtantsur: pish 15:45:07 *details* 15:45:19 * TheJulia sighs 15:45:22 (yeah, that's where the devil is, apparently) 15:45:24 dtantsur: right, I think we (nova/ironic) agreed to not to the "try / except 406" thing in the last ptg 15:45:36 * TheJulia just wants consensus and to stop trying to reinvent the wheel over... and over... and over again. 15:45:39 I think we should just move forward 15:45:43 jroll: which is... precisely how microversions are supposed to work.... 15:45:51 we have a clear path to good code 15:45:53 anyway. I'm fine with the patch. ship it! 15:46:04 dtantsur: "how microversions should work" has changed at least a few times since they began :) 15:46:20 jroll: and one of the changes was "let's not force SDK users to figure them out" ;) 15:46:35 anyway, I think we have a consensus here? 15:46:42 yeah, let's just do the damn thing 15:46:50 #agreed Do the damn thing 15:46:50 can I stop screaming and waving my arms in the air now? 15:46:53 mordred: YES 15:46:56 awesome 15:46:59 * mordred relaxes 15:47:00 Lol 15:47:04 * dtantsur removes earplugs 15:47:06 #undo 15:47:06 Removing item from minutes: #agreed Do the damn thing 15:47:12 Lol 15:47:41 #agreed We will start supporting specifying microversions per call in ironicclient 15:47:46 something like this? 15:47:54 ++ 15:48:15 wasn't the RFE approved for that? :) 15:48:34 how about #agreed don't change the plan yet again :P 15:48:39 yeah, we've already agreed to it 15:48:50 jroll: yes please 15:49:35 :) 15:49:41 anything else? 15:50:13 * jroll has nothing 15:50:23 * etingof is wondering how to attract people to his awesome virtualbmc patch.... 15:50:36 https://review.openstack.org/#/c/488874/ 15:50:37 patch 488874 - virtualbmc - multiprocess server, ZMQ-based management cli tool 15:50:38 beer and/or cookies usually work 15:50:43 ++ 15:50:52 (screaming and waving hands doesn't work for some people) 15:51:13 will you guys come to Brno some day? 15:51:15 * jroll just can't be bothered to review a giant patch to help people run a test tool in production :/ 15:51:16 * dtantsur hears crickets screaming and waving hands 15:51:29 etingof: will you man come to Berlin? ;) 15:51:34 okay, okay, it's offtopic 15:51:40 crickets 15:51:48 jroll: tbh we have problems in the CI as well, and with long-running QE environments 15:52:01 it's not only people shutting their legs 15:52:08 ehmm.. shooting? 15:52:12 lol 15:52:44 * dtantsur has a decorative plate saying "Today has been canceled. Go back to bed." 15:52:54 so, thanks all :) 15:53:01 #endmeeting