17:00:00 #startmeeting ironic 17:00:00 Meeting started Mon Mar 28 17:00:00 2016 UTC and is due to finish in 60 minutes. The chair is jroll. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:02 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:05 The meeting name has been set to 'ironic' 17:00:10 o/ 17:00:11 #chair NobodyCam devananda 17:00:12 Current chairs: NobodyCam devananda jroll 17:00:12 o/ 17:00:13 sup! 17:00:18 o/ 17:00:26 o/ 17:00:28 o/ 17:00:35 o/ 17:00:40 o/ 17:00:44 o/ 17:00:58 I expect most people are out for easter monday today 17:01:04 wazzup! 17:01:09 let's jump right in 17:01:12 ah yes! 17:01:19 #topic announcements and reminders 17:01:39 first things first, apparently y'all are stuck with me again this cycle 17:01:46 o/ 17:01:47 Contrats jroll 17:01:53 o/ 17:01:54 :P 17:01:57 o/ 17:02:07 ah yeah, well! I think I might endure that... :D 17:02:12 :) 17:02:16 congratz jroll 17:02:18 next up - we released all the things last week, they're all here 17:02:18 jroll: grats :) 17:02:19 And also TheJulia !!!!!!! 17:02:20 #link http://releases.openstack.org/mitaka/index.html 17:02:22 +1 17:02:24 and we have a new person... 17:02:34 yay! 17:02:41 and we welcomed TheJulia as our newest core \o/ 17:02:46 congrats! 17:02:55 wow, congratz TheJulia ! 17:02:59 :) 17:03:00 Thank you everyone 17:03:08 o/ 17:03:11 * jroll thanks folks for the congrats as well 17:03:43 another thing - our stable/kilo gate is green again, thanks goes to mriedem for fighting that 17:03:44 * thiagop thinks he missed a lot from last week... 17:03:51 congrats jroll, TheJulia :) 17:04:10 * NobodyCam gives congratz to everyone for the work that went into the release 17:04:11 jroll, one thing about this Kilo gate-fixing: it's sad that we've lost the pxe_ipa job 17:04:21 I see it as kind of a dangerous situation 17:04:24 and newton is open for development 17:04:29 yea newton! 17:04:35 dtantsur: it wasn't stable at all :/ 17:04:40 <[1]cdearborn> \o 17:04:46 dtantsur: maybe we don't deprecate bash ramdisk until kilo EOL then 17:04:57 s/deprecate/kill/ 17:04:58 jroll, I think we already agreed on that :) 17:05:05 right 17:05:08 so I think it's okay 17:05:38 * dtantsur checks if killing the bash ramdisk is on the list of summit topics 17:05:49 oh, one more gate thing, the IPA post job to publish ramdisks is currently broken, hoping https://review.openstack.org/297796 will fix it 17:06:09 I've been grinding on that, will continue until it's fixed - then making stable tests actually use the stable ramdisk 17:06:39 oh, good 17:06:44 more newton things: summit ideas go here: 17:06:46 #link https://etherpad.openstack.org/p/ironic-newton-summit 17:06:58 and starting on priorities here: 17:07:00 #link https://etherpad.openstack.org/p/ironic-newton-priorities 17:07:10 I expect to finish that during the summit 17:07:33 we need to finish the session planning by mid-april 17:07:39 I think this list is more than complete :D 17:07:46 summit schedule has been posted here: 17:07:48 #link http://lists.openstack.org/pipermail/openstack-dev/2016-March/090606.html 17:07:57 that's a pdf attachment, so check your actual email for the schedule 17:08:08 dtantsur: finish trimming it down, I should say :) 17:08:14 hehe 17:08:28 so that's a million announcements from me in no particular order. anyone have anything else? 17:09:16 jroll: has these ever been another version of that newton summit etherpad? 17:09:26 jroll: I feel like I added something to a completely different etherpad 17:09:29 JayF: not that I know of 17:09:31 heh 17:10:10 there was a netwon section onthe midcycle pad 17:10:16 ooo yeah 17:10:19 #link https://etherpad.openstack.org/p/ironic-mitaka-midcycle 17:10:32 https://etherpad.openstack.org/p/mitaka-ironic-design-summit-ideas is the other etherpad 17:10:36 that wasn't linked here 17:10:37 line 299 17:10:42 oh, mitaka vs newton 17:10:44 JayF: that's last summit heh 17:10:45 oy 17:11:19 okay, moving on then 17:11:29 #topic subteam status reports 17:11:35 I don't expect too much here 17:11:44 #link https://etherpad.openstack.org/p/IronicWhiteBoard 17:11:46 line 98 17:12:04 we'll need to update this for newton soon, once we figure out priorities 17:12:13 any questions/comments/concerns here? 17:12:43 * jroll waits about 2 minutes 17:14:32 * jlvillal hums Jeopardy theme song to himself... 17:14:37 lol 17:14:48 heh 17:14:51 moving on then 17:15:03 #topic RFE review 17:15:21 so I've added this section to the agenda, to discuss RFEs and if they need a spec 17:15:26 I just have one this week: 17:15:32 #link https://bugs.launchpad.net/ironic/+bug/1561100 17:15:32 Launchpad bug 1561100 in Ironic "[RFE] Centralize config options in Ironic" [Undecided,In progress] - Assigned to Ramamani Yeleswarapu (ramamani-yeleswarapu) 17:15:48 so this references a nova spec, and something nova has already been working on 17:16:05 there's also a cross-project spec in review right now: 17:16:07 #link https://review.openstack.org/#/c/295543/ 17:16:32 so I'm thinking this can be approved without a spec, but perhaps we wait for the cross-project spec to land? 17:16:47 I think that would be a good idea 17:17:10 As a note, we discussed the config stuff two weeks ago: http://eavesdrop.openstack.org/meetings/ironic/2016/ironic.2016-03-14-17.00.log.html 17:17:39 right, so that's about using oslo-config-generator, which we should do first 17:17:45 It was decided: "we will move to oslo config gen in newton, followed by nova-style refactor of config opts" 17:17:45 but this is an additional thing 17:17:51 oh. heh. 17:17:57 fair enough 17:18:14 I still think we should wait for the cross-project spec 17:18:26 which, btw, is being discussed in the cross-project meeting tomorrow, iirc 17:18:32 that is 2100 utc tomorrow 17:18:40 ++ we already have experience doing a big thing without waiting for a cross-project agreement ;) 17:18:56 * dtantsur never misses a chance to mention microversioning 17:18:56 ya, let's wait it out this time 17:18:59 Agreed on waiting 17:19:05 hah 17:19:07 ok cool 17:19:14 Added spec link to the RFE bug 17:20:01 cool 17:20:06 I posted a note on holding off for approval 17:20:31 #topic open discussion 17:20:37 now that newton is open would love to get eyes on https://review.openstack.org/#/c/275726 (Dynamic allocation of nodes for OneView) 17:20:46 I feel like I had a thing here but not sure what it was now :) 17:21:01 NobodyCam: I looked at that, was a little put off by the use of an external daemon :) 17:21:15 NobodyCam: but didn't wanna rain on the parade of a driver I'll never use :) 17:21:38 JayF: I beleieve it is worth a comment on the spec thou 17:21:49 I do want to point out that I'm working on using the stable builds of the agent, and getting tempest-full running. I made rfes for them to track it: https://bugs.launchpad.net/ironic/+bug/1561756 https://bugs.launchpad.net/ironic/+bug/1561760 17:21:49 Launchpad bug 1561756 in Ironic "[RFE] Use stable branch of agent when testing stable branches" [Wishlist,In progress] - Assigned to Jim Rollenhagen (jim-rollenhagen) 17:21:50 Launchpad bug 1561760 in Ironic "[RFE] CI job to run full tempest suite against ironic" [Wishlist,Confirmed] 17:22:06 NobodyCam: I mean, honestly, I think it's a not-awesome overall pattern; but I can't bring myself to care enough when it's isolated to the driver. 17:22:48 JayF: I believe they commented on a previous revision that they intend to make that a periodic task at a later point in time since it is just a cleanup/reconsilliation thing 17:23:18 TheJulia: my impression of it was the oppostie; they were going to use a periodic task but changed to a daemon so it could listen ont he event bus 17:24:09 JayF: Acutal discussion with them seemed otherwise since they already had the daemon written 17:24:16 JayF: Nope, the intention is to use periodic task. That's on our roadmap 17:24:32 thiagop: why wouldn't the spec show that then? 17:24:50 thiagop: Could we get the spec to be a little more clear on that? :) 17:24:55 If the spec said that I probably would've -1'd the intermediate state of having a separate daemon :) 17:25:27 jroll: that was discussed as a sugestion by lucas, I think we mentioned that we will do a follow up patch as soon as we figure out how to do that properly 17:25:34 JayF: agree, we should not have that intermediate state imo 17:25:53 why build in technical debt if we already have a plan to avoid it? 17:26:11 +1 17:27:12 Seems like the path forward then is to figure out how to properly have the periodic task 17:27:46 * jroll commented on the review 17:27:50 * JayF too 17:28:05 yep, let's do that before we start building it otherwise :) 17:28:33 I didn't have a chance to read this patch, but I still suspect we might want to agree on a common design for such things 17:28:41 jroll TheJulia ok then, we'll figure out how to do that now that Mitaka is out of our plans anyway 17:28:43 it's not the only driver that will have such functionality 17:29:08 dtantsur: I saw the topic you had proposed to the summit, TY 17:29:17 dtantsur: it will be interesting 17:29:19 didn't we have a spec for per-driver tasks.... 17:29:23 dtantsur: yeah, I'm wondering that as well 17:29:33 NobodyCam: yes, we have per-driver parallel tasks already :) 17:29:33 NobodyCam, we do have per-driver tasks 17:29:38 for a long time actually :) 17:29:44 thiagop: thanks in advance for updating that 17:29:46 I was looking for hte spec for that 17:29:49 the inspector inspection interface is using it 17:30:30 NobodyCam, thiagop, e.g. https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/inspector.py#L125-L140 17:30:31 found it 17:30:34 yep 17:30:43 we just thought that having that and evolving to periodic tasks later would be a good move 17:30:43 https://github.com/openstack/ironic-specs/blob/master/specs/kilo-implemented/driver-periodic-tasks.rst 17:31:18 thiagop: I'd rather do it right the first time 17:31:31 thiagop: when you think about how that requires documentation for the original way, then documentation on upgrading + new way, plus deprecation efforts and such 17:31:32 which is why I'm inclined to wait and discuss a more general approach at the summit 17:31:32 jroll: a'righ 17:31:38 thiagop: it makes a lot more since to get it 'right' from the start :) 17:31:50 JayF: makes sense 17:32:18 anything else for open discussion? 17:32:42 thanks all 17:32:47 for ^^^ 17:33:21 I'll note that I'll be working on Rescue specs this week; if anyone has interest in it LMK, I'm always keen to get some help 17:34:02 JayF: Ping me when you post the draft because that may have some implications on boot from volume 17:34:32 TheJulia: I'd very much hope they are sufficiently detached, but there are already pretty completely specs on the ironic side 17:34:41 TheJulia: https://review.openstack.org/#/c/171878/ :) 17:34:43 TheJulia: just needs a little bit more love + all the nova side bp completed 17:34:50 JayF: me too 17:34:57 ty for the link 17:35:33 ok, thanks all for showing up, good meeting 17:35:37 #endmeeting