17:01:52 <devananda> #startmeeting ironic
17:01:54 <openstack> Meeting started Mon Dec 15 17:01:52 2014 UTC and is due to finish in 60 minutes.  The chair is devananda. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:55 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:01:57 <openstack> The meeting name has been set to 'ironic'
17:01:59 <rloo> o/
17:02:15 <Nisha> o/
17:02:25 <devananda> g'morning / afternoon / evening / middle of the night to everyone :)
17:02:37 <naohirot> good morning
17:02:42 <devananda> #topic announcements
17:02:49 <clif_h> o/
17:02:52 <NobodyCam> I added a couple of items to the agenda late (just this morning)
17:03:05 <devananda> one hopefully quick announcement
17:03:12 <GheRivero> o/
17:03:38 <devananda> I'm working on getting details for the midcycle, like hotel and what not, and will distribute that once I have it
17:03:49 <devananda> along with an eventbrite signup thing
17:03:50 <NobodyCam> awesome :)
17:04:02 <lucasagomes> nice one
17:04:08 <devananda> grenoble seems farther away from airports than I realized
17:04:28 <devananda> but it's a nice flat city with good local transportation
17:04:33 <jroll> good opportunity to rent a fast car and drive through the back country :)
17:04:40 <devananda> jroll: yup
17:04:42 <NobodyCam> lol
17:04:46 <devananda> if that's your thing :)
17:05:15 <devananda> that's it for me
17:05:24 <NobodyCam> can I bring my own car?
17:05:37 <devananda> oh, also, welcome back to lucasagomes :)
17:05:37 <jroll> sounds expensive :P
17:05:42 <devananda> NobodyCam, lucasagomes - any announcements?
17:05:54 <lucasagomes> o/ thanks, it's good to be back
17:06:02 <lucasagomes> not really I'm still catching up with evrything I missed last week
17:06:09 <devananda> k
17:06:15 <devananda> #topic status board
17:06:19 <NobodyCam> not from me this time
17:06:41 <devananda> NobodyCam: woops. jumped early there, sorry
17:06:54 <devananda> following our new approach to weekly statuses, they are posted on the white board
17:06:57 <devananda> #link https://etherpad.openstack.org/p/IronicWhiteBoard
17:07:05 <devananda> let's take a minute to review, then raise any questions / concerns if we have them
17:07:09 <devananda> (anyone can do that)
17:07:17 <NobodyCam> Thanks to every for the new status update format... is supper great to read them over the weekend and digest slowly
17:07:32 <NobodyCam> everyone*
17:08:14 <jroll> iLO - "Setting up 3rd-party CI" \o/
17:08:17 <devananda> wanyen: ooh - third party CI will be great
17:08:26 <GheRivero> sorry for some late additions, but oslo weekly meeting just finished :)
17:08:26 <devananda> wanyen: have you been in contact with anteaya about that?
17:08:30 <rloo> not sure what the HI bugs are, but do we need to try to get any in for kilo-1?
17:08:33 <devananda> wanyen: I believe she coordinates the third-party-ci weekly meetings
17:08:52 <devananda> I don't see dtantsur here?
17:09:09 <lucasagomes> I think he's afk (judging by his nickname)
17:09:12 <wanyen> deva, no.  we requested an account already.
17:09:43 <wanyen> deva, okay.  we will have someone attend the CI meeting
17:09:47 <naohirot> devananda: I didn't see the status report email in openstack-dev last week, are we going to send the status to the maillist this week?
17:10:01 <devananda> jroll: ^
17:10:07 <anteaya> wanyen: you don't request an account anymore it is self-serve
17:10:10 <jroll> oops
17:10:26 <jroll> naohirot: devananda: last week ended at 10pm for me, I completely forgot
17:10:32 <naohirot> jroll: yes, are we?
17:10:35 <jroll> yes
17:10:37 <anteaya> wanyen: be sure to read this entire page: http://ci.openstack.org/third_party.html#creating-a-service-account
17:10:45 <naohirot> jroll: Okay
17:10:53 <lucasagomes> rloo, seems all the high bugs are in progress/fixed now
17:10:55 <wanyen> anteaya, tx
17:11:09 <anteaya> wanyen: welcome, I'll chat more in a third party meeting with you
17:11:18 <lucasagomes> there's one assigned to be that I don't think it's high priority as well (will reverify and change it accordingly)
17:11:28 <rloo> lucasagomes: thx
17:11:28 <devananda> naohirot, lintan: I noticed that the AMT driver proposal switched to using wsman instead of amttools
17:11:41 <Guest59984> yes
17:11:41 <Guest59984> terday
17:11:56 <naohirot> devananda: yes
17:12:36 <Guest59984> to support amt better
17:12:41 <devananda> naohirot: ack. just pointing it out here as I think that's something that other folks who were waiting for the AMT driver might want to know
17:12:42 <naohirot> devananda: I noticed that lintan is updating the spec.
17:12:58 <jroll> Guest59984: who are you? :P
17:13:13 <Guest59984> i am lintan
17:13:19 <jroll> ok
17:13:39 <devananda> 2 minutes before we move on
17:13:55 <jroll> Guest59984: you can fix your nickname with /nick lintan_
17:13:58 <jroll> since lintan is in use
17:14:24 <naohirot> devananda: I'm going to refer to AMT in order to know how to implement irmc.
17:14:38 <devananda> rloo: some of those "high" bugs don't seem appropriately triaged to me. they've been open and marked "high" for months ...
17:14:59 <devananda> rloo: perhaps it's time for us to have a bug-day again and just go through the list?
17:15:18 <NobodyCam> oh ++
17:15:34 <NobodyCam> good to do somehting like that before the holiday break
17:15:41 <rloo> devananda: yeah, but maybe next week? am wondering if we should spend this week trying to get patches in and maybe get specs moving along...
17:15:48 <devananda> ok, let's talk about that in open discussion
17:15:56 <devananda> I think it'd be great, but yea, when is a good question
17:15:57 <lucasagomes> devananda, +1 yeah there's one associated to me (the vendor passthru stuff) I just marked it as medium
17:16:01 <lucasagomes> cause def it wasn't high
17:16:10 <devananda> thanks all, let's move on
17:16:19 <jroll> bugday++
17:16:26 <devananda> #topic kilo-1 progress
17:16:37 <devananda> #link https://launchpad.net/ironic/+milestone/kilo-1
17:16:53 <devananda> rloo: ohhai - it's like you planned that ;)
17:17:11 <rloo> honestly, I didn't...
17:17:13 <devananda> so, it's probablyworth announcing for folks
17:17:27 <devananda> #info last week we landed the new state machine spec
17:17:42 <devananda> #link http://specs.openstack.org/openstack/ironic-specs/specs/kilo/new-ironic-state-machine.html
17:17:49 <devananda> I've marked it as "Informational" in launchpad
17:18:00 <devananda> because there is not a specific single body of code that's going to implement this
17:18:20 * lucasagomes will review the patches on gerrit asap
17:18:21 <devananda> it represents the direction we all agree the project is headed, and *many* individual specs are going to be needed to get there
17:18:28 <devananda> as we implement different aspects of the state machine
17:18:46 <devananda> since we actually dont have *any* formally modelleled state machine today, I've been implementing one
17:19:05 <devananda> I see NobodyCam put those links in the agenda in the next section -- so let's talk about that there
17:19:20 <devananda> for now, i'm just calling it out as "a big thing that was on our kilo-1 priority list"
17:19:36 <devananda> there are only 3 other blueprints targeted to kilo-1
17:19:54 <devananda> so, if you own one of those, please speak up with the status now :)
17:20:04 <NobodyCam> does the xml bug need to land for -1?
17:20:17 <jroll> lucasagomes: you still planning on doing pxe/configdrive stuff?
17:20:24 <devananda> we'll be tagging the k1 milestone as soon as that page is all "implemented" -- whether things land, or get removed from the page
17:20:27 <lucasagomes> jroll, I will! for sure
17:20:40 <lucasagomes> jroll, I'm finishing up the root device thing and I will jump on that
17:20:46 <jroll> lucasagomes: cool, I wonder if we should assign that blueprint to you then :)
17:20:50 <lucasagomes> probably won't finish on k-1 tho
17:20:56 <lucasagomes> jroll, +1
17:20:57 <jroll> devananda: we might need to push back configdriev
17:21:12 <lucasagomes> jroll, we need to assign to u the boot/deploy ifaces as well
17:21:12 <jroll> don't even have code up for the nova side yet so not a huge deal
17:21:23 <devananda> jroll: ack. bumping it now
17:21:25 <jroll> lucasagomes: yeah, is there a spec for that?
17:21:32 <lucasagomes> jroll, nop, but there's a blueprint
17:21:36 <NobodyCam> devananda: bumping to k-2?
17:21:40 <devananda> i see that mrda put up code for "logical names"
17:21:41 <lucasagomes> as it was something that everybody agreeded
17:21:43 <devananda> NobodyCam: yes
17:21:49 <lucasagomes> we didn't create a spec for it, we approved the bp straight away
17:21:54 <jroll> lucasagomes: cool, I think it'll be january when I get to that
17:21:55 <lucasagomes> agreed*
17:21:59 <devananda> #info "expose config drive to instances" bumped to kilo-2
17:22:03 <lucasagomes> jroll, ack
17:22:11 * lucasagomes assigns it to jroll
17:22:16 <jroll> thanks
17:22:55 <lucasagomes> done, thank you :)
17:23:26 <jroll> anything else on kilo 1?
17:23:36 <devananda> as for xml support, it looks like mrda's patch was abandoned a month ago, and I don't see any new work
17:23:53 <NobodyCam> thats why I asked
17:23:55 <Tanlin> as i update the amt spec, code is ready once the spec is approve
17:23:57 <devananda> there's support in wsme now which should make it easy for us to do
17:24:12 <jroll> Tanlin: is the code up already?
17:24:46 <devananda> rloo: so to your question, it seems like there are only 2 specs up that have code ready for this week
17:24:54 <devananda> logical names && AMT power driver
17:25:03 <rloo> jroll: Tanlin has code up. I (ahem) objected to some of it. I think that's why the spec is being updated.
17:25:10 <jroll> ha, ok
17:25:11 <Tanlin> jroll: yes, except one tiny conf
17:25:14 <rloo> jroll: would be good if others looked at it
17:25:19 <jroll> yep
17:25:21 <NobodyCam> rloo: link?
17:25:47 <devananda> https://review.openstack.org/135184
17:26:17 <rloo> NobodyCam: https://review.openstack.org/#/c/135184/
17:26:18 <jroll> oh, I remember this debate
17:26:19 <Tanlin> yes, i have some discussion with rloo and need your opinior
17:26:20 <jroll> cool
17:26:31 * jroll would like to see the spec update
17:26:36 <NobodyCam> #link https://review.openstack.org/#/c/135184
17:27:23 <devananda> #info AMT power driver spec needs revision, review is up. https://review.openstack.org/141269
17:27:38 <devananda> #info AMT power driver still targeted to kilo-1
17:28:06 <devananda> oh, also
17:28:16 <devananda> #info python-ironicclient 0.3.2 released last week
17:28:20 * lucasagomes adds to the todo list
17:28:27 <jroll> woohoo
17:28:33 <lucasagomes> oh nice
17:28:41 * NobodyCam will make an effort to review the AMT
17:28:43 <devananda> just a minor release, but had support for some new features that will show up in kilo-1
17:28:47 <NobodyCam> review*
17:29:05 <devananda> ok - about to move on, unless anyone raises something else
17:29:14 <sambetts> Cisco power driver spec is up for review
17:29:16 <sambetts> https://review.openstack.org/#/c/139517/
17:29:53 <devananda> #topic state machine
17:30:04 <NobodyCam> ah yes
17:30:14 <NobodyCam> this is the "CURRENT" state machine
17:30:42 <NobodyCam> devananda: have busted his rear on getting these reviews up (thank you devananda)
17:30:46 <sambetts> it likely needs work, any comments would be appreciated
17:31:12 <NobodyCam> I have listed the current reviews on the agenda page
17:31:40 <NobodyCam> the sooner we can land these patches, the better I feel
17:32:13 <devananda> #link https://review.openstack.org/#/q/status:open+project:openstack/ironic+branch:master+topic:bp/new-ironic-state-machine,n,z
17:32:16 <NobodyCam> this gives us a good clear path for what we have now to what we are looking for fromthe NEW state machine
17:32:27 <Shrews> i'm good with the first review in that series once rloo's comments have been addressed
17:32:41 <rloo> I started looking at those patches. I feel like they are a higher priority than the kilo-1 stuff. What do others think? (want to focus this week on hi-priority stuff)
17:33:13 <NobodyCam> i would say at teh same level as the k-1 stuff
17:33:59 <devananda> these are all related-to a spec which I feel was essential for kilo-1 (the state machine)
17:34:03 <NobodyCam> speaking of.. how many folks are not around after this week
17:34:18 <devananda> while it's informational, this is a non-trivial refactoring of code to support that spec
17:34:20 <Shrews> NobodyCam: i'm out beginning this friday until january
17:34:21 <rloo> devananda: do you have an idea of how to get from current states to new states?
17:34:24 <jroll> NobodyCam: /me is off dec 21-27
17:34:26 <lucasagomes> yeah I think that if we could get the state machine in k-1 that would be awesome
17:34:29 <devananda> rloo: so, I agree - this should all land in kilo-1 if possible
17:34:41 <lucasagomes> but I haven't looked at the code yet, I will do it asap
17:34:42 * JayF gone all next week
17:34:48 <devananda> rloo: right now, just an idea, but this refactoring is the beginning of that
17:35:01 <NobodyCam> so this is the week to land'um
17:35:09 <devananda> rloo: there's going to be some odd edge cases -- the last patch in the series starts to find them
17:35:15 <devananda> where drivers are currently manipulating the state directly
17:35:51 <devananda> rloo: the new state machine is only adding new states, not renaming or removing states
17:36:05 <Shrews> devananda: i spent a good bit of time pulling the fsm code out and playing with it independently to understand how it works. seems pretty sound.
17:36:22 <devananda> so I think we can migrate, as long as there are no in-progress operations during the migration
17:37:11 <devananda> but I don't yet have an implementation of the new state machine to migrate to. not sure I'll get to that this week. next week is more likely
17:37:13 <rloo> devananda: and somehow deal with backwards compatibility :-(
17:37:42 <rloo> devananda: I think if we can get the current states in an FSM this week, that'd be great ;)
17:37:48 <NobodyCam> devananda: how large do you think the edge case issue (of drivers changing state directly) is going to be
17:37:49 <jroll> I don't think backwards compat will be horrible, with us expressing it this way
17:38:00 <jroll> we'll just have extra states and transitions we're dragging around for a while
17:38:14 <devananda> rloo: I think "new server, old client" will be relatively easy -- the same three states are present and should be supported requests
17:38:14 <jroll> oh, out of tree drivers.
17:38:50 <devananda> jroll: we'll be expressing new states that the older clients dont understand
17:39:07 <jroll> mmm
17:39:11 <devananda> jroll: but POST ... {state: <state>} should still work with the old three states
17:39:17 <jroll> yeah, agree
17:39:21 <devananda> that's what I'm aiming for
17:39:35 <devananda> make sure that an older nova.virt.ironic driver still works with a new ironic service
17:39:41 <devananda> so that operators can upgrade Ironic first, then Nova second
17:40:08 * jroll wonders if he also wants the other direction to work
17:40:33 <devananda> anyhow. let's not birdwalk too far into that, and instead let's discuss that on the reviews
17:40:42 <devananda> jroll: please no
17:40:42 <jroll> yeah
17:40:45 <jroll> lol
17:41:38 <NobodyCam> devananda: are you planing on pushing up a new version of 139215 to address rloo comments?
17:41:49 <devananda> rloo: I'll address your comments on the first patch today
17:41:59 <NobodyCam> :)
17:42:05 <devananda> I'd love to get feedback on the rest of the patches this morning so I can try to address all of them this afternoon
17:42:11 <rloo> thx devananda
17:42:27 <devananda> in particular, the last two or three
17:42:33 <devananda> which start to show the ramifications of this change
17:42:42 <devananda> eg, https://review.openstack.org/#/c/140883/
17:42:51 * NobodyCam has not gotten that far in the chanin yet :(
17:42:52 <devananda> and https://review.openstack.org/#/c/140869/4
17:43:32 <devananda> 140883 should be covering all the in-tree drivers
17:43:48 <devananda> anyone working on a new driver, or maintaining out of tree drivers, is going to want to do similar
17:44:01 <devananda> it's also very likely I missed some state changes, so that patch may grow
17:44:15 <devananda> was just grepping ...
17:44:38 <NobodyCam> its a great start. thank you devananda :)
17:44:53 <devananda> anyhow, thanks in advance for the reviews. let's move on so we have time for open discussion
17:44:58 <devananda> #topic open discussion
17:45:21 <jroll> ok, bugday
17:45:33 <jroll> or review jam day or whatever
17:45:36 <kozhukalov> hey guys
17:45:40 <NobodyCam> lol okay this weekend I started to hack on a PixieBoots Irc bot
17:45:40 <kozhukalov> Folks, there was a discussion about Fuel Agent driver for Ironic. The spec https://review.openstack.org/#/c/138115/  was updated significantly and now should be much more compliant with what we have in Ironic. Did anyone have a chance to look at it? If so, do you think it’s possible to land this kind of feature to Ironic?
17:45:40 <jroll> any chance we can do that this week rather than next?
17:45:49 <devananda> bug day? holiday time?
17:45:50 <devananda> jroll: heh :)
17:46:05 <lucasagomes> jroll, I'm good with that, maybe at the end of the week like thursday
17:46:26 <NobodyCam> thursday works
17:46:26 <jroll> ++
17:46:36 <jroll> rloo: how is thurdday?
17:46:49 <devananda> jroll: I'd love to see a few of us spend a chunk of time going through the bug list and re-triaging things
17:47:14 <JayF> Hey guys; I was wondering how much folks would be in favor of a single repository with things needed to build Ironic's various required ramdisks. DIB elements needed for Ironic, the CoreOS ramdisk builder + logic to build IPA would go in there to start
17:47:26 <NobodyCam> do we need dtantsur fir that
17:47:33 <rloo> jroll: I'm around Thurs, but if we're still trying to get patches into kilo-1, I'd rather focus on that.
17:47:43 <devananda> thursday is good for me, though I may be out part of the day
17:47:49 <jroll> rloo: right, we could do a review day for those
17:47:56 <Shrews> I'm out part of the day on thursday for an appointment
17:47:58 <devananda> NobodyCam: it'd be good if he's there, yes
17:48:23 <jroll> JayF: I know you've gotten +1 from a few people on that, I feel like you should just write up a simple spec
17:48:29 <rloo> jroll: also (question for everyone I think). is it more important to do a bug review or maybe a spec review first? wondering if people are being delayed cuz specs aren't being reviewed fast enough or whatever
17:48:32 <devananda> jroll: how about a reviewday tomorrow focused on kilo-1 open reviews, and then visit the bug list after kilo-1 is tagged
17:48:34 <lucasagomes> JayF, I'm good with that too
17:48:43 <JayF> jroll: so do I :) I wanted to do it this weekend but didn't make the time
17:48:47 <JoshNang> JayF: +1 but i'm a bit biased
17:48:47 <lucasagomes> tho I really would like to see all the ramdisks functionalities merged with IPA
17:48:58 <wanyen> JayF, iLO team will be in favor of that
17:49:03 <lucasagomes> so we could replace all ramdisks and just use ipa instead
17:49:08 <devananda> JayF: i'm wondering a few things about that
17:49:09 <JayF> lucasagomes: I'm not as +1 on that as others are; but I think unifying where things are stored will help us in thinking more creatively about ramdisks
17:49:26 <jroll> devananda: I'm good til noon or so PST tomorrow, wfm
17:49:38 <lucasagomes> right, my concern is also having the dib element there
17:49:45 <devananda> JayF: how will that project be tested in relation to the various projects which contain the build tools?
17:49:51 <NobodyCam> JayF: will that not mean we are maintaining several out of tree "drivers/elements" for different projects?
17:49:54 <lucasagomes> because time to time they may change some stuff on the dib elements that might affect the ironic-deploy element
17:50:12 <devananda> JayF: what will that project look like, as an amalgam of *different* types of scripts?
17:50:30 <JayF> devananda: how is tripleo-image-elements tested in relation to other projects?
17:51:08 <JayF> devananda: I forsee something along the lines of tripleo-image-elements .. with DIB elements in one dir and a CoreOS builder in another
17:51:13 <devananda> JayF: I haven't kept up on changes in their gate structure. NobodyCam, do you know?
17:51:27 <JayF> I'm also willing to split out the CoreOS builder stuff into a completely separate project that then Ironic/IPA could consume
17:51:44 <NobodyCam> I have not followed the ci changes no :( maybe adam_g knows?
17:52:04 <devananda> JayF: then either that repo co-gates with both dib and coreosbuilder, or it does not gate and we trust folks not to break it
17:52:21 <devananda> I'm generally -1 on more things co-gating
17:52:31 * NobodyCam is not a fan of the latter option
17:52:34 <adam_g> i *think* t-i-e changes are tested with a tripleo devtest run, but IMBW
17:53:16 <adam_g> we did do some co-gating with ironic's devstack job + diskimage-builder (where the deploy-ironic element lives)
17:53:47 <JayF> devananda: I'm not sure if what I propose is the right solution; but I think it's better (gating issues taken into account as well) than us currently having builders scattered about :/
17:53:53 <jroll> let's leave it to the spec process, I think
17:54:03 <jroll> we're going to forget everything we said here
17:54:27 <devananda> kozhukalov: I haven't had a chance to look at the new spec version yet. will try to this week, but will be focusing on kilo-1 milestone
17:54:45 <wanyen> speaking of ramdisk, is there any plan to merge PXE ramdisk with IPA ramdiks?
17:54:46 <devananda> jroll: ++
17:55:04 <NobodyCam> *BEEP* five minute bell
17:55:11 <devananda> wanyen: they serve different use-cases today. so, not in the short term.
17:55:26 <wanyen> I meant add support of iscsi to IPA
17:55:33 <devananda> wanyen: that said, I think it'd be great if IPA supported the remote-iscsi-attach method as well
17:55:33 <lucasagomes> wanyen, i'd like to see those merged, but I don't know about anyone who's actually working on it
17:55:48 <devananda> jroll, JayF: any plans for ^ ?
17:55:49 <jroll> wrt fuel agent stuff, I still maintain that we can land those features in IPA quickly, but I may be biased etc. idk. going to stay out of it other than pointing this out.
17:55:59 <jroll> wanyen: devananda: should be pretty easy to do
17:56:04 * JayF == jroll w/r/t fuel agent
17:56:29 <wanyen> jroll, are u plan to do it for kilo?
17:56:34 <jroll> devananda: someone just needs to take on the work, there's two ways of going about it
17:56:34 <Nisha> remote-iscsi-attach method?
17:56:45 <lucasagomes> NobodyCam, any news on the DIB element for IPA?
17:56:49 <jroll> wanyen: I'm not planning on it, I will if nobody else does
17:56:59 <jroll> lucasagomes: it exists, it works, it also requires 3gb of RAM :(
17:57:06 <lucasagomes> :O
17:57:11 <lucasagomes> 3gb to build?
17:57:14 <devananda> jroll: yea, that is fail :(
17:57:15 <NobodyCam> lucasagomes: I restarted my effort. the original direction was not correct
17:57:21 <lucasagomes> or the machine needs 3g to boot the ramdisk?
17:57:22 <jroll> lucasagomes: 3gb to RUN
17:57:24 <lucasagomes> omg
17:57:27 <JayF> jroll: devananda: 1) Being hack the ramdisk to start an iscsi target daemon if an option is passed or 2) Give IPA a call to expose disks as an iscsi target
17:57:27 <jroll> right.
17:57:28 <lucasagomes> why's that?
17:57:49 <jroll> lucasagomes: because people didn't spend weeks optimizing it like we did for the coreos image
17:58:01 <JayF> both options lead to a heavier ramdisk and cause issues in the gate, right? (size of ramdisk is huge)
17:58:16 <jroll> lucasagomes: JayF quite literally spent weeks making CoreOS small
17:58:17 <lucasagomes> jroll, I see.. well seems like it's going on the right direction then
17:58:22 <lucasagomes> it's mostly optmizations left
17:58:23 <NobodyCam> I had started with a swift-temp-url element this was wrong. The logic reallys needed to be in the incubator scriots them self
17:58:33 <kozhukalov> devananda: thanx a lot, I really appreciate that, will wait for any news
17:58:47 <NobodyCam> *Two minutes*
17:59:55 <NobodyCam> please ping me any functions you think a PixieBoots Irc Bot should have: current code found here: https://github.com/NoBodyCam/PixieBootsIrcBot
17:59:55 <wanyen> Jroll, teh node zapping functions will be implmeneted for both IPA and pxe ramdisk?
18:00:01 <sambetts> Any comments on the Cisco power driver spec would be appricated https://review.openstack.org/#/c/139517/
18:00:15 <jroll> wanyen: I don't know, I hope so?
18:00:17 <JoshNang> wanyen: some will be in pxe
18:00:27 <NobodyCam> Time
18:00:28 <JoshNang> like, erasing disks i think we can do
18:00:32 <devananda> so we clearly can't pivot the upstream CI to use anything that requires a 3GB VM to run it ...
18:00:33 <wanyen> joshnang and jroll: that's great!
18:00:40 <lucasagomes> devananda, yeah :/
18:00:53 <JoshNang> wanyen: we can finish in -ironic
18:00:56 <lucasagomes> the problem with the default ramdisk is the communication between ironic and the running ramdisk
18:01:01 <devananda> thank's everyone - we're out of time
18:01:07 <NobodyCam> thank you all
18:01:08 * lucasagomes goes to the channel
18:01:10 <jroll> thanks
18:01:11 <lucasagomes> thanks
18:01:12 <devananda> reminder - next week's meeting is alternate time!
18:01:17 <wanyen> I have to go to a differnt meeting:(
18:01:25 <devananda> #endmeeting