17:00:09 <jroll> #startmeeting ironic
17:00:09 <openstack> Meeting started Mon Oct  5 17:00:09 2015 UTC and is due to finish in 60 minutes.  The chair is jroll. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:12 <openstack> The meeting name has been set to 'ironic'
17:00:16 <vdrok_> o/
17:00:17 <betherly> o/
17:00:18 <jroll> who's here for the ironic meeting?
17:00:21 <devananda> o/
17:00:24 <xek> o/
17:00:26 <[1]cdearborn> o/
17:00:26 <gabriel> o/
17:00:26 <wanyen> o/
17:00:26 <dtantsur> o/
17:00:27 <NobodyCam> good UGT morning all
17:00:30 <jlvillal> o/
17:00:33 <jroll> #chair devananda NobodyCam
17:00:33 <openstack> Current chairs: NobodyCam devananda jroll
17:01:00 <TheJulia> o/
17:01:01 <liliars> o/
17:01:05 <krtaylor> o/
17:01:10 <rloo> o/
17:01:10 <jroll> as always, the agenda can be found here:
17:01:12 <jroll> #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting
17:01:19 <jroll> super light today
17:01:22 <lucasagomes> o/  (here for few minutes only)
17:01:28 <jroll> #topic announcements
17:02:01 <jroll> we'll be releasing ironic 4.2.1 this week, to pick up a bug fix that was backported to liberty, and some translations
17:02:05 <jroll> #link https://review.openstack.org/230749
17:02:08 <trown> o/
17:02:14 <NobodyCam> nice :)
17:02:18 <Nisha> o/
17:02:28 <jroll> we're also planning on the first ironic-python-agent release this week, at 1.0.0 \o/
17:02:31 <jroll> #link https://review.openstack.org/229986
17:02:40 <NobodyCam> awesome!
17:02:43 <jroll> that will be the basis of a stable/liberty branch as well
17:03:03 <devananda> fwiw, I just found a backwards incompatibility bug in Liberty agent and pxe deploy drivers. I'm going to try to get that fixed ASAP
17:03:07 <jroll> I should info those I guess
17:03:12 <NobodyCam> did we cut bifrost release last week???
17:03:14 <jroll> #info ironic 4.2.1 release this week
17:03:23 <jroll> #info ironic-python-agent 1.0.0 release this week
17:03:32 <TheJulia> NobodyCam: week before if memory serves
17:03:37 <jlvillal> And ironic-lib last week?
17:03:38 <NobodyCam> ahh :)
17:03:39 <jroll> devananda: thank you for doing that
17:03:45 <jroll> does anyone else have any announcements today?
17:04:10 <xek> I added 3 bugs to address rolling upgrades issues in Mitaka
17:04:12 <rloo> do you want to remind about summit session/topics
17:04:14 <NobodyCam> summit like three weeks away
17:04:25 <jroll> yes, thanks rloo
17:04:31 <xek> https://bugs.launchpad.net/ironic?field.searchtext=rolling+upgrades
17:04:57 <lucasagomes> xek, nice one, thanks for that
17:05:00 <jroll> #info please submit your design session proposals. deadline is october 9. https://etherpad.openstack.org/p/mitaka-ironic-design-summit-ideas
17:05:07 <trown> woot woot on IPA release
17:05:20 <NobodyCam> #link https://etherpad.openstack.org/p/mitaka-ironic-design-summit-ideas
17:05:34 <jroll> thanks xek
17:05:50 <jroll> anything else?
17:06:03 <dtantsur> I've updated the driver composition spec
17:06:04 <Nisha> jroll, do we need to add the specs which were submitted during Liberty but didnt went through to the etherpad? or they will be discussed even without adding them on the etherpad?
17:06:07 <dtantsur> #link https://review.openstack.org/#/c/188370/
17:06:13 <dtantsur> must be an interesting reading :D
17:06:28 <jroll> Nisha: no need to re-approve specs, if that's your question?
17:06:35 <dtantsur> Nisha, only if they really need *discussion*, not just getting code landed
17:07:10 <jlvillal> jroll: ironic-lib 0.2.0 release last week. If you want to info that.
17:07:14 <Nisha> jroll, the spec didnt go thru
17:07:17 <Nisha> in liberty
17:07:18 <jroll> Nisha: let's come back to that in open discussion
17:07:25 <Nisha> jroll, SURE
17:07:40 <jroll> #info ironic-lib 0.2.0 released last week
17:07:46 <lucasagomes> Nisha, jroll sent an email to the ML about it
17:07:46 <jroll> much info, wow
17:07:51 <jroll> let's move on
17:07:59 <NobodyCam> :)
17:08:00 <jroll> #topic subteam status updates
17:08:04 <Nisha> lucasagomes, ok
17:08:08 <jroll> as usual, these can be found at
17:08:15 <jroll> #link https://etherpad.openstack.org/p/IronicWhiteBoard
17:09:29 <jroll> I don't see any surprises here, does anyone else have questions/comments/concerns?
17:09:35 * jroll gives folks a couple minutes to read
17:09:52 <dtantsur> we need to finish syncing ironic-lib from ironic... finally..
17:10:00 <jroll> yes
17:10:04 <jlvillal> +1
17:10:07 <devananda> dtantsur: ++
17:10:20 <jroll> dtantsur: there's a few ironic-lib patches to fix some out-of-sync things there, we can release again once those land
17:10:30 <rloo> dtantsur: wrt ironic-lib, i thought ramesh was going to push up a patch. does anyone know what patches are outstanding/need to be reviewed?
17:10:31 <devananda> jroll: on the neutron work, what does the upstream / automated testing look like at this point?
17:11:02 <jroll> rloo: https://review.openstack.org/#/q/project:openstack/ironic-lib+status:open,n,z
17:11:12 <jroll> devananda: I haven't seen anything yet
17:11:17 <devananda> it would be helpful to tag all the ironic-lib related work, across all projects, with a consistent topic so we can see them all at once in gerrit
17:11:23 <dtantsur> rloo, for now we review everything in ironic-lib. after that we can release - bump g-r - port ironic
17:11:32 <dtantsur> devananda++
17:11:54 <rloo> jroll, dtantsur: i was thinking of the ironic patch to use ironic-lib.
17:11:56 <devananda> jroll: are there instructions for how to reproduce tests locally?
17:12:11 <devananda> jroll: like, for the network stuff, how do i validate it?
17:12:41 <jroll> devananda: I'm not sure, Sukhdev? ^^
17:12:42 <jlvillal> rloo: ramesh pointed me to: https://review.openstack.org/184443
17:13:07 * lucasagomes have to go, see y'all
17:13:10 <rloo> i really hope the patches to ironic-lib are only for sync'ing between ironic & lib, not new stuff
17:13:14 <NobodyCam> night lucasagomes
17:13:27 <jlvillal> rloo: My patches are just syncing
17:13:46 <devananda> jroll: also, what's the status of pyhton-ironicclient work to add the portgroup commands?
17:13:59 <jlvillal> rloo: Except the iteritems one...
17:14:06 <jroll> devananda: there's a patch, it mostly works
17:14:10 <devananda> nice
17:14:18 <gabriel> OneView driver updates: we would love to have some reviews in our patch :)  https://review.openstack.org/#/c/191822/
17:14:31 <jroll> ^^ we should get that in soon
17:14:47 <rloo> jlvillal: maybe defer that one until we get ironic using it? otherwise, adds one more 'thing' to check when looking at discrepancies between ironic & lib
17:15:08 <Sukhdev> jroll devananda: most of the testing as of now is manual - once all is well cooked, next will be to write some automated tests
17:15:10 <jlvillal> rloo: sure
17:15:30 <rloo> thx jlvillal
17:15:51 <jroll> Sukhdev: I'd love to have that in the gate before we land this, even if it's non-voting. we can come back to that...
17:16:10 <devananda> jroll: Sukhdev: ++
17:16:42 <jroll> anything else on subteam updates?
17:16:44 <devananda> it looks like client and nova patches use a different gerrit topic, so I added links to the 'pad
17:16:52 <jroll> thanks
17:16:55 <NobodyCam> ty devananda
17:17:15 <gabriel> We (oneview guy) have worked on ramesh's reviews. It is ready for new reviews.
17:17:18 <jroll> actually, anything else can continue in open discussion
17:17:24 <jroll> #topic open discussion
17:17:46 <jroll> Nisha: so, can you clarify your question?
17:18:43 <jroll> (if it isn't answered already)
17:18:53 <Nisha> jroll, as i said in channel...i was just asking if priorities gets decided based on summit discussions?
17:19:29 <Nisha> jroll, so i think for seting up priorities for Mitaka cycle even the specs which are not at Mitaka etherpad would be considered
17:19:34 <jroll> Nisha: I expect we'll spend some time going over mitaka priorities during the friday session, however the priorities list is always changing
17:19:40 <jroll> we'll be reviewing all specs/blueprints for that
17:19:51 <Nisha> jroll, thanks that answers my ques
17:20:04 <jroll> Nisha: the etherpad is only for sessions about a topic, not a list of things to work on in mitaka :)
17:20:16 <Nisha> jroll, :)
17:20:54 <devananda> and to repeat - previously approved but not completed specs do not need another round of approvals (unless they're special, in which case, we'll reach out to the author)
17:21:04 * jroll wonders out loud if we can finish the ironic-lib transition before the summit
17:21:14 * jroll also thinks we should land the oneview driver before the summit
17:21:22 * devananda thinks that both of those would be great
17:21:33 <wanyen> Jroll: For specs not yet got approved in Liberty, do they needs to be re-submitted for Mitaka?
17:21:34 <jlvillal> jroll: I think ironic-lib transition is possible
17:21:44 * thiagop would be very happy with that
17:21:46 <jroll> wanyen: no, I sent an email about that last week
17:21:58 <wanyen> jroll, oh. tx
17:22:07 <jroll> please follow the mailing list, that's where the important things go
17:22:13 <devananda> jroll: for completed specs, we had a little discussion around my patch last week
17:22:16 <gabriel> thank you a lot, jroll :)
17:22:31 <jroll> devananda: ya?
17:22:57 <devananda> tldr; should the spec for completed work get moved to the folder for the cycle it was implemented with or without a placeholder in the /approved/ folder ?
17:23:24 <jroll> oh, that
17:23:32 <NobodyCam> :)
17:23:53 <jroll> so one of the main reasons I like the placeholder, is that I set blueprint links to the approved URL
17:24:22 <devananda> jroll: I agree with the usefulness of that launchpad link within the context of launchpad
17:24:26 <jroll> it's certainly possible to fix that after the spec move, but is yet another thing to do
17:24:44 <devananda> jroll: but then, the /approved/ folder grows indefinitely and ceases to represent the current work backlog
17:24:46 <jroll> devananda: keep in mind our release emails/notes have launchpad links
17:24:47 <rloo> personally, i think the specs shouldn't get moved at all. seems odd to be changing the URL as to where the spec is accessed. we 'just' need a way of classifying/showing them as <cycle>implemented on the table of contents or whatever.
17:24:58 <thiagop> The approved folder would become very large with all the specs on it, right?
17:25:10 <jroll> yeah :(
17:25:23 <devananda> I also don't like changing the links. but I do like having a single place to look for "what is the ongoing work"
17:25:26 <jroll> that's a good point
17:25:48 <jroll> devananda: I'd probably argue that launchpad should be the source of truth for that
17:25:58 <gabriel> we can place symbolic links on an implemented or implementing directory to the approved spec
17:26:21 <devananda> rloo: the TOC is built from folders, not manually edited
17:26:24 * jroll wonders if that works with the publishing job
17:26:32 <jlvillal> I could see everything goes in one folder, and that could be the link that does not change. But then use symbolic links to arrange things.
17:26:36 <devananda> rloo: so if we stopped moving them into /XX-implemented/ then the TOC would not get updated
17:26:49 <devananda> jroll: I think you know how I feel about LP ....
17:26:58 <rloo> devananda: i wonder if there is some other way to build the toc, or can we use symbolic links as gabriel suggested
17:27:21 <jroll> devananda: it doesn't matter how we feel about LP; it's what we use.
17:27:51 <rloo> it seems like we have a 'db' of files, and they each have states associated with them, and we want to display them in various ways...
17:27:58 <jroll> devananda: I don't see any sense in having a different source of truth for ongoing work vs bugs, release tracking, etc
17:28:02 <devananda> fwiw, a few of you may want to talk with / look at Nova's process for this, and what several other projects are doing
17:28:24 <devananda> it generally requires deleting or marking "not-implemented" lots of things and making folks go through yet another round of reviews at the start of the next cycle
17:28:25 <jroll> indeed
17:28:31 <devananda> not doing ^ is a benefit of our current approach
17:28:49 <thiagop> devananda: +1(000)
17:28:57 <devananda> perhaps this would be a good summit topic? :)
17:29:13 <jroll> ++
17:29:15 <rloo> ewwww, not sure i want to spend one of our topics on this.
17:29:26 <jroll> because I also wanted to talk about how we track work in general
17:29:32 <jroll> e.g. requiring a blueprint for any feature work
17:29:35 <rloo> can't someone propose/email on that stuff?
17:29:44 <jroll> everything is super unorganized right now
17:29:49 <devananda> rloo: I've spent several hours on this with other projects // in the cross project track, in previous summits. you should come join the fun ;)
17:30:04 <devananda> jroll: ++
17:30:06 <rloo> devananda: clearly, our idea of fun is different :)
17:30:23 <jroll> rloo: I think this would be a broader "let's get organized" track. which could be useful. think about all the times you've wondered what you should be reviewing
17:30:34 <rloo> devananda: i think if it does make sense, for projects to be consistent with how specs/etc are handled.
17:30:49 <NobodyCam> jroll: ++ sounds like a good track to me
17:30:51 <rloo> jroll: just tell me what i should be reviewing.
17:31:07 <jroll> rloo: I don't know because everything is a mess :((((
17:31:13 <rloo> jroll: guess for me, i'd prefer technical discussions at the summit.
17:31:58 <devananda> jroll: added to the bottom of the pad with your name on it ;)
17:31:59 <jroll> rloo: I tend to think team dynamics and getting everyone on the same page is just as important as the tech
17:32:03 <jroll> devananda: heh thanks
17:32:32 <rloo> jroll: oh it is important, but we're a community that isn't all present at the summit.
17:32:42 <rloo> jroll: and such processes involve everyone contributing.
17:32:59 <jroll> rloo: indeed, but we can get general consensus and propose that back after the summit
17:33:26 <jroll> the non-technical pieces are just as difficult to talk about in IRC, IMO
17:33:31 <rloo> jroll: I'm just voicing my opinion. if others feel like we should discuss, then i am outvoted.
17:33:58 <jroll> rloo: yeah, well, the session is now proposed, we can decide next week whether to include it
17:34:05 <jroll> fwiw, we have ~12 proposals and 8 slots
17:34:28 <jroll> (so far)
17:34:34 <thiagop> wow
17:35:30 <thiagop> jroll: is that a pad or something with the proposals?
17:35:48 <jroll> thiagop: https://etherpad.openstack.org/p/mitaka-ironic-design-summit-ideas
17:36:03 * rloo wonders if people read the emails or understand what they are about
17:36:23 <jroll> heh
17:36:28 <devananda> jroll: 14 by my count
17:36:37 <rloo> seriously, maybe there is a communication issue
17:37:02 <thiagop> rloo: not always for me...
17:37:11 <jroll> devananda: 7 and 8 are the same IMO
17:37:40 <jroll> anyway, I must have miscounted earlier this morning :P
17:38:06 <jlvillal> Random question: Is there a reason we don't use the requests library in python-ironicclient?
17:38:09 <rloo> well, we don't really need to count them, but who has numbered them already?
17:38:24 <jroll> I believe deva just numbered them
17:38:30 <jroll> jlvillal: don't we use something from oslo?
17:38:48 <jlvillal> oslo.utils?
17:38:49 <rloo> :)
17:39:05 <jroll> jlvillal: yeah, and it uses requests https://github.com/openstack/python-ironicclient/blob/master/ironicclient/openstack/common/apiclient/client.py#L39
17:39:43 <jlvillal> jroll: Thanks. Looking at: https://review.openstack.org/#/c/229286/   Which is trying to use proxy support. And requests supports proxies directly....
17:40:29 * jlvillal has taken a detoru...
17:40:30 <jroll> jlvillal: idk
17:40:38 <jlvillal> s/detoru/detour/
17:40:51 <jroll> anything else folks want to talk about?
17:41:13 <NobodyCam> i'm good taday
17:41:19 <NobodyCam> today even :p
17:41:34 <rloo> i was thinking about Nisha's questions. would be good to try to articulate how things get prioritized. (not exactly now but)
17:42:11 <jroll> rloo: and that goes back to: I'm not sure we even know how we prioritize things :|
17:42:29 <rloo> jroll: i thought you and deva did it over drinks :)
17:42:40 <Nisha> rloo :D
17:42:50 <jroll> rloo: unfortunately, no, we get way off topic when that happens
17:42:50 <rloo> for me, 'infrastructure' is first.
17:42:52 <jroll> :P
17:43:11 <rloo> jroll: :D
17:43:26 <jroll> alright, should we wrap this up?
17:43:32 <jroll> going once...
17:43:32 <jlvillal> Works for me :)
17:43:40 <jroll> thanks everyone :)
17:43:42 <jroll> #endmeeting