14:00:03 <johnthetubaguy> #startmeeting nova
14:00:04 <openstack> Meeting started Thu Jan 22 14:00:03 2015 UTC and is due to finish in 60 minutes.  The chair is johnthetubaguy. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:08 <openstack> The meeting name has been set to 'nova'
14:00:16 <johnthetubaguy> https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting
14:00:23 <johnthetubaguy> lets do something similar
14:00:24 <bauzas> \o
14:00:31 <n0ano> o/
14:00:34 <johnthetubaguy> #topic Kilo Specs
14:00:36 <PaulMurray> hi
14:00:38 <gilliard> hi
14:01:13 <johnthetubaguy> so, sepcs
14:01:14 <gibi> o/
14:01:28 <alexpilotti> hello!
14:01:30 <johnthetubaguy> lots of exception requests, not actually got through them all myself
14:01:30 <gilliard> abhishekk mikal tjones cburgess jgrimm adrian_otto funzo mjturek jcook ekhugen irina_pov krtaylor danpb alexpilotti flip214 raildo jaypipes gilliard garyk edleafe dims moshele anteaya Nisha
14:01:30 <edleafe> o/
14:01:56 <ekhugen> o/
14:02:03 <flip214> gilliard: I think the wakeup-call would be better in #openstack-nova, in case some people are not always joined to -meeting.
14:02:05 <johnthetubaguy> gilliard: I did that in openstack-nova, thats the norm I think
14:02:15 <gilliard> OK
14:02:23 <johnthetubaguy> so any questions on specs process at this point?>
14:02:24 <gilliard> missed that, sorry
14:02:27 <bauzas> and then, someone invented calendars
14:02:36 <johnthetubaguy> #link http://lists.openstack.org/pipermail/openstack-dev/2015-January/054039.html
14:02:46 <johnthetubaguy> we have deadlines coming up frast
14:02:47 <johnthetubaguy> fast
14:02:50 <bauzas> are we planning a Spec review day ?
14:03:02 <garyk> hi
14:03:03 <bauzas> mikal said he was considering it last time
14:03:17 <bauzas> but we're only 2 days lefy
14:03:21 <johnthetubaguy> bauzas: nope, that time has passed, I was hoping for a code review day on friday, but no sure anyone has stepped up at this point
14:03:28 <bauzas> yeah
14:04:00 <johnthetubaguy> so 1/22 (today) is Feature Proposal Freeze day, we said
14:04:13 <bauzas> oh right
14:04:14 <johnthetubaguy> for all non priority code
14:04:30 <johnthetubaguy> #info FeatureProposalFreeze day today
14:04:33 <garyk> johnthetubaguy: is that specs or code?
14:04:37 <johnthetubaguy> garyk: code
14:04:50 <flip214> johnthetubaguy: is a spec, exception mail, and code up for review enough, or is there some other part of the process that needs to be followed?
14:04:56 <bauzas> johnthetubaguy: it's also the Spec Freeze Exception deadline nope ?
14:05:01 <johnthetubaguy> Feature freeze is 5th Feb
14:05:18 <johnthetubaguy> bauzas: in theory we said 5th Feb, I think, which was a bit dumb
14:05:27 <bauzas> flip214: FPF means that you can only iterate on your existing patches
14:05:29 <johnthetubaguy> flip214: for now, yes, I think so
14:05:39 <garyk> johnthetubaguy: why dumb. the fact that there is a meetup and everyone is in the same room could help
14:05:51 <johnthetubaguy> bauzas: right, its a grey area with a spec exception pending, and code already up
14:06:13 <johnthetubaguy> garyk: fair point
14:06:25 <johnthetubaguy> I mean we should have picked today really
14:06:38 <johnthetubaguy> but yeah, I wanted to have a chance to debate / fight at the mid cycle
14:06:47 <bauzas> garyk: not all people will be present, so it's like 2nd grade citizenship for people who can't attend and expect approvals
14:07:10 <johnthetubaguy> bauzas: well not totally, but I get your point
14:07:28 <bauzas> johnthetubaguy: agreed, wording was harsh
14:07:54 <johnthetubaguy> bauzas: its fair really, but hopefully its fairly even handed
14:07:55 <bauzas> johnthetubaguy: but you got my idea, if we allow the possibility to push up some reviews during midcycle, we need to also do that for remotees
14:08:14 <johnthetubaguy> bauzas: totally agreed
14:08:23 <johnthetubaguy> there are exceptions for anything remote or not
14:08:30 <bauzas> johnthetubaguy: +1
14:08:34 <johnthetubaguy> priority stuff can continue as normal
14:08:36 <bauzas> johnthetubaguy: hence the mail thing
14:08:45 <johnthetubaguy> right
14:08:59 <johnthetubaguy> anyways, we are going to a rat hole, all be it an important one
14:09:06 <johnthetubaguy> zooming back up...
14:09:14 <garyk> bauzas: it is not a matter of second class.
14:09:23 <johnthetubaguy> the goal here is to get more priority code merged, and be fair to everyone else
14:09:31 <garyk> the fact is there may be a number of cores present in one place and if one can reach out to them then it would be easier
14:09:42 <johnthetubaguy> by fair, when there is little chance of getting merged, we don't pretend it will still get into kilo
14:09:42 <garyk> i am sure yhat you can get a proxy to help you if it is importnat
14:09:58 <johnthetubaguy> garyk: thats fair
14:10:10 <johnthetubaguy> do email me if you want me to fight for something
14:10:18 <johnthetubaguy> but better
14:10:22 <dims__> o/
14:10:25 <johnthetubaguy> add your arguments in the spec review
14:10:37 <johnthetubaguy> so, any more questions on process?
14:10:40 <johnthetubaguy> dims__: you got one?
14:10:40 <edleafe> I think the point is we have to be aware not to ignore those not present at the meetup
14:10:51 <johnthetubaguy> edleafe: agreed
14:10:52 <dims__> johnthetubaguy: am good
14:10:59 <bauzas> finding the etherpad with all the asked exceptions
14:11:23 <johnthetubaguy> lets move on
14:11:29 <johnthetubaguy> #topic Priorities
14:11:48 <johnthetubaguy> #info priorities following general kilo featurefreeze dates
14:11:50 <garyk> bauzas: feel free to ping me too and i can try and get some poeople to look
14:12:34 <johnthetubaguy> https://etherpad.openstack.org/p/kilo-nova-priorities
14:12:39 <johnthetubaguy> anyone got things to talk about?
14:12:55 <johnthetubaguy> anteaya: you want to cover nova-network migration stuff?
14:13:20 <bauzas> garyk: thanks but I'll be in, I was just speaking for others' voices
14:13:34 * johnthetubaguy watches tumble weeds
14:13:36 <dims__> johnthetubaguy: ext4 ephemeral disk has been pending for a while - https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bug/1266262,n,z
14:13:37 <gilliard> #link https://etherpad.openstack.org/p/kilo-nova-priorities-tracking
14:13:47 <gilliard> That has a few links to reviews linked to the priorities
14:13:55 <gilliard> anteya added some links to specs
14:13:57 <bauzas> I need to update this etherpad
14:14:08 <johnthetubaguy> right, posted that above, but yes, please update the etherpad
14:14:25 <johnthetubaguy> if there is nothing to actively discuss lets move on
14:14:27 <bauzas> and I have a stuck review for a priority BP, but let's discuss it later
14:14:36 <johnthetubaguy> #topic Gate status
14:14:47 <johnthetubaguy> I don't see any folks around for that
14:15:04 <garyk> johnthetubaguy: back to priorities - we have features that were in review in juno and are still in review for kilo….
14:15:05 <bauzas> FYI, I proposed a change for adding cells v1 testing in check pipeline
14:15:36 <johnthetubaguy> seems like we need this: https://bugs.launchpad.net/nova/+bug/1410622
14:15:42 <johnthetubaguy> garyk: right, but they are not a priority
14:16:12 <garyk> johnthetubaguy: is there a list of the priority features?
14:16:13 <johnthetubaguy> well thats unfair, they could be
14:16:15 <johnthetubaguy> yes
14:16:29 <johnthetubaguy> garyk: its merged in the specs repo, as agreed at the summit
14:16:29 <dims__> johnthetubaguy: the last matt and i debugged was that the signatures on the wire is different from what keystone expects (boto)
14:16:51 <johnthetubaguy> well agreed technicaly in the nova-spec review that merged with the details that conainted the summit discussion
14:16:52 <garyk> johnthetubaguy: ack
14:16:54 <bauzas> dims__: the categorization rate is low, do you need help ?
14:17:14 <dims__> bauzas: yes
14:17:16 <johnthetubaguy> dims__: I think miridem was requesting help, if you could reach out to him, thats probably best
14:17:32 <bauzas> dims__: ack, will see what I can do
14:17:34 <dims__> johnthetubaguy: need someone with keystone-y skills
14:17:38 <dims__> thanks bauzas
14:17:46 <johnthetubaguy> dims__: gotcha
14:18:00 <bauzas> dims__: any idea why the rate is so low ? (<50%)
14:18:01 <johnthetubaguy> we should have someone drop in their meeting I guess
14:18:14 <bauzas> mriedem seems to be not present today
14:18:17 <dims__> bauzas: we've capped the boto version
14:18:30 <dims__> bauzas: till this gets fixed
14:18:38 <bauzas> dims__: but the logstash query is not categorized yet ?
14:18:40 <johnthetubaguy> right, fix the bleeding, sort the problem when we can
14:19:15 <johnthetubaguy> lets move on
14:19:21 <bauzas> dims__: I can understand that there is a boto problem related to keystone mw auth, but that shouldn't impact the rate, right ?
14:19:26 <johnthetubaguy> #topic Mid Cycle meetup
14:19:26 <bauzas> johnthetubaguy: ack
14:19:35 <johnthetubaguy> #link https://etherpad.openstack.org/p/kilo-nova-midcycle
14:19:41 <dims__> bauzas: let;s switch to nova channel
14:19:48 <bauzas> dims__: agreed
14:19:50 <johnthetubaguy> so mostly, we will discuss priorties, and keep them moving forward
14:20:02 <johnthetubaguy> bauzas: dims__: thank you for digging, appreciated
14:20:13 <johnthetubaguy> any one got questions on the mid cycle
14:20:23 <johnthetubaguy> mikal send an email via eventbright with directions, etc
14:20:32 <bauzas> eventbrite :)
14:20:40 <johnthetubaguy> lol, yeah, that
14:20:57 <johnthetubaguy> #topic Bugs
14:21:12 <johnthetubaguy> so dims__ you mentioned ext4, thats still an issue I guess?
14:21:18 <johnthetubaguy> did we agree its not a feature?
14:21:34 <dims__> treated as performance defect originally
14:21:51 <dims__> there was an email to operators mailing list as was requested here last time
14:22:10 <dims__> only feedback was, there should be a switch to not format the disk at all
14:22:32 <johnthetubaguy> ah, OK
14:22:41 <dims__> http://markmail.org/message/mmxa5kot7v6c27oy
14:23:08 <johnthetubaguy> hmm, its sucks, but I think a spec would help get agreement on this
14:23:28 <dims__> johnthetubaguy: last meeting i did create a blueprint as mentioned
14:23:32 <johnthetubaguy> but, honestly, this feels like something that should be user visible in the flavor
14:23:39 <johnthetubaguy> dims__: ah, sorry, I missed that
14:24:04 <dims__> johnthetubaguy: agreement was to log a blueprint and not do a full spec
14:24:14 <johnthetubaguy> I am OK with that
14:24:24 <johnthetubaguy> its just the spec would help pin down agreement I guess
14:24:32 <dims__> thanks, i will rebase and get the 2 reviews ready for cores
14:24:34 <johnthetubaguy> you got a link to the blueprint for the log?
14:24:35 <bauzas> possible midcycle topic ?
14:24:43 <dims__> one sec
14:25:04 <dims__> #link https://blueprints.launchpad.net/nova/+spec/ext4-as-ephemeral-disk
14:25:04 <johnthetubaguy> bauzas: yeah, its worth adding I guess, to move it along
14:25:13 <johnthetubaguy> dims__: sweet, thanks for the link
14:25:45 <bauzas> sounds like something manageable to discuss during midcycle and do feedback to operators afterwards
14:25:46 <johnthetubaguy> #link https://review.openstack.org/#/c/141883/
14:26:20 <bauzas> johnthetubaguy: yup ?
14:26:22 <johnthetubaguy> bauzas: yeah its a issue of changing the "ABI" in a way
14:26:23 <sean-k-mooney> hi i added this to the agenda to ask for some core review.
14:26:30 <johnthetubaguy> ah cools
14:26:35 <johnthetubaguy> it needs some review
14:26:37 <bauzas> sean-k-mooney: ah ok
14:26:42 <johnthetubaguy> which is a good segway...
14:26:47 <johnthetubaguy> #topic Stuck Reviews
14:26:52 <bauzas> eh
14:27:08 <bauzas> I got one
14:27:09 <bauzas> https://review.openstack.org/#/c/146913/4
14:27:11 <johnthetubaguy> OK, anyone got stuff they want to discuss, things with −2 or −1 thats an issue, etc?
14:27:22 <bauzas> that's technically a stuck change
14:27:35 <johnthetubaguy> looking
14:27:45 <johnthetubaguy> whats the dissagreement
14:27:54 <bauzas> long story short, we need to create a NovaObject for the image meta
14:28:09 <bauzas> properties field is the problem
14:28:16 <bauzas> I'm in favor of unversion it
14:28:22 <bauzas> because it's a mess atm
14:28:53 <bauzas> but the disagreement comes from the fact I've been pointed out to use another patch which versions it
14:28:58 <bauzas> thoughts ?
14:29:19 <bauzas> fix the mess ans wait, or unversion and go forward ?
14:29:38 <johnthetubaguy> can we add the extra change after your work?
14:29:40 <bauzas> tbp, I don't need an answer *now*
14:29:50 <johnthetubaguy> so yours puts the unversion, then add the version stuff on top?
14:29:56 <bauzas> I'm pointing the review to get feedback in the change itself
14:30:05 <bauzas> johnthetubaguy: that's just bumping a version
14:30:05 <johnthetubaguy> right
14:30:26 <johnthetubaguy> I need to read more to be clear I guess
14:30:29 <bauzas> we can move on, I shed the light on my pathc :)
14:30:34 <johnthetubaguy> good to raise it, yeah
14:30:39 <johnthetubaguy> any more stuck reviews?
14:30:41 * johnthetubaguy ducks
14:31:07 <johnthetubaguy> oh wait, we have some links from last week
14:31:23 <sean-k-mooney> its not technically stuch but if you have time to reveiw #link https://review.openstack.org/#/c/141883
14:31:37 <Nisha> johnthetubaguy, need reviews on https://review.openstack.org/133534
14:31:44 <johnthetubaguy> https://review.openstack.org/#/c/141010/
14:31:51 <Nisha> johnthetubaguy, yes
14:32:11 <johnthetubaguy> bauzas: I think you have good comments on there
14:32:12 <Nisha> ^^^ above reviews are required fr ironic
14:32:34 <bauzas> johnthetubaguy: yeah, the patch you mentioned is very similar to https://review.openstack.org/#/c/141883
14:32:37 <dims__> johnthetubaguy: Minesweeper has blessed https://review.openstack.org/#/c/142118/ (CRITICAL bug)
14:32:57 <bauzas> johnthetubaguy: I personnaly dislike any string hackling like we previously did
14:33:06 <bauzas> hacking*
14:33:13 <Nisha> johnthetubaguy, https://review.openstack.org/#/c/141010/ is required for ironic
14:33:15 <bauzas> we need to be more robust
14:33:21 <johnthetubaguy> bauzas: got it, you don't want to add more mess using the old method
14:33:28 <Nisha> bauzas, its not hacking in nova
14:33:40 <bauzas> Nisha: it's hacking a string, correct
14:33:47 <johnthetubaguy> Nisha: problem is we are rewriting that bit of code
14:33:50 <Nisha> bauzas, no
14:34:06 <johnthetubaguy> right now its not very well versioned, and we are making lots of things more explicit
14:34:14 <Nisha> bauzas, https://review.openstack.org/133534 the related spec for same
14:34:16 <johnthetubaguy> passing objects not strings, etc
14:34:35 <johnthetubaguy> right, we probably need to discuss the issues on the spec
14:34:51 <johnthetubaguy> Nisha: did you request a spec freeze exception?
14:35:01 <Nisha> johnthetubaguy, yes...after discussion with Alexis Lee on the spec he was fine and
14:35:03 <bauzas> Nisha: see my comments on https://review.openstack.org/#/c/141883
14:35:13 <Nisha> johnthetubaguy, yes i do sent a mail to mailing list
14:35:30 <bauzas> Nisha: we support a string as of now, we'll still support a string
14:35:40 <johnthetubaguy> Nisha: I couldn't remember if I saw it thats, cool, so we need to view the spec and agree what we can do
14:35:48 <bauzas> Nisha: but the difference is that the format of the string needs to be safier
14:35:53 <bauzas> johnthetubaguy: agreed
14:36:14 <johnthetubaguy> Nisha: honestly, given our kilo priorities, we might want to defer to L so we can get the code we need in place, but I need to give it more thought
14:36:15 <Nisha> bauzas, i think the misunderstanding from the spec /code is that nova filter will support list
14:36:39 <johnthetubaguy> hmm, thats a point
14:36:48 <johnthetubaguy> anyways, lets look at that offline more
14:36:52 <bauzas> Nisha: let me provide further explanations in the spec
14:36:54 <Nisha> while the list (a string seperated with delimiters aprt from comma) is in ironic
14:36:54 <johnthetubaguy> #topic Open Disscussion
14:37:04 <johnthetubaguy> so any more to discuss?
14:37:07 <flip214> yeah.
14:37:17 <flip214> does anyone know of a diagram about data flow within openstack?
14:37:19 <Nisha> johnthetubaguy, when can i catch you guys
14:37:24 <Nisha> what time means
14:37:36 <flip214> which function gets called in which order, and what data gets passed from one to the other?
14:37:43 <johnthetubaguy> flip214: the docs have a nice diagram from ken pepple, did you see it?
14:37:56 <johnthetubaguy> flip214: its not that explicit, you need to read the code for the functions
14:38:02 <bauzas> johnthetubaguy: is it still relevant ?
14:38:12 <flip214> no, not yet. I just noticed that many docs are outdated.
14:38:28 <baoli> johnthetubaguy: where is the list of spec freeze exceptions?
14:38:34 <johnthetubaguy> updates welcome, which isn't a great answer, but the overall stuff is quite accurage
14:38:34 <Nisha> johnthetubaguy, need review on https://review.openstack.org/141012
14:38:38 <bauzas> I saw some diagrams when googling but they're now pretty old and don't explain the conductor nor the objects
14:38:48 <johnthetubaguy> baoli: I am using my email inbox right now, there is an etherpad somewhere
14:39:07 <johnthetubaguy> #help update our docs to include more conductor and objects info
14:39:13 <bauzas> flip214: honestly, take a starting point, like the API endpoint for booting, and follow the golden path
14:39:35 <gilliard> you  can raise bugs for docs, too
14:39:36 <flip214> bauzas: that's not easy, because of the many indirections.
14:39:45 <johnthetubaguy> flip214: what bauzas said, just understand, API, DB, queue, compute, conductor, scheduler, and thats the key
14:39:55 <Nisha> johnthetubaguy, bauzas , needed review on https://review.openstack.org/141012 . The spec for same is merged and code is placed for review long back but no reviews yet....
14:39:59 <johnthetubaguy> flip214: we should take this into the nova channel after the meeting
14:40:07 <baoli> johnthetubaguy: I see. I guess that https://review.openstack.org/#/c/136077/ is there?
14:40:29 <bauzas> gilliard: fair point
14:40:46 <johnthetubaguy> Nisha: please update your blueprint to say it NeedsCodeReview, if you have all your code up for review
14:40:57 <johnthetubaguy> we should really look at that in the code push
14:41:23 <Nisha> johnthetubaguy, ok. thanks
14:42:32 <johnthetubaguy> any more for any more?
14:42:36 <Nisha> johnthetubaguy, done
14:43:05 <Nisha> johnthetubaguy, i will like to take up the flavor spec in the channel after this meeting...
14:43:16 <Nisha> if that helps to clear the things
14:43:26 <johnthetubaguy> OK, sounds good
14:43:27 <johnthetubaguy> thanks all
14:43:42 <johnthetubaguy> safe journey to the mid cylce if you are heading there
14:43:45 <johnthetubaguy> #endmeeting