12:00:06 <alex_xu> #startmeeting nova api
12:00:07 <openstack> Meeting started Fri Jul 31 12:00:06 2015 UTC and is due to finish in 60 minutes.  The chair is alex_xu. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:00:10 <openstack> The meeting name has been set to 'nova_api'
12:00:23 <alex_xu> hello, anyone at here today?
12:00:40 <gmann_> o/
12:00:59 <alex_xu> gmann_: hello!
12:01:25 <kumar> hello
12:01:34 <alex_xu> kumar: welcome
12:03:02 <alex_xu> one more minute for waiting for more people
12:03:09 <gmann_> yea
12:03:34 <gmann_> johnthetubaguy might be away as he mentioned in ML
12:04:05 <alex_xu> gmann_: ok, I may hope johnthetubaguy and sdague can update some midcycle news to us
12:04:18 <gmann_> alex_xu: yea, that would be nice
12:04:26 <alex_xu> gmann_: if they aren't here, let do that next week
12:04:48 <gmann_> ye
12:04:50 <gmann_> yea
12:04:59 <sdague> hey folks
12:05:00 <alex_xu> #topic  status of liberty priority items
12:05:09 <alex_xu> sdague: hey, just waiting for you :)
12:05:37 <alex_xu> #link https://review.openstack.org/#/q/status:open+project:openstack/python-novaclient+branch:master+topic:bp/keypair-x509-certificates,n,z
12:05:54 <sdague> so I think the biggest items from the midcycle was the bits of the API that we are sure we need landed
12:06:06 <sdague> 1. the schema relaxation
12:06:20 <sdague> 2. the paste changes to default v2.0 on v2.1
12:06:29 <sdague> 3. the client changes
12:06:59 <alex_xu> client changes means microversion client?
12:07:58 <sdague> yes
12:08:02 <gmann_> sdague: 2 point, it means we can get rid of all v2 code also or later?
12:08:18 <sdague> gmann_: it should be deprecated if it's enabled
12:08:34 <sdague> we'll want some way to warn people if they have that code stack loaded
12:08:46 <sdague> I expect it will be a couple of cycles before we can delete it
12:09:02 <gmann_> sdague: ohk
12:09:43 <alex_xu> sdague: I saw https://etherpad.openstack.org/p/YVR-nova-liberty-summit-action-items mention remove extension, it looks like action for you?
12:09:52 <sdague> yes, I need to write up the document for that
12:10:11 <sdague> I'm working on a LinuxCon presentation today, but will try to do that next week
12:10:26 <alex_xu> sdague: ok, there is spec https://review.openstack.org/202900 , just initial version, so let me waiting for your doc.
12:10:37 <alex_xu> sdague: ok
12:10:43 <sdague> looking
12:11:22 <sdague> alex_xu: ok, I'll work from there, I think this is going to be more in the dev ref
12:11:35 <alex_xu> sdague: ok, no problem
12:12:02 <sdague> I did get the when you need a microversion guide up - http://docs.openstack.org/developer/nova/api_microversion_dev.html - now with flowchart
12:12:25 <alex_xu> sdague: thanks, that's cool
12:12:35 <gmann_> sdague: yea, flowchart was really nice
12:13:24 <alex_xu> sdague: so we can still add more thing, when we found more? looks like this case https://review.openstack.org/#/c/189544/ didn't include in it.
12:14:22 <sdague> alex_xu: sure, that's just a bug on the existing implementation, right?
12:14:57 <sdague> kenichi still has that as WIP
12:14:59 <alex_xu> sdague: that is provide more strict validation
12:15:58 <gmann_> sdague: alex_xu : Actually that came from https://bugs.launchpad.net/python-openstackclient/+bug/1434172
12:15:59 <openstack> Launchpad bug 1434172 in OpenStack Compute (nova) "security group create errors without description" [Low,In progress] - Assigned to Adrien Vergé (adrien-verge)
12:16:01 <uvirtbot> Launchpad bug 1434172 in nova "security group create errors without description" [Low,In progress]
12:16:02 <uvirtbot> Launchpad bug 1434172 in nova "security group create errors without description" [Low,In progress] https://launchpad.net/bugs/1434172
12:16:24 <sdague> gmann_: right, I thought that was mostly a client side issue
12:17:31 <alex_xu> gmann_: sdague , this is the server side fix https://review.openstack.org/#/c/192043/
12:17:36 <alex_xu> and that fix is used to backport
12:17:59 <alex_xu> so I think adding json-schema need new microverion
12:18:34 <sdague> so, description is not supposed to be manditory
12:19:04 <sdague> I feel like that's fixing the bug in the wrong direction
12:19:09 <gmann_> sdague: alex_xu : yea, that is optional but nova expected that mandatory
12:19:20 <sdague> gmann_: right, but the docs say it is optional
12:19:25 <gmann_> yea
12:19:26 <alex_xu> sdague: nova-network expect mandatory, but neutron not
12:19:34 <sdague> alex_xu: right, but that's a bug
12:19:40 <sdague> it's supposed to be optional
12:19:46 <sdague> and it used to be
12:19:50 <sdague> at some point it stopped being
12:20:00 <gmann_> alex_xu: :) that was main issue and its end up having json-schema adjustment etc
12:20:24 <alex_xu> sdague: ok...if we make it optional, then we need a microversion?
12:20:44 <sdague> alex_xu: honestly, I don't think so, because being non optional is a new buggy behavior
12:20:52 <sdague> we should fix it and back port the fix
12:20:57 <gmann_> alex_xu: IMO, its just bug and we just fix what Nova published in doc
12:21:00 <alex_xu> I guess that api behavior like that for a long time
12:21:15 <sdague> alex_xu: no, I think it was impacted by the jsonschema changes
12:21:29 <sdague> it really hasn't been like this for very long
12:22:31 * edleafe wanders in late
12:22:45 <alex_xu> sdague: it also a bug for v2 api?
12:22:50 <gmann_> sdague: we have same for v2 code base too. we do not have security grp schema just python code validation for name and description
12:22:58 <sdague> gmann_: hmmm...
12:23:08 <sdague> I really really thought we didn't need this before
12:23:26 <sdague> but if it's been here forever, I guess we can enforce it everywhere
12:24:23 <alex_xu> sdague: let me check the code later, and feedback on the patch. if it's been here forever, then we fix it in the right direction, right?
12:24:33 <sdague> alex_xu: correct
12:24:38 <alex_xu> sdague: ok
12:25:13 <alex_xu> any question on midcycle meetup? then we move on
12:25:13 <gmann_> https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/contrib/security_groups.py#L201
12:26:00 <alex_xu> gmann_: yea
12:26:12 <gmann_> alex_xu: also you might want to look -https://review.openstack.org/#/c/183909/5
12:26:18 <gmann_> which tried to fix that
12:27:12 <alex_xu> gmann_: similar fix with ken'ichi one?
12:27:36 <alex_xu> gmann_: let me check that later, let us move on
12:27:38 <gmann_> alex_xu: cannot remember actually, needs to look
12:27:40 <gmann_> ye
12:27:49 <alex_xu> gmann_: ok, thanks, will take a look
12:28:13 <alex_xu> #link https://review.openstack.org/#/q/status:open+project:openstack/python-novaclient+branch:master+topic:bp/keypair-x509-certificates,n,z
12:28:24 <alex_xu> ^ microversion client patch on good shape
12:28:57 <alex_xu> I work on those patches before andrey back from vacation
12:29:16 <alex_xu> hope we can merge them soon. we already have 2.9 on server side!
12:29:48 <alex_xu> #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/api-relax-validation,n,z
12:29:58 <alex_xu> this on good shape also
12:30:26 <alex_xu> sdague: gmann_ hope you can take a look at the check I added for ensure the test running on right wsgi stack
12:30:27 <gmann_> alex_xu: yea, thanks
12:30:36 <alex_xu> gmann_: np
12:30:43 <gmann_> alex_xu: ll look
12:30:49 <alex_xu> gmann_: thanks
12:31:04 <sdague> alex_xu: I'll try, I've got to get this presentation into shape before I really dive back into much code or review
12:31:33 <alex_xu> sdague: ok, no problem, no hurry
12:31:55 <alex_xu> #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:merge_sample_tests,n,z
12:32:04 <alex_xu> gmann_: ^ I think just left few patches
12:32:29 <alex_xu> same as the policy stuff
12:32:31 <alex_xu> #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/nova-api-policy-final-part,n,z
12:32:38 <gmann_> alex_xu: yea, but after those gets merged i need 2-3 patches to cleanup remaining thing some TODO things
12:32:49 <alex_xu> gmann_: cool
12:33:23 <alex_xu> ok, that is all priority items we tracked
12:33:26 <alex_xu> any question?
12:34:11 <alex_xu> ok, let move mon
12:34:16 <alex_xu> #topic open
12:34:28 <alex_xu> gmann_: ^ your turn, I saw some item from you
12:34:32 <gmann_> alex_xu: yea
12:34:39 <gmann_> sdague: lock patch- https://review.openstack.org/#/c/202391/
12:35:06 <gmann_> spec is merged
12:35:13 <sdague> oh, great, I'll take a look
12:35:19 <alex_xu> gmann_: did johnthetubaguy agree on we can merge this even after freeze?
12:35:30 <sdague> alex_xu: yes
12:35:51 <alex_xu> sdague: ok, cool, let me try help on review also
12:36:12 <gmann_> sdague: humm, actually we were on spec thing but i did not confirm from johnthetubaguy that we can merge or not
12:36:14 <sdague> ok, I just approved it
12:36:22 <gmann_> sdague: Thanks
12:36:28 <alex_xu> heh, cool
12:36:37 <sdague> I'm going to provide an update to the docs as a follow up patch
12:36:57 <gmann_> sdague: ohk, Thanks a lot
12:37:37 <gmann_> next if v2.1 VIF net-id bug - to remove VIF from v2.1 extension list - https://review.openstack.org/#/c/198934/
12:37:43 <gmann_> master one is approved
12:37:59 <gmann_> Need review for Kilo backport too - https://review.openstack.org/#/c/198944/
12:38:14 <gmann_> sdague: can you please look so that both can go in togehter
12:38:56 <alex_xu> gmann_: cool, thanks for the fix
12:39:20 <gmann_> but no hurry on this in case you busy with presentation :). i think this cn go later too as this is bug fix
12:39:24 <gmann_> alex_xu: np
12:39:54 <alex_xu> gmann_: yea, it is a bug fix, we didn't freeze for it
12:40:25 <alex_xu> any open?
12:40:48 <alex_xu> kumar: anything we can help?
12:40:49 <gmann_> alex_xu:  even I need to update doc also to mention the diff between v2 and v2.1. Will try to do next week
12:41:10 <alex_xu> gmann_: cool, that is useful
12:41:11 <kumar> #link https://review.openstack.org/#/c/183716/
12:42:45 <alex_xu> kumar: hope api team review on the api part?
12:44:16 <alex_xu> kumar: are you stil here?
12:44:17 <kumar> i have no idea on api review, but was looking for availability zones with cells
12:46:34 <alex_xu> kumar: looks like you should looking for help on nova weekly meeting. this meeting is more specific on nova api.
12:47:25 <kumar> alex_xu i have got this review on our api list - liberty design
12:47:48 <alex_xu> kumar: which list?
12:49:03 <kumar> under scheduler list
12:50:19 <sdague> kumar: REST API ?
12:52:45 <alex_xu> ...let free the meeting first.
12:52:56 <sdague> yeh, I think we're done
12:53:07 <alex_xu> kumar: let's move to openstack-nova channel, I guess we can get more help
12:53:16 <sdague> I guess the only other thing I had was about the Nova API docs
12:53:32 <sdague> I was reading them on the plane on the way home, and they are really really out of date at this point
12:53:41 <sdague> they basically describe the API circa 2011
12:53:58 <sdague> I really feel like we need to figure out how to get that to something more current
12:54:13 <alex_xu> sdague: you mean this https://github.com/russell/fairy-slipper?
12:54:13 <sdague> because if people have to read the Nova code to use the API, we've failed
12:54:36 <sdague> no
12:54:47 <gmann_> sdague: API ref docs
12:55:14 <alex_xu> sdague: this one http://developer.openstack.org/api-ref-compute-v2.1.html?
12:55:48 <sdague> I mean http://docs.openstack.org/developer/nova/v2/2.0_general_info.html
12:56:32 <sdague> I think there has been far too much focus on a magic solution here, but the concept guide is the thing that people actually need
12:56:34 <gmann_> alex_xu: sdague : yea even ref doc has some URL  wrong , I saw recently while doing JSON-Home stuff
12:57:09 <sdague> so, honestly, I feel like json home needs to go on the back burner until we explain what we have. Or have some metric on docs matching reality
12:57:16 <sdague> because the current state of things is really bad
12:59:05 <sdague> anyway, I'll kick off a thread about it next week. But we need a documentation plan, and I don't feel like we have one
12:59:18 <alex_xu> sdague: yea
12:59:42 <alex_xu> v2.1 is also mess
12:59:48 <gmann_> yea
13:00:07 <alex_xu> not sure https://github.com/russell/fairy-slipper can ready for Liberty
13:00:09 <sdague> so I honestly think, that after the bits to get v2.1 working and v2.0 on v2.1
13:00:20 <sdague> that writing documentation is the most important thing
13:00:45 <sdague> because we have a huge debt there
13:01:06 <alex_xu> sdague: yea, agree
13:01:46 <alex_xu> ok, let see your thread, then see how we can figure out a plan
13:01:48 <sdague> ok, got to run, thanks folks
13:01:52 <alex_xu> and the meeting is out of time
13:01:55 <gmann_> yea
13:02:01 <alex_xu> thanks all
13:02:08 <alex_xu> #endmeeting