12:00:02 <alex_xu> #startmeeting nova api
12:00:03 <openstack> Meeting started Tue Nov 24 12:00:02 2015 UTC and is due to finish in 60 minutes.  The chair is alex_xu. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:04 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:00:06 <openstack> The meeting name has been set to 'nova_api'
12:00:12 <alex_xu> who is here today?
12:00:19 <Kevin_Zheng> hi
12:00:22 <gmann_> hi
12:00:23 <oomichi> hi
12:00:42 <alex_xu> hello everyone
12:01:03 <alex_xu> let's start the meeting
12:01:10 <alex_xu> #topic actions from last meeting
12:01:10 <edleafe> o/
12:01:14 * oomichi all kids still wake up even now 9pm..
12:01:19 <alex_xu> sdague ask docs team about approval rights by api subteam on our wadl
12:01:27 <alex_xu> oomichi: heh
12:01:28 <jichen> o/
12:01:36 <alex_xu> sdague: are you around?
12:01:59 <gmann_> alex_xu: cool
12:02:09 * johnthetubaguy waves hello
12:02:18 <alex_xu> I guess not, let me check with him, when he is up
12:02:24 <edleafe> alex_xu: this is more or less a holiday week in the US
12:02:30 <alex_xu> gmann_: yea, if we get right, that will more easy for us
12:02:31 <edleafe> he might not be around
12:02:41 <alex_xu> edleafe: ah, thanks
12:02:47 <alex_xu> edleafe: why you are still working :)
12:02:51 <gmann_> alex_xu: yes, that will be helpful to fast the things
12:03:02 <alex_xu> so let's move on
12:03:07 <alex_xu> #topic content patches up for review
12:03:08 <edleafe> alex_xu: I keep asking myself that :)
12:03:19 <gmann_> edleafe: heh :)
12:03:19 <alex_xu> #link https://etherpad.openstack.org/p/mitaka-nova-priorities-tracking
12:03:27 <alex_xu> we have a set of patches ready for review
12:03:40 <alex_xu> so please review them
12:04:02 <alex_xu> I'm thinking for the service and live-migration I can ask scheduler and live-migration team members help to take a look at.
12:04:03 <oomichi> nice progress :)
12:04:08 <gmann_> alex_xu: i will start allocating my time from tomorrow, was busy in tempest things.
12:04:22 <alex_xu> gmann_: thanks
12:04:33 <alex_xu> And there is api-ref patch up https://review.openstack.org/248534
12:04:55 <alex_xu> I hope this can be an example of what api-ref expected, so please help me!
12:05:11 <alex_xu> Basically as my understand the api-ref is focus on the API behavior. The concept doc is focus on the use-case
12:05:17 <edleafe> alex_xu: I'll try to find time today for that
12:05:29 <alex_xu> edleafe: thanks
12:05:53 <gmann_> alex_xu: how about giving ref of concept guide in API description in api-ref ?
12:06:31 <alex_xu> gmann_: yea, maybe on the top of api ref, we needn't link which api to one concept guide
12:06:37 <alex_xu> s/which/each
12:07:06 <johnthetubaguy> we could, but that does feel like overkill I guess
12:07:29 <gmann_> johnthetubaguy: but that gives benefit of reading all from one place
12:07:32 <johnthetubaguy> would be nice to link to the live-migrate use cases, etc, in the concept guide from the complete ref, I supose
12:08:10 <gmann_> yea otherway around also helps
12:08:52 <johnthetubaguy> I think it will become clearer once we fill out one in full
12:09:12 <alex_xu> yea, just keep fill the doc
12:09:16 <johnthetubaguy> I know how alex_xu is doing just that, lets give live-migrate looking good, and see how it looks
12:09:30 <johnthetubaguy> s/I know how/I like how/
12:09:41 <gmann_> ok.
12:09:46 <alex_xu> yea, please help me make it better, them we have template for api ref
12:10:09 <alex_xu> so let's move on
12:10:12 <alex_xu> #topic most needed next content patches
12:10:19 <alex_xu> So how about we select topic for people focus on next week? For example we focus on this one https://github.com/openstack/nova/blame/master/api-guide/source/general_info.rst#L115
12:11:08 <jichen> +1
12:11:55 <johnthetubaguy> the alternative, I guess, is to all pick different bits, like with the API ref, but either way should work
12:12:41 <alex_xu> yea, based on people flavor
12:13:26 <gmann_> yea
12:13:46 <alex_xu> I just thought if people can focus one area, then review quick, feedback quick.
12:13:56 <alex_xu> but anyway either way works
12:14:28 <edleafe> we should pick the parts to work on so that we don't duplicate effort
12:14:59 <alex_xu> edleafe: put the part you work on in concept doc section at https://etherpad.openstack.org/p/nova-v2.1-api-doc
12:15:33 <alex_xu> most of time we use this way to sync work.
12:15:43 <edleafe> alex_xu: great!
12:16:09 * alex_xu just quck add note to etherpad
12:16:40 <jichen> alex_xu: there is a section in https://etherpad.openstack.org/p/mitaka-nova-priorities-tracking
12:16:45 * edleafe just signed up for a couple
12:16:45 <alex_xu> edleafe: ah, I see you take the host one, thanks :)
12:17:18 <alex_xu> jichen: that is just for patch ready for review I think, we shouldn't put too much our work detail to confuse other review
12:17:34 <johnthetubaguy> should we not be sure to only take the ones we are working on?
12:17:36 <alex_xu> s/other review/other reviewer/
12:18:02 <jichen> alex_xu: ok,
12:18:25 <alex_xu> johnthetubaguy: yea, one or two, not two much, release timely when you aren't work on anymore
12:18:37 <alex_xu> s/not two much/not too much/
12:18:40 <johnthetubaguy> alex_xu: cool, sounds like we are agreed
12:19:01 <johnthetubaguy> well, if you have patches up for seven, thats cool, but yeah
12:19:18 <alex_xu> yea
12:19:22 <alex_xu> so let's move on
12:19:27 <alex_xu> #topic API futures - specs
12:19:51 <alex_xu> any spec people want to bring up, I didn't get time select some
12:20:09 <oomichi> can I ?
12:20:21 <alex_xu> oomichi: sure, please
12:20:28 <oomichi> #link https://review.openstack.org/#/c/249167/
12:20:53 <oomichi> we have api-wg guideline and some specs don't fit to the guideline
12:20:57 <johnthetubaguy> yeah, need to agree the way forward for the pause and cancel live-migrate
12:21:01 <oomichi> s/and/but/
12:21:09 <alex_xu> oomichi: good catch!
12:21:13 <oomichi> the above patch fixes one
12:21:42 <gmann_> oomichi: ahh, yea
12:21:51 <johnthetubaguy> oomichi: which spec from API working group is that in?
12:22:14 <oomichi> I am wondering how about containing the part of the guideline on spec template.
12:22:15 <oomichi> it is difficult to know the guideline by all people
12:22:24 <oomichi> johnthetubaguy: sorry, cannot understand the comment
12:22:47 <alex_xu> #link https://github.com/openstack/api-wg/blob/master/guidelines/naming.rst
12:22:53 <alex_xu> I guess this one ^
12:22:57 <johnthetubaguy> oomichi: sorry, I was meaning which API WG doc tells us about the rules we need to follow, is it this one? http://specs.openstack.org/openstack/api-wg/guidelines/naming.html
12:23:08 <johnthetubaguy> ah, yeah, looking at the same thing
12:23:23 <oomichi> johnthetubaguy: yeah, you are right
12:23:25 <oomichi> alex_xu: you also
12:23:31 <alex_xu> johnthetubaguy: yea, that's one, at the bottom of doc
12:24:16 <johnthetubaguy> hmm, that is messy, its the opposite to most of our current action names, but agreed with the approach
12:24:35 <oomichi> this kind of violation seems common, I feel
12:24:42 <gmann_> johnthetubaguy: yea, many action names we changed back to v2 style in v2.1
12:25:09 <alex_xu> johnthetubaguy: nvm....we have a lot of still in nova. like 'os-liveMigrate'
12:25:13 <oomichi> johnthetubaguy: yeah, right. nova api is nice practice for getting better api design :)
12:25:15 <alex_xu> s/still/style/
12:25:22 <johnthetubaguy> do we have any that follow the new style yet?
12:25:40 <johnthetubaguy> action names that is
12:26:26 <alex_xu> johnthetubaguy: sorry, I didn't get you
12:26:26 <oomichi> johnthetubaguy: I'm fixing the api on nova-spec reviews, and trigger_crash_dump is one
12:26:39 <johnthetubaguy> I am just looking at the current API
12:26:55 <johnthetubaguy> we don't seem to have any server actions that follow the API WG rules
12:27:15 <johnthetubaguy> so following the rules is going to make our API even more inconsistent, which seems like a worry
12:27:16 <jichen> yeah, do a quick look and seems no
12:27:47 <johnthetubaguy> we seem to have camalCase or os-camalCase
12:27:48 <gmann_> johnthetubaguy: yea, we changed those back in v2.1 (in v3 all of them were corrected :))
12:28:06 <johnthetubaguy> gmann_: well v2.1 == v2.0 yes
12:28:12 <edleafe> johnthetubaguy: we could accept 'os-liveMigrate' and 'os_live_migrate' to do the same thing. Then change the docs to only mention the latter
12:28:15 <gmann_> yea
12:28:40 <johnthetubaguy> edleafe: we could, yes
12:29:09 <gmann_> how about changing those in single microversion? but we need to support old also.
12:29:18 <johnthetubaguy> I am just worried we are making the API worse by following the new rules, anyways
12:29:31 <oomichi> I found three
12:29:46 <oomichi> disassociate_host disassociate_project and associate_host
12:29:51 <alex_xu> we said we will correct api as api-wg guide in next release in one microversion
12:29:53 <edleafe> johnthetubaguy: sometimes you need to move backwards before moving forwards :)
12:30:10 <oomichi> on the current APIs
12:30:12 <johnthetubaguy> edleafe: agreed, just asking the question
12:30:22 <johnthetubaguy> oomichi: I guess they are missing from the complete ref?
12:30:35 <oomichi> johnthetubaguy: nice point :)
12:30:39 <oomichi> johnthetubaguy: I feel so
12:30:43 <edleafe> johnthetubaguy: of course - it's an excellent point that has come up frequently in the api-wg
12:30:58 <johnthetubaguy> anyways, seems like we are going into this eyes open, so I am fine with it
12:31:25 <johnthetubaguy> need to go back and fix some of this next release, and add some alias bits into action and parameter names
12:32:01 <alex_xu> yea
12:32:13 <oomichi> johnthetubaguy: nice way
12:32:51 <alex_xu> oops, I have one want to ask at here again
12:32:57 <alex_xu> #link https://review.openstack.org/245543
12:33:21 <alex_xu> so there is opinion said, we should deprecated the feature before we remove the feature from the API
12:33:49 <alex_xu> my answer is no, because we have microversion, the old version api always give people chance to upgrade
12:34:19 <alex_xu> and asked oomichi, we agreed on the same thing. But want to ask this again in api team.
12:34:20 <oomichi> alex_xu: +1
12:34:34 <johnthetubaguy> alex_xu: we can remove it now in the microversion
12:34:46 <edleafe> alex_xu: agree
12:34:52 <johnthetubaguy> the problem, is we can't really ever remove it from the old version, but thats a different issue
12:35:16 <alex_xu> johnthetubaguy: yea
12:35:18 <oomichi> johnthetubaguy: yeah, completely same thing we discussed today
12:35:34 <gmann_> looks fine with microversion.
12:35:38 <oomichi> difficult to bump minimum microversion
12:36:38 <alex_xu> ok, thanks all, this looks like clear, let me bring this to live-migrate meeting after few hours
12:36:57 <edleafe> oomichi: why do we need to bump the minimum?
12:37:04 <gmann_> oomichi: yup, but that's should be fine as we provide better API in latest version :)
12:37:22 <oomichi> edleafe: for removing maintenance cost on the community
12:37:28 <alex_xu> edleafe: drop some burden?
12:37:36 <alex_xu> but that won't happen
12:37:42 <edleafe> oomichi: ah, I see. I don't think it'll ever happen
12:38:13 <oomichi> edleafe: for us ;)
12:38:14 <oomichi> edleafe: actually as you said
12:38:19 <alex_xu> ok, so any more want to bring up?
12:38:48 <alex_xu> ok, let's move on
12:38:50 <alex_xu> #topic API futures - patches for approved specs
12:39:08 <alex_xu> I guess this one is no also, as people focus on nova-spec recently
12:39:49 <johnthetubaguy> do we have any API concept guide stuff that needs discussing?
12:40:54 <alex_xu> I guess no
12:40:56 <oomichi> #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/complete-todo-in-api-concept-doc,n,z
12:41:05 <oomichi> ^^^ api-concept patches on the gerrit
12:42:05 <alex_xu> so let's move on
12:42:11 <alex_xu> #topic open
12:42:35 <alex_xu> any open? if not, we will have short meeting today
12:42:51 <zhipeng> do we have time for new specs ?
12:42:58 <zhipeng> just arrived
12:43:00 <zhipeng> sorry
12:43:05 <alex_xu> zhipeng: sure, please go ahead
12:43:56 <zhipeng> thx alex_xu, we have i think 3 specs in Nova that deals with API in Mitaka
12:44:07 <alex_xu> zhipeng: do you have link?
12:44:14 <zhipeng> I'm trying to find them :)
12:44:22 <zhipeng> in a minute
12:44:34 <alex_xu> zhipeng: no problem :)
12:46:09 <zhipeng> https://www.google.com/url?q=https%3A%2F%2Freview.openstack.org%2F%23%2Fc%2F241066%2F&sa=D&sntz=1&usg=AFQjCNExQRyhLq36c0GptMBn3qQo1oCcHQ
12:46:15 <zhipeng> ah sorry
12:46:43 <zhipeng> https://review.openstack.org/#/c/241066/
12:46:58 <zhipeng> https://review.openstack.org/#/c/241065/
12:47:07 <zhipeng> https://review.openstack.org/#/c/241063/
12:47:29 <jichen> I also have one ... https://review.openstack.org/#/c/219431/
12:48:25 <zhipeng> so basically we want to enable running storage functions in VM or container, that requires Nova supports attach with options that deals with high speed bus type
12:48:51 <johnthetubaguy> zhipeng: we have rejected blueprints to similar to those suggested in the past, but I will need to go through each one, one by one and understand what is wanted
12:49:17 <zhipeng> johnthetubaguy no problem
12:49:43 <johnthetubaguy> zhipeng: the problem I am facing right now, is we have around 100 in review right now
12:49:44 <alex_xu> yea, need take a look at more.
12:50:11 <johnthetubaguy> zhipeng: been accepting them since the end of july, and its proving hard work getting through them all
12:50:23 <zhipeng> yes totally understodd, and I just want to raise awareness :)
12:50:40 <zhipeng> johnthetubaguy understodd
12:50:54 <alex_xu> jichen one looks like ok for me
12:51:12 <alex_xu> but anyway let review them offline
12:51:41 <alex_xu> if no more question, I will close the meeting early
12:51:57 <johnthetubaguy> +1
12:52:01 <zhipeng> thanks alex_XU, we already faced BP dropped in L due to lack of review resource
12:52:03 <zhipeng> +1
12:52:13 <alex_xu> thanks all
12:52:15 <oomichi> alex_xu: nice leading
12:52:15 <jichen> ok, thanks
12:52:15 <alex_xu> zhipeng: np
12:52:18 <alex_xu> #endmeeting