12:00:05 #startmeeting nova api 12:00:06 Meeting started Tue Nov 10 12:00:05 2015 UTC and is due to finish in 60 minutes. The chair is alex_xu. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:00:07 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 12:00:09 The meeting name has been set to 'nova_api' 12:00:15 who is here today? 12:00:17 o/ 12:00:26 hi 12:00:33 hi 12:00:59 hello everyone 12:01:41 o/ 12:02:16 ok, let's start the meeting 12:02:23 #topic actions from last meeting 12:02:27 hi 12:02:31 gmann_ create bp for microversion top bottom and changed testing 12:02:41 actually this is from before summit 12:02:50 alex_xu, yes, Done today 12:03:01 gmann_: cool, do you have link? 12:03:56 #link https://blueprints.launchpad.net/nova/+spec/nova-microversion-functional-tests 12:04:08 gmann_: cool, thanks 12:04:27 gmann_: looks great 12:04:43 so there will be spec later? 12:05:02 I don't think we need a spec for that, so I approved it 12:05:05 alex_xu, do we need spec for that? or we can do direct 12:05:15 johnthetubaguy, yea 12:05:26 ok, cool, got it 12:05:34 so let's move on 12:06:19 wait, so this will have a lot of work, right? do you need people help on that? 12:06:57 honestly, I want us to focus on the docs instead of this testing 12:07:10 I am tempted to make it low priority for now 12:07:42 johnthetubaguy, sounds good. even i ws thinking to start this after all extesnion tests work 12:07:56 so that it will be easy to share sample files in mocroversion testing also 12:08:11 ok, cool 12:08:17 #topic API Documentation 12:08:24 alex_xu, but should not reuire much work, we can do after doc thing done 12:08:37 gmann_: ok, got it 12:08:41 API concept doc and API reference 12:08:46 #link https://etherpad.openstack.org/p/nova-v2.1-api-doc 12:08:55 I wrote down the workflow and expectation for existed api refer, I wonder hope give feedback about whether is ok before I advert it out. 12:09:07 Basically it focus on improve the description of each api and parameter, after we have swagger, we can port them into the nova code directly. But that is huge number works. 12:10:33 that looks good to me 12:10:34 * alex_xu guess people busy on read my notes 12:10:45 johnthetubaguy: cool, thanks 12:11:07 alex_xu, nice plan to import to code from there. 12:11:09 I was wondering if it would be helpful to try to jump start this with something like a 2 day virtual doc sprint 12:11:12 sorry folks I have to run now, my food is ready, not used to the daylight savings changes yet 12:11:24 sdague: thats a very good idea 12:11:26 gmann_: thanks 12:11:29 sdague, +2 12:11:31 sdague: interesting idea 12:11:41 we can clean up a lot in that 12:11:42 +1 12:11:48 johnthetubaguy: no problem 12:11:48 +1 12:12:11 so, my suggestion for that would be to do it the first week of December some time 12:12:21 because there is a holiday in the US the end of nov 12:12:37 and that whole week will be patchy 12:12:47 sdague, Thanks giving? 12:12:51 gmann_: yep 12:13:01 that week is Mitaka-1, will people busy on spec freeze? 12:13:10 alex_xu: hmmm good point 12:13:18 maybe week after then? 12:13:30 yea after M-1 it will be good 12:13:31 yea, that will be good 12:13:57 what process for that? send email advert it 12:14:36 yeh, so I'd say pick 2 days. Try to get commitments from folks in advance that they'll contribute during it. 12:15:20 ok 12:15:29 Have an etherpad with things people should be working on, stuff ready for review, and we'll try to just go heads down and push through a bunch during that 12:15:45 it will be virtual sprint so it will be those 2 days in local timezone for folks 12:16:01 yup 12:16:11 sounds cool 12:16:54 ok, alex_xu you up for organizing it? 12:17:06 or any other volunteers to organize 12:17:13 anyone could take the lead on that 12:17:47 sdague: yea sure, I happy to help on, but I may need ask many question to you, as I haven't experience on organize those thing 12:18:06 Hi, is it a work to improve the doc and cleanup ? I'd like to be part of it. 12:18:17 sure, no probably, feel free to ask me about it any time 12:18:24 sdague: thanks 12:18:29 tangchen_: yea 12:18:35 tangchen_, yes. 12:18:45 alex_xu, I can also help if you need. 12:18:53 OK, please let me join. 12:18:59 gmann_: cool, thanks 12:19:13 #info alex_xu to organize nova doc virtual sprint for 2nd week of Dec 12:19:16 also thanks jichen and Kevin_Zheng, I saw them already work on the api refer 12:19:35 alex_xu: NP 12:19:50 API Review guideline 12:19:51 alex_xu: yep 12:19:54 alex_xu: so this will be bundle together and virtual sprint will handle them all ? 12:20:16 jichen: I'd say keep up the good work now 12:20:23 yea 12:20:30 needn't wait the virtual sprint 12:20:34 this is more about trying to pull in more people and push a bunch of work through quickly 12:20:34 jichen, we keep doing those and try to finish in sprint 12:20:36 sdague: yea, that's my thought too, ok got it 12:20:38 yea 12:21:18 ok, let's move on 12:21:33 the api review guide is bring up in the summit 12:21:48 anyone sign up to work on it? 12:22:37 emm...if not, I guess that will be my work, although I'm not good at English ;) 12:22:48 alex_xu: Sorry, I didn't go to the summit. What do you mean by "sign up" ? 12:23:04 tangchen_: I mean any volunteers to work on it 12:23:11 It need more experiance I guess 12:23:35 yea, so let me try when I have time 12:23:36 alex_xu: I'd like to. But currently, I don't what to do in detail. 12:23:47 tangchen_: yea, I know 12:23:51 so let's move on 12:23:55 Swagger Doc Generate 12:24:00 Thanks oomichi discussed swagger with me at summit, and we summary something in the etherpad 12:24:07 Thanks oomichi discussed swagger with me at summit, and we summary something in the etherpad 12:24:11 #link https://etherpad.openstack.org/p/mitaka-nova-swagger 12:24:20 oops 12:24:26 #link https://review.openstack.org/243365 12:24:31 And work out the initial version nova-spec, still need some detail to put in. 12:25:04 appreciate gmann_ and oomichi_ if you guys can help me input some tempest consideration in the spec 12:25:13 alex_xu, i did not get chance to look into that. I will look tomorrow and add if needed from Tempest side etc 12:25:24 alex_xu, yea, sure 12:25:25 gmann_: thanks a lot 12:25:35 and Next is I should work out more PoC to show how to doc in the code. 12:26:04 #action alex_xu update the PoC to show more example 12:26:17 alex_xu, Also will check with swagger team about request sample example in spec 12:26:29 gmann_: ok, cool 12:27:06 I think swagger can't support that in short time, we can try add some extend field into the swagger to support request sample example 12:27:45 and those can by picked by UI ? 12:27:59 yea, UI is coded by doc team 12:28:18 #link https://github.com/russell/fairy-slipper 12:28:21 * edleafe yawns 12:28:24 gmann_: this one I guess ^ 12:28:26 yea, 12:28:31 alex_xu, cool 12:29:06 actually I hope the Russell give some input how to support microversions and actions, but I didn't get response from him 12:29:33 and looks like Anne isn't here today also :( 12:30:11 anyway, any more question on swagger? 12:30:55 if not, let's move on, this is all about api doc 12:30:59 #topic Service Catalog 12:31:01 alex_xu, it will be good if we can have some fairy-slipper thing also with your POC or patches for those 12:31:23 so that we can see how those going to be looks like and improve at early stage if needed 12:31:35 but as you said need input from doc team on those 12:31:43 gmann_: yea, I should catch Anne get some input 12:32:00 alex_xu, ok cool 12:32:11 for this part is about remove project_id from URL 12:32:17 #link https://etherpad.openstack.org/p/removing-project-id-in-nova-urls 12:32:30 sdague: ^ anything you want to talk about? 12:33:21 I saw part of patches isn't ready for review yet 12:34:42 I guess sdague busy on other things 12:34:45 yea 12:35:12 any more question on this, otherwise let's move on 12:35:12 yeh, sorry, baby just got up 12:35:16 back now 12:35:28 so, I think it's largely doable in one patch 12:35:39 sdague: yea, baby is more important :) 12:35:45 mmmpork is assisting getting that patch into shape with tests 12:36:00 sdague, devtsack patch also will be landed right? - https://review.openstack.org/#/c/233079/3 12:36:05 https://review.openstack.org/#/c/233076/ 12:36:07 sdague, means not actualy WIP 12:36:15 gmann_: yes, it requires the nova patch though 12:36:19 yea 12:36:43 a novaclient patch already merged 12:36:53 right 12:37:22 we need the novaclient patch in a released version before we can change devstack as well 12:37:35 otherwise things like nova flavor-list blow up 12:37:37 ok, so I guess we just waiting for the nova patch ready to review 12:37:45 because of the microversion detection going a bit haywire 12:38:06 #action sdague to circle with mmmpork today on nova patch 12:38:18 #action sdague to circle with mmmpork today on nova patch for optional project_id 12:38:19 sdague: cool, thanks 12:38:53 ok, cool, so if no more question, let's move on 12:39:14 #topic Novaclient 12:39:37 there is one work bring up by sdague, it is about Caching the version negotiation. 12:40:02 at least for now, there isn't any assignee yet 12:40:21 yes, that was a "nice to have" from the API session to reduce our round trips 12:40:55 I think that if no one else takes a look I'll see if mmmpork is up for it 12:41:13 alex_xu, sdague what is that actually? 12:41:15 sdague: ok, cool 12:41:50 do we have some etherpad/bug etc so that i can look tomorrow 12:42:03 gmann_: so, we're doing up to 3 round trips for every nova cli call now 12:42:10 because of the microversion negotiation 12:42:28 but the microversions supported on the server don't change very often 12:42:36 #link https://etherpad.openstack.org/p/mitaka-nova-api 12:42:41 gmann_: there are some notes ^ 12:43:00 so we should ideally cache that result for like 24hrs 12:43:29 sdague, ah i see. 12:43:45 alex_xu, Thanks for link 12:43:54 gmann_: np 12:44:03 anyway, we should move on. It's a pretty low priority thing, just a nice to have 12:44:11 ok, let's move on 12:44:19 #topic Testing 12:44:29 #link https://review.openstack.org/#/c/169126/ 12:44:47 I saw this spec about test Microversion in tempest side 12:45:15 Emm...ken'ichi isn't here 12:45:33 I think people just need to review it 12:45:57 alex_xu, yea 12:46:12 sdague, has +2 on it. 12:46:12 another thing about test is merge all extension for sample test 12:46:22 https://blueprints.launchpad.net/nova/+spec/api-sample-tests-with-all-extensions 12:46:24 in parallel we have started implementation also 12:46:34 gmann_: cool 12:46:49 alex_xu, and for all extesnion- 12:46:56 gmann_: anything update for all extension? 12:47:03 started for image and flavor merging 12:47:06 * alex_xu jump too fast... 12:47:06 #link https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/api-sample-tests-with-all-extensions,n,z 12:47:16 reviews are welcome 12:47:32 gmann_: cool! 12:47:36 manly i need feedback about patch size. if it is ok to do by resource vise 12:47:58 because flavor one looks large but easy to review 12:48:04 I'll take a look later 12:48:13 sdague, Thanks. 12:48:18 yea, me too 12:48:29 ok, let's move on 12:48:29 alex_xu, Thanks 12:48:35 time is tight again 12:48:36 something that mmmpork pointed out last night was that we still have a very different validation model in api samples from the jsonschema 12:48:41 get_regexes 12:48:49 might be a thing to clean that up 12:49:11 sdague, ok, need to look on verfication part 12:49:46 yea, good point 12:50:00 sdague, once we do validation with response schema (in swagger work) then we can clean up 12:50:31 then we might get rid of validation based on regex 12:51:25 gmann_: ok 12:51:39 I need get into the code to know how to clean up it 12:51:48 10 mins left, let's move on 12:51:54 #topic API Deprecations 12:51:57 so, the swagger work seems really abstract to me still. Is there a piece of code that takes something from the tree and publishes it to docs.openstack.org anywhere to see it all in action? 12:52:56 sdague: not yet, the swagger ui is still in progress https://github.com/russell/fairy-slipper 12:53:02 sdague, do not have for docs.openstack.org thing. 12:53:04 yea 12:53:41 if russell can hook things with swagger POC then it wil be nice 12:53:48 yea 12:54:08 need poke doc team to give some feedback 12:54:40 sdague: any thought, or you see some risk? 12:54:41 ok, so we should set a date of when that POC will be there, it should be before the doc sprint 12:55:02 because it will help encourage people if their work hits docs.openstack.org immediately 12:55:02 sdague: ok, I will busy on it 12:55:28 so I would say the status of that publishing pipeline should be a standing item on this agenda 12:56:33 sdague: yea, this really talk with doc team 12:56:44 s/really/really need/ 12:56:56 alex_xu: ok, you have the action for this? 12:57:27 #action alex_xu contact the doc team about how to swagger stuff 12:57:38 #undo 12:57:39 Removing item from minutes: 12:57:54 #action alex_xu contact the doc team about swagger stuff 12:58:10 Sorry, busy getting the family off to work/school. I'll read through the scrollback and see if there is anything I can help with. 12:58:13 ok two mins left... 12:58:22 edleafe: no prolbme 12:58:41 sdague: anything we need to for deprecated the ec2 and legacy api for now? 12:58:55 I mean any actions from summit? 12:59:45 alex_xu: not yet 12:59:55 we don't have feedback from operators yet on using them 12:59:56 sdague: ok, cool 13:00:12 I guess I have an action on asking about the stackforge project 13:00:18 and how it's working for folks 13:00:24 so I guess we talk all of mitaka tasks today 13:00:34 sdague: ok 13:00:40 sorry, it's time to close meeting 13:00:45 thanks all 13:00:53 #endmeeting