02:01:49 <zhenguo> #startmeeting mogan
02:01:50 <openstack> Meeting started Thu Aug 24 02:01:49 2017 UTC and is due to finish in 60 minutes.  The chair is zhenguo. Information about MeetBot at http://wiki.debian.org/MeetBot.
02:01:51 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
02:01:53 <openstack> The meeting name has been set to 'mogan'
02:02:34 <zhenguo> liusheng, shaohe_feng, Xinran, luyao, zhangyang: o/
02:02:44 <liusheng> o/
02:03:16 <zhenguo> wanghao and litao can't attend today's meeting!
02:04:38 <zhenguo> liusheng: only we two here?
02:04:52 <liusheng> zhenguo: :(
02:04:54 <liusheng> zhenguo: hah
02:05:15 <zhenguo> shaohe_feng, Xinran: are you around?
02:06:25 <liusheng> seems they all are absent
02:06:28 <shaohe_feng> zhenguo: I'm here
02:06:42 <zhenguo> shaohe_feng: o/
02:06:45 <liusheng> shaohe_feng: o/
02:07:21 <zhenguo> ok, it seems like a core team meeting, lol
02:07:34 <zhenguo> as always, the agenda:
02:07:34 <zhenguo> #link https://wiki.openstack.org/wiki/Meetings/Mogan#Agenda_for_next_meeting
02:07:45 <zhenguo> #topic announcements and reminders
02:07:56 <zhenguo> liusheng helped to move our docs to RTD http://mogan.readthedocs.io/projects/api-ref/en/latest/
02:08:07 <zhenguo> but there are still some problems during building, we will try to address it before Pike
02:08:24 <zhenguo> shaohe_feng: are you familiar with readthedocs
02:08:35 <liusheng> yes, I have no idea about that problem :(
02:08:50 <shaohe_feng> zhenguo: not familiar. let me look into it.
02:08:59 <shaohe_feng> zhenguo: I will give some introduce for all
02:09:10 <zhenguo> shahe_feng: cool
02:09:56 <zhenguo> There's a patch up to replace 'baremetal' with 'baremetalcompute' in our CLI
02:09:56 <zhenguo> any objections?
02:11:01 <zhenguo> https://review.openstack.org/#/c/496138/
02:11:16 <shaohe_feng> seem no.
02:11:44 <liusheng> seems we  can merge it quickly to avoid conflicts :D
02:12:09 <zhenguo> shaohe_feng, liusheng: so please help to +A it :D
02:12:46 <shaohe_feng> zhenguo: sure.
02:12:53 <liusheng> zhenguo: +2ed
02:13:39 <zhenguo> liusheng, shaohe_feng: do you think when we should cut the Pike release
02:13:46 <zhenguo> by the end of next week?
02:14:19 <liusheng> sounds good
02:14:49 <zhenguo> ok, so let's try to land more things this week
02:15:15 <shaohe_feng> zhenguo: yes.
02:15:15 <liusheng> we can finish our high priority works until next week
02:16:12 <zhenguo> yes, we can check the tasks during next topic
02:16:39 <zhenguo> Contribution into mogan for the last 30 days
02:16:39 <zhenguo> #link http://stackalytics.com/report/contribution/mogan/30
02:16:43 <Xinran> o/ Sorry for late ..
02:16:49 <zhenguo> Xinran: np
02:17:21 <zhenguo> please take a look at the last 30 days contributions
02:17:50 <zhenguo> really happy to see Xinran's reviews, thanks!
02:18:36 <Xinran> zhenguo,  hah thanks
02:18:46 <zhenguo> Xinran: hah please keep it :D
02:19:10 <zhenguo> not much of announcement, anyone else have a thing here?
02:19:28 <zhenguo> ok, let's move on
02:19:32 <zhenguo> #topic task tracking
02:19:42 <zhenguo> #links https://etherpad.openstack.org/p/MoganWhiteBoard
02:19:51 * zhenguo gives folks a few minutes to review
02:24:09 <zhenguo> I only got one patch left for the node aggregates bp https://review.openstack.org/#/c/493495/
02:24:31 <zhenguo> please help to look at it shaohe_feng, already got a +2
02:24:45 <shaohe_feng> zhenguo: Ok. then you can try to play with mogan
02:25:03 <zhenguo> shaohe_feng: hah
02:25:11 <shaohe_feng> zhenguo: find bug, and file them.  and let's fix them.
02:25:23 <shaohe_feng> then we can make mogan mor stable.
02:25:26 <shaohe_feng> :)
02:25:39 <zhenguo> shaohe_feng: yes, everytime I setup the physical env, will find lots of bugs :D
02:26:27 <shaohe_feng> zhenguo: great. do you have physical env?
02:26:35 <zhenguo> shahe_feng: sure, lol
02:26:50 <zhenguo> shaohe_feng: I need to test my socat console with that
02:27:36 <shaohe_feng> zhenguo: nice. so usrs can have multi ways to access the host.
02:27:43 <zhenguo> so my socat console task seems can be done by this week
02:27:49 <shaohe_feng> nice.
02:28:29 <zhenguo> hi guys, please remember to update the etherpad,
02:28:40 <zhenguo> wanghao will send a tasks report out tommorow
02:30:15 <shaohe_feng> good.
02:30:39 <zhenguo> shaohe_feng, Xinran: do you have servers for testing?
02:31:20 <shaohe_feng> zhenguo: we are try to find some servers from our architecture.
02:31:29 <zhenguo> shaohe_feng: hah, cool
02:31:34 <shaohe_feng> architect
02:31:52 <shaohe_feng> zhenguo: you have meet he in China Day.
02:32:09 <zhenguo> shaohe_feng: yes, he mentioned that
02:32:19 <liusheng> that is cool :)
02:32:41 <zhenguo> liusheng: what about your tasks?
02:32:53 <zhenguo> liusheng: seems patches are ready for review
02:33:12 <liusheng> zhenguo: yes, but didn't finish yet
02:33:37 <zhenguo> liusheng: ok, seems almost done :D
02:33:37 <liusheng> zhenguo: will finish it until this week
02:33:42 <liusheng> zhenguo: yes
02:34:11 <zhenguo> liusheng: seems we will work 6 days this week
02:34:38 <liusheng> zhenguo, shaohe_feng and please take a look this patch: https://review.openstack.org/#/c/492370/ it is convenient for function testing :)
02:34:44 <liusheng> zhenguo: will hanle the conflicts
02:34:54 <litao__> I am later, just an internal meeting
02:35:05 <zhenguo> liusheng:  yes
02:35:11 <zhenguo> litao__: o/
02:35:37 <zhenguo> litao__: we will cut the Pike release by the end of next week
02:36:08 <litao__> zhenguo: So we have little time
02:36:22 <zhenguo> litao__: yes, time is tight
02:36:39 <litao__> zhenguo: Ok, i will hurry up
02:36:46 <zhenguo> litao__: thanks
02:36:59 <zhenguo> seems the first managing running server patch is amost good,
02:37:10 <zhenguo> please help to review
02:37:22 <litao__> zhenguo:  sure
02:37:34 <zhenguo> litao__: you can wait for that landed, then rebase your patch :D
02:38:01 <shaohe_feng> zhenguo: OK.
02:38:09 <litao__> zhenguo: yes, there will be many conflicts
02:38:19 <zhenguo> litao__: hah
02:39:25 <zhenguo> I would like to return the addresses with server API object instead of the whole networks info
02:39:31 <zhenguo> https://review.openstack.org/#/c/495826/
02:39:54 <zhenguo> we also provide a separated URI for server networks
02:40:25 <liusheng> zhenguo: it is only change in the api layer, right ?
02:40:36 <zhenguo> liusheng: yes
02:40:42 <liusheng> zhenguo: ok
02:40:57 <zhenguo> liusheng: seems client side also need to be updated
02:41:04 <liusheng> zhenguo: yes, sure
02:42:00 <zhenguo> ok, anything else here?
02:43:17 <zhenguo> Xiran: do you find using patch method for flavor access list is a bit inconvinient?
02:43:17 <shaohe_feng> seems no
02:43:25 <zhenguo> Xiran^^
02:43:30 <zhenguo> Xinran
02:43:32 <zhenguo> hah
02:43:59 <shaohe_feng> I will talk with her.
02:44:02 <Xinran> zhenguo,   yes
02:44:24 <shaohe_feng> Xinran: is online. :)
02:44:24 <Xinran> zhenguo,  I found we use index to delete
02:44:36 <zhenguo> Xinran: yes
02:44:37 <liusheng> seems the patch almost looks good
02:44:40 <zhenguo> Xiran: seems hard to use
02:45:06 <zhenguo> liusheng: you mean the addresses one?
02:45:27 <Xinran> zhenguo,  must get the index of the project to be delete
02:45:41 <liusheng> zhenguo: no, I mean the flavor access one, hah
02:45:45 <zhenguo> Xinran: yes
02:45:55 <zhenguo> liusheng: we are talking about the patch list method
02:46:07 <zhenguo> liusheng: we need to get the index of the list to delete it
02:46:50 <zhenguo> Xinran: not sure if the sort of the list is fixed
02:47:49 <Xinran> zhenguo,  seems it depends to order of adding , I will check it
02:48:01 <Xinran> s/to/on
02:48:22 <zhenguo> Xinran: anyway, seems we changed to make it's inconvenient than before
02:48:52 <zhenguo> Xinran, liusheng: maybe it's ok to keep the original flavor access URI?
02:49:27 <liusheng> zhenguo: you mean still use a saparate API
02:49:33 <zhenguo> liusheng: yes
02:49:44 <Xinran> zhenguo,  hah,  but I think use a controller is redundant.
02:50:05 <zhenguo> Xinran: you mean the flavor access controller?
02:50:12 <Xinran> zhenguo,  yes
02:50:45 <Xinran> is there another way to delete with PATCH method
02:50:51 <zhenguo> Xinran: we can use someting like /flavors/flavor_uuid/access ?
02:51:19 <zhenguo> Xinran: oh, it's the current way
02:51:24 <liusheng> zhenguo: sorry, may I need to review again, I even didn't get the inconvenience you talking about :(
02:51:48 <zhenguo> liusheng: PATCH method is easy to update key/value
02:52:01 <zhenguo> liusheng: the flavor access is a list
02:52:07 <zhenguo> liusheng; there's no key
02:53:04 <zhenguo> liusheng: when you adding a project in, you PATCH {"op": "add",  "path": "/projects/-", "value": "xx"}
02:53:06 <liusheng> zhenguo: got it, but it is easy to work arount isn't it ?
02:53:14 <zhenguo> liusheng: how
02:53:15 <liusheng> zhenguo: arount/ around
02:53:30 <zhenguo> liusheng: when you delete a project, you need to find the index of it in the list
02:53:58 <liusheng> zhenguo: don't use the json patch method
02:54:00 <zhenguo> liusheng: PATCH {'op': 'remove', 'path': '/projects/0'}
02:54:27 <zhenguo> liusheng: you mean use PUT?
02:54:29 <shaohe_feng> zhenguo: I remeber you have send the rfc link of PATCH before.
02:54:42 <zhenguo> shaohe_feng: yes
02:55:14 <liusheng> zhenguo: no I mean don't use the apply_jsonpatch method, we just query the projects and set it
02:55:19 <zhenguo> seems there's not a better way to update a list, either PUT or PATCH
02:55:50 <zhenguo> liusheng: can you give a example
02:56:01 <zhenguo> liusheng: we must follow waht the REST defination
02:57:06 <liusheng> zhenguo: it don't effect the REST definition, we just need query the projects, and add the project then save it in db, isn't right ?
02:57:34 <zhenguo> liusheng: I mean, how the API like
02:58:07 <zhenguo> liusheng: we can do anything under the API
02:58:27 <zhenguo> liusheng: but what do you want users use the API for update the access list
02:59:09 <liusheng> zhenguo: let me think more
02:59:20 <liusheng> zhenguo: we can disscuss after this meeting
02:59:31 <Xinran> zhenguo, liusheng just check,  the sort of list depends on order of adding
02:59:33 <zhenguo> liusheng: ok
02:59:56 <zhenguo> Xinran: so, users need to remember the order themselves?
03:00:15 <liusheng> why need to use the index ?
03:00:21 <Xinran> zhenguo,  seems yes ...
03:00:27 <zhenguo> Xinran: :(
03:00:59 <zhenguo> liusheng: that's what PATCH works for a list
03:01:15 <zhenguo> liusheng: it's not a dict, we can't find a key
03:02:01 <zhenguo> Xinran: if so, I have to say sorry, I'd like to keep the origianl way :(
03:02:27 <liusheng> zhenguo: got it, will check  more
03:02:39 <zhenguo> liusheng: ok
03:02:48 <Xinran> zhenguo,  hah it's ok. if we dont have other solutions :)
03:02:56 <zhenguo> Xinran: hah
03:03:08 <zhenguo> oh, it's almost time,
03:03:14 <zhenguo> anything else?
03:03:29 <zhenguo> thanks all
03:03:33 <zhenguo> #endmeeting