02:01:07 <liusheng> #startmeeting Mogan
02:01:08 <openstack> Meeting started Thu Jun 29 02:01:07 2017 UTC and is due to finish in 60 minutes.  The chair is liusheng. Information about MeetBot at http://wiki.debian.org/MeetBot.
02:01:09 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
02:01:11 <openstack> The meeting name has been set to 'mogan'
02:01:33 <liusheng> hi mogan folks
02:01:35 <zhenguo> o/
02:02:09 <liusheng> shaohe_feng, Xinran zhangyang wanghao litao__
02:02:14 <liusheng> o/
02:02:17 <zhangyang> o/
02:02:22 <Xinran> o/
02:02:48 <liusheng> ok, because zhenguo is on a business in shenzhen
02:03:08 <wanghao> 0/
02:03:10 <wanghao> o/
02:03:12 <litao__> o/
02:03:13 <liusheng> I will hold this meeting temporarily
02:03:16 <liusheng> :D
02:03:36 <liusheng> thanks you all, let's move on
02:03:49 <liusheng> as always, the agenda:
02:03:55 <liusheng> #link https://wiki.openstack.org/wiki/Meetings/Mogan#Agenda_for_next_meeting
02:04:00 <liusheng> #topic announcements and reminder
02:05:09 <liusheng> seems nothing need to be announced here, zhenguo, wanghao do you have anything here ?
02:05:27 <liusheng> just reminder you guys about the summit presentation :D
02:06:06 <wanghao> emm
02:06:37 <wanghao> Haven't decide what the topic we should commit,  but we still work on this.
02:07:32 <liusheng> wanghao: ok, no hurry, the Summit call for presentations open until July 14
02:07:50 <wanghao> yes, okay
02:08:02 <liusheng> let's take a look the recently contribution
02:08:14 <liusheng> #link http://stackalytics.com/report/contribution/mogan/30
02:09:28 <liusheng> looks better than before, thanks for wanghao and litao's review contribution
02:09:42 <wanghao> liusheng: haha thanks.
02:10:36 <liusheng> wanghao: it is important if you guys want to be core reviewers, (just told you a secret :D)
02:10:53 <wanghao> got that :)
02:11:10 <litao__> liusheng: yes, come on
02:11:45 <liusheng> ok, if nothing else need to be announced here, let't move on
02:11:53 <liusheng> #topic task tracking.
02:12:03 <liusheng> #links https://etherpad.openstack.org/p/MoganWhiteBoard
02:12:20 <shaohe_feng> morning liusheng
02:12:34 * liusheng give folks a few minutes to review the tasks status
02:12:34 <shaohe_feng> sorry for late.
02:12:43 <liusheng> shaohe_feng: o/ it is ok
02:12:49 <liusheng> shaohe_feng: thanks
02:14:34 <liusheng> we have two bp specs need to be review these days, one is wanghao's bm adoption feature
02:15:42 * litao__ yes,  we should do it in this week.
02:16:06 <wanghao> still some thing we need to make a deal, and then we can go to code for that.
02:16:19 <liusheng> litao__: thanks, will review that today, shaohe_feng, zhenguo if you have time, please take a look for that
02:16:28 <wanghao> I have updated the spec for liusheng and zhengguo's review
02:17:19 <liusheng> wanghao: yes, thanks, it is this a real requirement of your products, right ?
02:18:04 <wanghao> liusheng: yes, it will be a good feature that we can support that to our coustemers.
02:18:19 <wanghao> liusheng: nova can't support this, you know
02:18:44 <shaohe_feng> wanghao: great, that's mogan beyond nova
02:19:01 <wanghao> shaohe_feng: yes, totally agree
02:19:10 <litao__> liusheng: so if mogan is steady enough, we will change nova to mogan
02:19:48 <liusheng> wanghao: ok, I just found you have given a -1 for my bms migration patch, hah. you can feel free to take over or abandon that patch, if there is conflicts with your idea
02:20:22 <liusheng> shaohe_feng, wanghao yes, sure
02:20:29 <wanghao> liusheng: baisclly it's ok for me,  just a little think.. haha
02:21:05 <liusheng> liusheng: hah, we all hope that
02:21:13 <liusheng> wanghao: :D
02:21:45 <liusheng> ok, another spec is about tracking resource with placement service
02:22:04 <liusheng> #link https://review.openstack.org/#/c/475700/
02:22:36 <wanghao> yeah, have reviewed it today,  I think it's good start.
02:22:43 <liusheng> actually, I have uploaded almost all the code of this feature :-P
02:22:50 <liusheng> wanghao: thanks
02:23:13 <wanghao> liusheng: okay, will check the code later.
02:23:38 <liusheng> wanghao: thanks
02:24:10 * litao__ liusheng: I will give you a -1 later :D
02:24:42 <liusheng> btw, must to reminder, after we switched to user placement to track node resources, we have removed the listing node api and list availability zone api
02:24:51 <liusheng> litao__: hah, welcome
02:25:19 <wanghao> liusheng: you mean to use placement api directly?
02:25:50 <wanghao> liusheng: get resource providers?
02:25:54 <liusheng> wanghao: as discussed with zhenguo, we decided won't provide listing nodes api to end users
02:26:24 <liusheng> wanghao: we will don't store the node info in Mogan's db after switching to use placement service
02:26:47 <wanghao> liusheng: okay, I see,  nodes should be exposed to user by mogan
02:27:26 <wanghao> liusheng: they just need to care about the servers.
02:27:30 <liusheng> for listing az api, currently, it depends on the nodes objects stored in Moga's db
02:27:52 <wanghao> emm , since we also remove the compute node objects
02:27:57 <liusheng> wanghao: yes, the only reason we expose the api I can find out is we need to support node aggregate
02:28:02 <liusheng> wanghao: yes
02:29:03 <liusheng> but for list az api, I am not sure if we need to provide a alternative implementation
02:29:28 <liusheng> we will leverage placement api to support nodes aggregate
02:30:08 <wanghao> liusheng: en,  so  do we need to support az?  I'm not clear with it yet.
02:30:44 <wanghao> liusheng: I mean if we have used nodes aggregate for placing server.  Do we need az more?
02:30:58 <liusheng> wanghao: not very sure, but we cannot support it in current way, since it depends on the node object stored in Mogan's db.
02:31:32 <liusheng> wanghao: but az is not similar with nodes aggregate
02:31:53 <litao__> wanghao: I think we need,  node aggregate is admin only and AZ is not
02:32:14 <liusheng> wanghao: have talked with zhenguo, may we can also remport the az info of nodes to placement service
02:33:11 <zhangyang> liusheng: so we need the driver to provide az info?
02:33:16 <wanghao> liusheng: you mean like adding a trail to node?
02:34:00 <liusheng> zhangyang: ironic nodes have the az attribute
02:34:18 <liusheng> zhangyang: that can be configured by operators
02:34:32 <wanghao> litao__: so we need to bind az and node aggregate.
02:34:37 <liusheng> wanghao: yes, placement support traits of nodes
02:34:41 <zhangyang> liusheng: emmm...but other driver may not have this value 😂
02:35:04 <liusheng> traits is like metadata of a node object in placement service
02:35:43 <wanghao> zhangyang: maybe can be configured in driver config group as default az.
02:35:56 <wanghao> liusheng: yes,  so we can use traits to set the az info
02:36:11 <liusheng> zhangyang: the az info can be None, and we have defaut az config  option
02:36:19 <liusheng> yes
02:36:21 <zhangyang> wanghao: haha, all in default az equals no az info.
02:36:39 <wanghao> zhangyang: that's true....
02:36:44 <liusheng> hah, yes
02:36:59 <wanghao> zhangyang: or push the driver to support it. :)
02:37:19 <zhangyang> wanghao: ok, i'll check it out.
02:37:32 <wanghao> zhangyang: cool
02:38:18 <litao__> liusheng: for az, we can learn from nova
02:38:44 <liusheng> zhangyang: btw, we can also push node info from cloudboot to placement in futrue
02:39:19 <liusheng> litao__: Nova have store the node info to itself db, but we don't
02:39:53 <zhangyang> liusheng: yeah, but the info cloudboot collected is much different from ironic, need to check the difference of abilities between them.
02:39:55 <liusheng> litao__: so someting maybe different, but we hope to provide similar functionality. hah
02:40:31 <litao__> liusheng: yeah
02:40:32 <liusheng> zhangyang: actually, it don't need much info to be report to placement
02:41:04 <liusheng> zhangyang: btw, what's the status of cloudboot driver task ?
02:42:58 <zhangyang> liusheng: its almost done on the driver side, but not tested with cloudboot yet...
02:44:07 <liusheng> zhangyang: cool, you can feel free to push your draft code to gerrit, maybe we can help to review it, even we cannot test it, hah
02:44:37 <zhangyang> liusheng: haha, ok, it's very nice of you guys.
02:44:50 <wanghao> zhangyang: that's we should do
02:45:06 <liusheng> hah, yes
02:45:22 <litao__> zhangyang: I like review codes, hah
02:45:49 <zhangyang> litao__: cool
02:46:17 * liusheng thumbs up for litao__
02:47:11 <liusheng> ok, anyone have other task state updates here ?
02:48:03 <liusheng> zhenguo is not here, we got less words. hah
02:48:20 <liusheng> #topic open disscussion
02:49:16 <litao__> liusheng: BTW, pls review my two patches , hah
02:49:25 <liusheng> oh, just reminder you to update the task status in etherpad, it is covenient for wanghao help to send the task tracking email
02:49:26 <wanghao> yeah, just don't forget update the task status
02:49:43 <liusheng> liusheng: ok, please add me in reviewers list
02:50:08 * litao__ liusheng:ok
02:50:15 <wanghao> and we need a etherpad to put the review list for review hour
02:50:48 <wanghao> do we have that?  if not, I create one: https://etherpad.openstack.org/p/MoganReviewHour
02:50:52 <liusheng> wanghao: oh, yes, forgot that
02:51:00 <liusheng> wanghao: thanks, cool
02:51:02 <wanghao> you can put the patch into it.
02:51:08 <wanghao> np
02:53:21 <liusheng> ok, we can end this meeting if nothing else need to be disscussed here
02:54:20 <shaohe_feng> wanghao: cool.
02:54:29 <shaohe_feng> we use this link.
02:54:35 <liusheng> pleas feel free to ping each other in mogan channel
02:54:41 <liusheng> shaohe_feng: yes
02:55:19 <liusheng> shaohe_feng: do you have anyting else to talk ? :D
02:55:50 <liusheng> ok, thank you all, bye
02:55:56 <liusheng> end meeting
02:55:56 <shaohe_feng> liusheng: no, just do more reviews. :)
02:56:10 <liusheng> #endmeeting