01:00:28 <zhangchi> #startmeeting tricircle
01:00:29 <openstack> Meeting started Wed May  8 01:00:28 2019 UTC and is due to finish in 60 minutes.  The chair is zhangchi. Information about MeetBot at http://wiki.debian.org/MeetBot.
01:00:30 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
01:00:32 <openstack> The meeting name has been set to 'tricircle'
01:01:09 <zhangchi> Morning everyone, welcome back to tricircle meeting
01:01:20 <ziyu> hi
01:01:46 <ziyu> godd day bruce zhang :)
01:01:47 <zhangchi> hi ziyu
01:02:20 <Yipei> hi, all
01:02:33 <XSM_HUST> hi,all
01:02:33 <ziyu> you would better tell other people in wechart we begin to meeting.
01:02:51 <ziyu> from today zhangchi will hold the meeting instead of me.haha!
01:03:49 <zhangchi> ok Message send
01:04:38 <zhangchi> #info To Do list https://etherpad.openstack.org/p/tricircle-train-plan
01:05:26 <zhangchi> ziyu how things going about your Finish the PR "ipv6_address_mode" not found in unit test
01:05:39 <ziyu> I have done it.
01:06:07 <zhangchi> good job is the patch submitted?
01:06:23 <ziyu> have merged.
01:06:35 <zhangchi> Greate
01:07:02 <zhangchi> weclome XSM_Hust
01:07:26 <zhangchi> what about your Deploy containers in tricircle environment
01:07:42 <XSM_HUST> After my experiments and bug tracking
01:08:37 <XSM_HUST> I found that the reason for deploying the container in the tricircle environment failed was that zun and kuryr did not specify region_name when using neutron api.
01:10:02 <XSM_HUST> This is a configuration bug in the zun and kuryr projects
01:11:50 <XSM_HUST> Howerver,the multi_region scene needs to specify which region's neutron api to use.
01:12:05 <ziyu> you can try seed a email to zun in openstack maillist?
01:12:19 <ziyu> or submit a bug report in the zun project?
01:12:38 <ziyu> how do you think? how about this? @zhangchi
01:13:33 <XSM_HUST> After I modified the neutron api calling code for the zun and kuryr projects, I was able to start the running container normally in the tricircle environment.
01:14:00 <zhangchi> good job  Is the problem in multi_region scene still exists
01:16:51 <Yipei> sent requests to different neutron?
01:18:57 <XSM_HUST> In other words, when creating a network or port in zun and kuryr, the region_name option is not specified, and the request will be sent to RegionOne by default.
01:19:43 <Yipei> is it can be configured to central region?
01:20:24 <Yipei> then send requests to central region be default
01:20:38 <Yipei> s/be/by
01:21:39 <XSM_HUST> Maybe possible. Also need specify region_name or endpoint option.
01:23:30 <XSM_HUST> The main problem is that there are no configuration items for region_name or endpoint in the neutron configuration of zun and kuryr.
01:23:41 <XSM_HUST> They
01:24:30 <zhangchi> agree with this  the region name may be set by keystone endpoint and your enviroment variable from Central pod
01:26:39 <zhangchi> or may be set by the neutron cli command
01:29:02 <XSM_HUST> openstack --os-region-name RegionTwo appcontainer run --name container1 --net network=net2 cirros sudo nc -l -p 80
01:29:29 <XSM_HUST> create a container command like this.
01:30:47 <zhangchi> region name has already been contained in CLI command
01:31:12 <XSM_HUST> But in this request,port request can't be sent to RegionTwo.
01:31:20 <zhangchi> then the request is still send to region one by default
01:32:01 <XSM_HUST> zun and kuryr use the RegionOne's neutron client.
01:34:22 <zhangchi> ok we will check the zun and kuryr mechanism for endpoint option
01:34:38 <XSM_HUST> yes, in zun and kuryr. port request didn't dispatch to diffrent region(neutronclient) according to real demand.
01:35:34 <Yipei> i thins all the networking requests should be sent to central region, and then central neutron decide where to create the resource
01:36:01 <Yipei> and it seems that availability_zone can be specified
01:36:11 <Yipei> #link https://developer.openstack.org/api-ref/application-container/?expanded=start-a-container-detail,commit-a-container-detail,attach-a-network-to-a-container-detail,execute-command-in-a-running-container-detail,enable-service-detail,create-new-container-detail#create-new-container
01:36:30 <XSM_HUST> you can see "/kuryr/kuryr/lib/utils.py" line38
01:36:58 <XSM_HUST> Also "/zun/zun/commen/clients.py"
01:36:59 <XSM_HUST> Line71
01:39:54 <Yipei> ok, maybe require more investigation, we can discuss it offline
01:40:02 <XSM_HUST> ok
01:40:19 <zhangchi> Yipei is experienced
01:42:36 <zhangchi> Ok we wils continue the next topics
01:43:52 <Yipei> @Hust_xwq, wanquan will take the l3 networking model over
01:45:16 <zhangchi> nice
01:47:54 <zhangchi> so what about the remaining tasks about the l3 networking model
01:50:45 <Yipei> he is not available today for some personal business, he is currently studying l3 networking.
01:51:10 <Yipei> zhuangzhuang and I will help him
01:51:44 <zhangchi> Ok I got it Thanks Yipei
01:52:36 <Yipei> you are welcome.
01:55:06 <zhangchi> From my side the exception about tricircle unitest cannot loading the network_ranges model from neutron have been fixed
01:57:37 <zhangchi> This is because some unit test make the fake client and is assigned to directory.get_plugin method
01:58:37 <zhangchi> is xiaohan zhang and yaguagnchen on line now?
02:01:12 <zhangchi> ok  their  irc enviroment have not already been configured
02:01:34 <zhangchi> any other more topics
02:02:18 <zhangchi> Thanks for attending nice day
02:02:33 <zhangchi> #endmeeting
02:02:36 <XSM_HUST> Bye
02:02:47 <zhangchi> Bye
02:02:56 <Yipei> bye
02:03:24 <zhangchi> Bye yipei and XSM_HUST
02:04:17 <zhangchi> #endmeeting