14:02:36 <joehuang> #startmeeting tricircle 14:02:36 <openstack> Meeting started Wed Mar 15 14:02:36 2017 UTC and is due to finish in 60 minutes. The chair is joehuang. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:02:37 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:02:39 <openstack> The meeting name has been set to 'tricircle' 14:02:44 <joehuang> hello 14:02:54 <joehuang> #topic rollcall 14:02:58 <joehuang> #info joehuang 14:03:05 <dongfeng> #info dongfeng 14:03:22 <Yipei> #info Yipei 14:03:31 <fanyishi> #info fanyishi 14:03:32 <yinxiulin> #info 秀林 14:03:42 <yinxiulin> #info xiulin 14:03:42 <RongHui11> #info ronghui 14:03:46 <zhiyuan> #info zhiyuan 14:04:02 <longxiongqiu> #info longxiongqiu 14:04:07 <WuCheng> #info WuCheng 14:04:20 <joehuang> #topic Test coverage requirement 14:04:45 <joehuang> there is one patch submitted to set threshold for test coverage 14:04:56 <joehuang> #link https://review.openstack.org/#/c/444394/ 14:05:12 <joehuang> the current coverage is 71% 14:06:06 <joehuang> so all new patch should add enough test code 14:06:28 <joehuang> to keep the test coverage >-71% 14:07:04 <joehuang> if a patch leads to the coverage <71% 14:07:22 <joehuang> then the check test will fail 14:07:28 <longxiongqiu> ok 14:08:06 <RongHui11> got it 14:08:08 <joehuang> that means each new patch must add test cases to cover more than 71% code, otherwise it'll fail in test 14:08:28 <dongfeng> will the doc be ignored? 14:08:45 <joehuang> doc test is another test 14:09:09 <dongfeng> got it. 14:09:47 <joehuang> for new patch, we hope it can reach >90% for its own code 14:10:17 <joehuang> that means all branch clauses will be covered as much as possible 14:11:17 <joehuang> you can check the coverage of test in each patch, just click the link of tricircle-coverage-ubuntu-xenial-nv 14:12:12 <joehuang> then you can find each file coverage test report in the folder "cover" 14:12:48 <joehuang> do you have any question to check the coverage report? 14:13:04 <dongfeng> no for me 14:13:22 <joehuang> click each file, you can see what tested, what not 14:13:23 <Yipei> no 14:14:13 <RongHui11> no 14:14:33 <joehuang> #info for old file, it's encouraged to add test cases to increase the test coverage 14:15:03 <joehuang> good, if no question, zhiyuan, I think we can approve this patch 14:15:15 <zhiyuan> ok, that's fine 14:15:42 <joehuang> #topic Pike-1 prioritized patches progress review 14:16:20 <joehuang> #info two patches of VxLAN feature merged 14:16:37 <joehuang> #info router az support merged 14:17:11 <joehuang> other patches are still in review 14:17:37 <joehuang> QoS spec is still in discussion 14:18:04 <joehuang> VxLAN feature spec is close to merge 14:18:55 <RongHui11> yes we need to speed up the rhythm of the discussion 14:19:03 <joehuang> do you have topic to discuss for the spec/code patch in review 14:19:15 <dongfeng> no for me. 14:19:57 <zhiyuan> no for me, I will update my patches tomorrow based on the comments 14:20:39 <joehuang> tricircleclient patch is close to be ready, I think 14:20:59 <electrocucaracha> sorry this savings time day 14:21:05 <electrocucaracha> yes I guess it's 14:21:10 <joehuang> hello, victor 14:21:18 <joehuang> o, your time changed 14:21:28 <joehuang> what's the current time now 14:21:39 <electrocucaracha> Sindhu told me that she will fix it, but now she is serving as bug deputy in neutron 14:21:49 <electrocucaracha> 9:21 am 14:22:06 <joehuang> is it too early for you? or it's good 14:22:18 <joehuang> understand 14:22:19 <electrocucaracha> its much better 14:22:21 <electrocucaracha> hehehe 14:22:35 <sindhu> yes I will fix it today 14:22:41 <sindhu> sorry for the delay 14:22:54 <joehuang> there are several patches are almost close to merge, please have a look 14:23:23 <joehuang> to Sindhu, it's ok, not delayed yet :) 14:24:19 <joehuang> We just discussed that the coverage patch will be merged to look after tricircle code test coverage 14:24:36 <joehuang> thanks to victor to bring this one 14:24:46 <joehuang> really good 14:25:03 <electrocucaracha> yeah, and now makes more sense to have coverage jobs 14:25:19 <electrocucaracha> that validates the code coverage and not the creation of the reports 14:26:20 <joehuang> +1, it's very helpful, and I saw that the coverage increased from 71% to 72% 14:26:39 <electrocucaracha> \o/ 14:27:17 <joehuang> after Tricircle pike-1 is released, we'd like to have a demo in OPNFV beijing summit 14:27:23 <joehuang> #topic Demo of VNF high availability across OpenStack with Tricircle in OPNFV Beijing summit 14:27:55 <joehuang> based on the VxLAN cross Neutron L2 network and allowed address pair 14:28:32 <joehuang> we can help VNF(telecom application) to realize high availability across OpenStack 14:28:48 <joehuang> and I discussed with contributors in OPNFV 14:29:32 <joehuang> we'd like to setup a demo of this in OPNFV beijing summit, it'll be held in the middle of June 14:30:00 <joehuang> and plan to integrate tricircle into OPNFV in its E-release 14:31:59 <joehuang> do you have some questions on this? there is one project in OPNFV called multisite, I am also the PTL of that project 14:33:30 <electrocucaracha> joehuang: did you know if those sessions are going to be available online? 14:33:51 <joehuang> you mean OPNFV summit? 14:33:58 <electrocucaracha> yes 14:34:27 <joehuang> last year, video was recorded and published in youtube 14:34:55 <joehuang> OPNFV is call for presentation for the summit, we may submit a session 14:35:31 <joehuang> except the sessions, OPNFV summit also provide expo like that in OpenStack summit 14:35:48 <joehuang> for expo, no video will be recorded 14:36:10 <joehuang> If the session was selected, then we can have video published in youtube 14:36:33 <joehuang> would you like to submit a session together? 14:37:52 <electrocucaracha> well, I can try it but I'm not sure that I've fully support of my manager 14:38:14 <electrocucaracha> anyway, it's something that we can discuss offline 14:38:20 <joehuang> ok, let's discuss the session offline 14:38:51 <joehuang> and one session was selected in OpenStack summit together with telecom NFV WG 14:39:32 <joehuang> so I am planning to attend OpenStack Boston summit 14:39:39 <joehuang> we may meet there 14:41:21 <joehuang> #topic open discussion 14:41:35 <joehuang> any topics? 14:41:39 <Yipei> yes 14:41:52 <joehuang> please, Yipei 14:41:55 <Yipei> for LBaaS, when creating lb, the service plugin will use the local plugin to create port. but the port body has no ip address. instead, the service plugin fills fixed_ip with subnet_id, leading None in ip_address 14:42:12 <Yipei> https://github.com/openstack/tricircle/blob/master/tricircle/network/local_plugin.py#L443 14:43:55 <joehuang> this is a bug in the code 14:44:12 <Yipei> so how it is not belong to special port 14:44:31 <zhiyuan> no, that's to reject port creation request from dhcp agent 14:44:51 <Yipei> just skip the part? 14:45:07 <joehuang> have you checked the request body passed from LBaaS service plugin? 14:45:24 <joehuang> for DHCP port processing, it's correct 14:45:52 <zhiyuan> so LBaaS will also pass the body only with subnet_id? 14:45:56 <joehuang> but we need to make sure it works for LBaaS or not 14:46:27 <Yipei> to zhiyuan, yes, no fixed_ip 14:47:03 <zhiyuan> no fixed_ip or no ip_address? 14:47:16 <Yipei> o, no ip_address 14:47:27 <Yipei> fixed_ip is filled with subnet_id 14:47:53 <zhiyuan> does the body contain device_owner? 14:48:00 <Yipei> yes 14:48:02 <paul-andreraymon> quit 14:48:09 <Yipei> DEVICE_OWNER_LOADBALANCERV2 14:49:11 <joehuang> we may judge by device owner to process the request from LBaaS 14:49:49 <zhiyuan> cool, then we can also add this port to the special port list 14:50:40 <Yipei> so we need to add port in special ports 14:51:35 <Yipei> https://github.com/openstack/tricircle/blob/master/tricircle/network/local_plugin.py#L453 14:51:53 <Yipei> then it will be processed here 14:52:14 <Yipei> oh, it wrong 14:52:15 <Yipei> sorry 14:53:31 <joehuang> to Yipei, I think you can figure out offline how to fix this issue, it's much more clear nor 14:53:47 <Yipei> ok 14:53:50 <Yipei> got it 14:54:03 <electrocucaracha> I'd like to mention the upgrades requirements patch that was submitted earlier, https://review.openstack.org/#/c/445153/ it contains a change in SQLAlchemy 14:54:13 <joehuang> few minutes left, other topics? 14:54:24 <RongHui11> no for me 14:54:31 <dongfeng> no for me 14:54:46 <longxiongqiu> no for me 14:55:02 <Yipei> no 14:55:03 <joehuang> ok, will approve soon 14:55:36 <joehuang> please review the patches 14:55:43 <zhiyuan> ok 14:55:48 <yinxiulin> no 14:56:56 <joehuang> ok, thank you for attending the meeting 14:57:06 <joehuang> #endmeeting