13:00:54 #startmeeting tricircle 13:00:56 Meeting started Wed Nov 23 13:00:54 2016 UTC and is due to finish in 60 minutes. The chair is joehuang. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:57 hello 13:00:58 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:01:00 The meeting name has been set to 'tricircle' 13:01:07 #hello 13:01:08 #topic rollcall 13:01:09 hello 13:01:13 #info joehuang 13:01:19 #info ronghui 13:01:23 #info zhiyuan 13:01:24 #info longxiongqiu 13:01:24 #info xiulin 13:01:28 #info Yipei 13:02:09 #info dongfeng 13:02:16 #info jiawei 13:02:18 #topic Ocata feature development review 13:03:02 let's have a short review for the feature development, the coming milestone is in the middle of Dec 13:03:34 please have a brief description of the feature you are working for 13:04:49 the patch for documentation for resource routing is in review, and the blueprint for resource routing is also in review. 13:05:33 (1) the manual installation guide is under review. (2) the spec for combined bridge network is under review (3) the code for combined bridge network is under development, I will start to test the code with DevStack tomorrow 13:05:37 to Dongfeng, yes, just receive the notification of the blueprint 13:06:36 for the configuraton i still try to use the document tools to get the configuraton guide 13:06:42 is there anyone install tricircle according to the manual 13:07:25 i install the tircircle according th the manual for singlenode 13:08:02 sorry 13:08:15 not the manual 13:08:21 is the devstack 13:09:38 you can install the devstack without tricircle first, then try to configure the tricircle according to the manual guide 13:09:49 hello Junsik, not see you some time 13:10:24 ok 13:10:29 i will try it 13:10:38 to Zhiyuan, thank you for your great effort 13:10:45 implement LBaaS for tricircle, still studying LBaaS. plan to run LBaaS in one OpenStack instance. 13:10:47 thank you, ronghui 13:11:03 to yipei, great 13:11:20 LBaaS is applying openstack bigtent 13:11:35 will try installing devstack tomorrow by following the manual 13:11:37 the project name is Ocataiva 13:12:36 ok, i will pay more attention to the project 13:13:26 you can refer to this guide: http://docs.openstack.org/mitaka/networking-guide/config-lbaas.html 13:13:54 this is mitaka version, you'd better to read the newton release 13:14:08 ok, got it. thx for sharing 13:14:30 I am learning tricircle design documents. plan to start resource update function next week 13:15:21 you can discuss any questions in mail-list or irc 13:15:56 Hello, there are several patches in review 13:16:28 I hope that everyone can spend some time to review other's patch 13:17:15 you can have more broad knowledge on what's being developed, and improve the quality of each patch 13:18:36 ok. 13:18:40 ok, got it 13:18:45 sure 13:18:56 ok 13:19:07 ok 13:19:11 ok 13:19:38 thank you all, we are a team, team work can make everything better 13:19:53 #topic Py35, Ubuntu Xenial gate and check jobs 13:21:30 according to community requirement 13:21:51 all project needs to adapt to pyhon3 13:22:40 and tested in ubuntu xenial release 13:23:16 #link https://review.openstack.org/#/c/400493/1/jenkins/jobs/projects.yaml 13:23:39 to support python 3, we need to use "six" library in the list and dict traverse 13:24:29 after the jobs enabled, two more jobs will be added in the gate and check test, but at first it's not voting job, that means it's ok if it failed 13:24:41 to zhiyuan, yes, some update is needed 13:25:46 when will the two jobs become voting job? 13:26:05 we needs to fix the error in the failure job step by step, at last in this release, we should be able to migrate to python3 and xenial 13:26:13 no, non-voting jobs 13:27:05 after all failure are fixed, I hope by the end of Ocata release 13:27:23 the earlier, the better 13:27:30 the sooner, the better 13:28:04 it'll take one week or so to enable these jobs 13:28:37 and one more job doc-publisher-site: docs.openstack.org is also added 13:28:43 i think we just enable the jobs then we know what error needs to be fixed 13:29:26 that means all docs of tricircle will be published to docs.openstack.org website 13:30:20 the jobs will report failure, but will not block any patch's merge 13:32:06 so if the doc job fails, the doc will not be updated in the website? 13:32:30 it's better if someone of us can install tricircle under python35, and execute the pep8/unit/functest to see what error occured 13:32:59 yes 13:34:16 infra team always suggest to enable a job as non-voting at first, after it's stable for the job, then make it as voting job 13:35:14 currently our doc test is successfully for each patch 13:35:51 we can use virtualenv to setup a python3 environment 13:36:22 But it's strange that all jobs are already ubuntu-xenial based 13:36:53 for example https://review.openstack.org/#/c/396564/, I am not sure whether we need to enable special job for xenial 13:37:18 yes, venv is a good way to try python3 13:37:28 or container 13:38:19 i think our test is already runnint on xenial, we just need to enable python3 test 13:38:49 will consult that in infra team 13:39:22 #action consult infra team whether tricircle needs to enable special xenial jobs or alreay support that 13:40:58 #topic open discussion 13:41:09 any topic to discuss? 13:41:45 for me, there may be risk to support shared vxlan network in this circle 13:41:45 no for me 13:42:25 if we are going to follow the "vtep info in port" change in neutron 13:43:40 to zhiyuan, understand the risk 13:44:02 we need to work closely with anil for this feature 13:44:18 at the same time, enhance Tricircle basic capabilities 13:44:41 for example, qos, update operation, metering, etc 13:45:07 do you have other proposal to reduce the risk? 13:45:22 or bypass the risk? 13:47:29 there's other solution to support shared vxlan network, like the "remote port" way, but if we would like to use the "vtep info in port" feature, it will be a waste to implement "remote port" in this circle 13:50:07 we can discuss if it's possible, or how to reduce the waste, at least we need to know how Anil will implement the VTEP info in port bindings 13:51:33 ok, so let's keep our eye on Anil's work 13:52:00 at least we can finish the bridge network combination change 13:52:05 sure 13:52:31 actively involve in Anil's improvement 13:52:38 if necessary 13:53:11 ok, other topics? 13:53:18 no 13:53:28 no 13:53:30 no 13:53:31 no 13:53:32 no 13:53:42 ok, thank you for your attending 13:53:49 #endmeeting