13:01:07 #startmeeting tricircle 13:01:11 Meeting started Wed Jan 18 13:01:07 2017 UTC and is due to finish in 60 minutes. The chair is joehuang. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:01:12 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:01:15 The meeting name has been set to 'tricircle' 13:01:30 #topic rollcall 13:01:37 #info joehuang 13:01:46 #info xiulin 13:01:47 #info Yipei 13:02:33 hi 13:02:42 hi joe 13:03:11 #info zhiyuan 13:03:13 hi 13:03:22 #info jiawei 13:03:48 due to the Chinese spring festival is approaching 13:04:04 today ronghui will absence 13:04:09 and most of us will be on holiday from next week 13:04:47 He have something urgent 13:04:55 so the weekly meeting of Jan 25, Feb 1 will be cancelled 13:05:06 to Jiawei, that's fine :) 13:05:31 #info weekly meeting of Jan 25, Feb 1 will be cancelled 13:06:03 #topic Ocata release preparation 13:06:43 there are only few patches in the review queue 13:07:07 #link https://review.openstack.org/#/q/project:openstack/tricircle 13:08:02 let's work on these patches and make it being merged before Jan 24, except QoS spec 13:08:40 Is there any challenge for the goal? 13:09:18 no for me, i only have one patch in queue 13:10:18 Xiuling, how about yours? 13:10:38 no 13:10:49 ok 13:10:52 i only have one too 13:11:15 at the same time, I hope that we can fix the bugs listed in https://bugs.launchpad.net/tricircle as far as possible 13:13:15 this bug no longer exist. how to close it? 13:13:19 https://bugs.launchpad.net/tricircle/+bug/1647284 13:13:19 Launchpad bug 1647284 in Tricircle "api_v1.rst needs updating after AZ deletion" [Undecided,New] - Assigned to Dongfeng Huang (southeast02) 13:13:50 maybe i can fix this one #1647924 Xjob of router setup not triggered 13:14:26 to Dongfeng, just close it 13:14:41 to Zhiyuan, thanks 13:15:08 ok. i'll have a try. 13:15:23 I'll check these bug reports one by one tomorrow 13:15:54 to define the priority 13:16:47 is there a patch for 1647284? you can also add a comment in the bug page to link the related patch 13:16:47 according to Doug's mail http://lists.openstack.org/pipermail/openstack-dev/2017-January/110218.html 13:17:25 no 13:17:51 because Tricircle is one of "cycle-with-intermediary" projects, so the stable branch could be created just before Feb.16 13:19:16 so we can postpone our stable branch creation to Feb.13 13:19:51 after that we test the stable branch until Feb.16, then publish the Ocata release 13:20:51 to Dongfeng, you may add some comments to point out the bug does not exist now, then close the bug 13:20:54 since the deadline is postponed, will we introduce new features? 13:21:11 the problem in patch 1647284 no longer exist since the api_v1.rst has been updated. 13:21:32 mostly bug fix, but not large feature 13:22:03 just change its status to fix committed? 13:22:07 ok 13:23:05 how about mark it as "invalid" 13:23:23 ok. 13:24:19 but new feature could be developed and review in parallel 13:24:36 just not being merged into Ocata branch 13:25:06 new features will be merged after Ocata stable branch is created 13:25:14 is it clear? 13:25:22 yes 13:25:28 yes 13:25:32 yes 13:25:38 yes 13:26:00 yes 13:26:36 #info new feature could be developed and review in parallel, just not being merged into Ocata branch, new features will be merged into master branch after Ocata stable branch is created 13:27:20 #topic PTL election 13:27:42 one more topic, the Pike PTL election is also approaching 13:27:59 will be held during Chinese spring festival 13:28:19 if you want to work on PTL in Pike cycle 13:28:33 please send your self-nomination mail in the mail-list 13:28:51 if more than one candidates, than election will be held 13:29:06 if only one candidate, then no election is needed 13:29:22 I hope that you can have a try 13:31:00 the election week is Jan 30 - Feb 03, send your self-nomination according to the community schedule 13:31:42 is there any question on PTL election? 13:32:32 no 13:32:42 no 13:32:44 no 13:32:45 no 13:33:06 ok 13:33:44 #topic Pike planning and design topics 13:34:24 I prepared one etherpad for Pike cycle design topics and planning 13:34:30 #link https://etherpad.openstack.org/p/tricircle-pike-design-topics 13:41:14 oops, don't know why I can't send a message in etherpad, then I suggest to discuss during PTG 13:42:04 I saw your mesage 13:42:07 I saw your message 13:42:37 we can have a preparation meeting next Tuesday morning 13:42:49 oh, i see your message in etherpad now 13:43:38 design topics could be discussed during PTG 13:43:50 before that we need to do planning 13:44:03 i.e, preparation for the PTG discussion 13:44:06 oh i see 13:44:49 so I suggest let's discuss this etherpad next Tuesday morning 9:30am Beijing time 13:45:14 ok, no problem 13:45:25 how about others? 13:45:29 that's fine 13:45:33 it's fine. 13:45:45 fine 13:45:59 it's planning meeting 13:46:55 #info let's discuss the planning etherpad next Tuesday morning 9:30am Beijing time (UTC1:30 am) 13:47:15 #info planning etherpad https://etherpad.openstack.org/p/tricircle-pike-design-topics 13:47:27 ok, thanks 13:47:53 please input your ideas into the etherpad before the meeting as needed 13:48:20 #topic QoS and LBaaS feature support 13:48:34 hello, how about these two features? 13:49:06 for lbaas, i tried to create a load balancer in tricircle environment 13:50:09 http://paste.openstack.org/show/595364/ 13:50:29 it seems that octavia cannot find the driver 13:50:59 allowed_address_pairs driver 13:51:36 oh allow-address-pairs not supported by central plugin currently 13:51:48 you need to enable this extentions, Tricircle has not supported allowed_address_pairs yet 13:51:56 to zhiyuan, yes 13:52:24 but I doubt it's reported in central neutron or Ocatavia 13:53:06 it is reported in central neutron 13:53:25 the request is still not send to octavia 13:53:39 sorry 13:53:39 opt/stack/octavia/octavia/common/utils.py\", line 51, in get_network_driver\n invoke_on_load=True\n\n 13:54:56 i mean octavia controller worker 13:55:29 hello 13:55:53 the error occured here /opt/stack/octavia/octavia/network/drivers/neutron/allowed_address_pairs.py\", line 47 13:56:24 https://github.com/openstack/octavia/blob/master/octavia/api/v1/controllers/load_balancer.py#L108 13:56:47 here it validate whether the subnet exists 13:57:27 then it create the driver 13:57:57 central Neutron with tricircle plugin does not support address pair yet 13:58:26 need to investigate how to implement it in Tricircle central plugin 13:58:44 ok, got it 13:59:13 It seems that Ocatavia relies on address pair feature 13:59:19 yes 13:59:32 highly depends on it 13:59:43 #info investigate how to implement address pair in Tricircle central plugin 13:59:50 other topics? 14:00:11 no from me 14:00:12 no for me 14:00:16 no 14:00:16 no for me 14:00:32 ok, thank you for attending 14:00:43 wish you have a pleasant holiday 14:00:55 #endmeeting