13:00:52 #startmeeting senlin 13:00:53 Meeting started Tue Mar 28 13:00:52 2017 UTC and is due to finish in 60 minutes. The chair is Qiming. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:54 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:56 The meeting name has been set to 'senlin' 13:00:59 evening 13:01:12 hi 13:01:37 hi, yanyan 13:02:02 evening : ) 13:02:09 evening, Ruijie_ 13:02:16 hi Qiming 13:02:48 hi yanyanhu 13:02:48 ethan texted me that he will join a bit later 13:03:03 hi, Ruijie_ :) 13:03:42 let's get started 13:03:46 ok 13:04:15 #topic pike work items 13:04:23 #link https://etherpad.openstack.org/p/senlin-pike-workitems 13:04:59 not sure we have progress in API 13:05:23 Ethan is not here to give us an update on feature rich server 13:06:08 Engine improvement, CLUSTER_CHECK is still there for review 13:06:25 Node adopt 13:06:31 I'm still working on it 13:07:00 got some problems when implementing this at profile base layer 13:07:07 https://review.openstack.org/446867 13:07:09 this one 13:07:40 right, that one is about adopting a heat stack 13:08:08 a more trick thing is about doing this at profile base class 13:08:21 we will need a profile object to invoke do_adopt() method 13:08:43 we can use a fake node object for this method to work 13:09:02 however, we still need a profile instance in order to invoke its do_adopt() 13:09:07 can we define do_adopt as class method? 13:09:17 when we create a profile object, we need a spec 13:09:38 a class method cannot be used to hole the _computeclient ... properties 13:09:54 oh, right 13:10:23 or we can just do a class method, but we don't use the _computeclient ... instance property 13:10:59 still trying to find a decent workaround 13:11:08 a little tricky here... 13:12:00 I hope we can work this out without introducing messy workarounds 13:12:22 understand 13:12:29 next thing is about scaling improvement 13:12:42 Ruijie_, do you have an update? 13:12:56 yes Qiming, filed a bp 13:13:11 oh, ... I didn't notice it 13:13:16 had a plan, can we add the parameter to scaling_policy ? 13:13:55 I'd plan to add one parameter to scaling_policy and request obj to support health_check 13:14:02 I think so 13:14:16 it is something optional 13:14:17 so we need to bump the version of scaling policy 13:14:41 yes 13:15:05 okay Qiming, will start this work soon. 13:15:09 it means a property that is only supported by a new version 13:15:12 thanks 13:15:24 np :) 13:15:51 We don't have xuefeng online today? 13:16:02 he said on wechat 13:16:34 so, we can skip rdo item 13:16:53 no update on health 13:17:31 policy one 13:17:35 just reviewed 13:17:40 it looks fine to me 13:18:01 thx Qiming. 13:18:18 may need to revise this part when doing health check 13:18:35 that's all for the etherpad 13:18:48 #topic summit prep 13:19:11 have any of you confirmed your presence at the summit? 13:19:42 Qiming, will you join it? 13:19:48 yes, I will 13:19:52 great 13:20:07 may not be able to boston :( 13:20:23 I guess xinhui and ethan will join the summit as well? 13:20:33 I believe xinhui is going and maybe also xuefeng 13:21:08 ok, so you do have several presentations need to cover 13:21:16 I'm a little worried about the mistral integration talk, since they haven't shown up for quite a few weeks 13:21:26 ... 13:21:44 has xinhui communicated with them about the detail? 13:21:59 like content preparation and presentation arrangement 13:22:01 not sure 13:22:08 she looks very very busy 13:22:24 ok, hope to have a meet after qingming festival 13:22:53 we don't have a lot time to prep 13:23:13 very curious about the mistral support 13:23:23 yes 13:23:27 only a month 13:23:35 ethan mentioned that their VNF case is in pretty good shape 13:23:41 that is really good news 13:23:46 good to hear that 13:23:56 #topic open discussions 13:24:08 anything else you want to add? 13:24:15 nope from me 13:24:22 em Qiming, I met a problem about the lock 13:24:30 yes? 13:24:38 Cluster is already locked by action [u'b112a879-41ac-4581-97e1-a0d1e27c2cf1'], action e61586b8-8c0f-4b46-9eb8-e9c0673b9cf0 failed grabbing the lock 13:24:47 this cluster had been locked for days :) 13:25:04 engine will not clear the dead lock? 13:25:10 it cannot 13:25:20 unless the engine is dead 13:25:32 seems like a bug of action 13:25:46 what's the status of the b112 action? 13:26:10 let me check it 13:26:37 its cluster_scale_in 13:26:49 and the status reason is 'Action execution failed' 13:27:18 is is possible that we provide API or anything to clear the dead lock 13:27:23 em, ... not sure why that would happen 13:27:44 if you have an idea, feel free to propose it 13:28:16 okay Qiming, will lock forward to it :) 13:28:19 the thing is that we have to make sure the db is consistent 13:28:45 looks to me like a bug 13:29:01 it would be helpful if you can help reproduce it 13:29:23 sure Qiming 13:29:28 anything else? 13:29:50 no from me :) 13:29:59 alright 13:30:04 elynn, evening 13:30:15 we are about to end the meeting, :D 13:30:18 o/ 13:30:21 .... 13:30:24 haha 13:30:26 okay... 13:30:40 you can check the meeting minutes 13:30:43 it is short 13:30:49 do you have anything to share? 13:30:54 I will~ 13:31:01 No... 13:31:16 thank you for joining, :) 13:31:22 :P 13:31:38 we can release the channel early today 13:31:44 yea, looks so 13:31:59 #endmeeting