13:02:44 #startmeeting senlin 13:02:45 Meeting started Tue Jan 19 13:02:44 2016 UTC and is due to finish in 60 minutes. The chair is Qiming. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:02:46 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:02:48 The meeting name has been set to 'senlin' 13:02:56 sorry, a bit late 13:03:08 o/ 13:03:22 trapped by the placement policy 13:03:22 hi 13:03:53 pls check meeting agenda 13:04:09 very short one, :) 13:04:12 #link https://wiki.openstack.org/wiki/Meetings/SenlinAgenda 13:04:35 first things first, work items etherpad 13:04:43 #link https://etherpad.openstack.org/p/senlin-mitaka-workitems 13:05:13 I tried keep updating this page 13:05:34 location header can be got now, elynn? 13:06:01 yes 13:06:16 I try and can get location now. 13:06:21 cool 13:06:27 But openstacksdk doesn't release new tag. 13:06:39 for action header, may need to add that asap 13:06:45 only master support location. 13:07:05 Action header? 13:07:05 yes, after action header is added, we can push them to do a 1.0 release 13:07:27 oh, sorry, location header points to action 13:07:54 see what a mess I'm 13:08:19 we can get location header now 13:08:31 great 13:08:41 I can split location string and get action id in heat. 13:09:04 moving on, I don't think anyone is working on advanced filters 13:09:26 so you will continue to work on the heat resource type thing 13:09:34 have forgotten it 13:09:42 Qiming: yes 13:09:54 haiwei, no worry, we have been talking empty, :P 13:09:55 I will add more resource type in heat for senlin in this release 13:10:02 great 13:10:12 senlinclient test case coverage is 91% 13:10:38 there are a few units a little bit tricky to test 13:10:53 will make a new release? 13:10:53 we can get back to that after this m2 release 13:11:12 yes, haiwei, client should be released first 13:11:22 I think it is in pretty good shape now 13:11:37 seems so 13:12:21 health management part, xinhui has been pushing that end 13:12:29 great progress so far 13:12:56 we won't expose those interfaces from API in m2 13:13:22 why not 13:13:25 we can discuss later when those bits all merged 13:13:38 and see if it is approriate 13:13:58 haiwei, we can never deprecate things once exposed from API 13:14:19 that's the reason why we spent so much effort on deleting things 13:14:46 will we make a release of API, and put health polices into micro-version? 13:14:50 Documentation part, no progress, I guess 13:15:01 haiwei, that is an option 13:15:17 before doing micro-versioning, we need an infra to support it 13:15:34 micro-version specs are under review in openstack-specs project 13:15:37 ok, we don't need to make API release so fast I think 13:15:41 we should watch that 13:16:05 sure, keep the minimum set of APIs, and make them usable, stable 13:16:28 no progress on profile update, right? 13:16:45 yes 13:16:53 didn't get time to work on it... 13:17:09 no worry, not high priority at the moment 13:17:36 yea, will focus on policy this week 13:17:46 okay, the highest priority is to make policies work 13:18:20 especially those ones which used to work before we remove soft-delete, priority, level, cooldown ... etc 13:18:39 that will be the baseline for this m2 release 13:18:49 yes, lots of thing need to clean 13:18:57 yep 13:20:16 deleting clusters which have receivers associated 13:20:25 this one has been solved 13:20:35 em ... can be improved though 13:20:58 so it is no longer a bug 13:21:13 can we remove it? 13:21:19 yes, I think so 13:21:23 yes 13:22:14 policy fix is not related to receiver 13:22:31 lb rework was listed on line 46 13:22:48 builtin policy captured on line 43 13:22:49 yes, so we can remove this item 13:22:55 do it! 13:22:56 it is in wrong place 13:23:05 done 13:23:06 :P 13:23:10 it feels good deleting things! 13:23:20 sure, haha 13:23:24 engine status listing 13:23:38 haven't started working on functional test for receiver as well 13:23:39 that depends on a db table, which means a db version bump 13:23:47 Qiming: yes 13:23:50 yanyanhu, no worry 13:23:52 hope can start this work from next week 13:23:53 I will modify my patch 13:24:14 have to prioritize things 13:24:26 okay, done 13:24:26 add 02_xxx.py for db migration. 13:24:29 :) 13:24:52 anything missed from work items etherpad? 13:25:26 nope, I think 13:25:28 can you give me the link? 13:25:46 https://etherpad.openstack.org/p/senlin-mitaka-workitems 13:26:35 #topic high priority reviews/bugs 13:26:50 about deletion policy for cross az/region one, It seems I am not in the right way 13:27:38 haiwei, will propose a revision to region placement policy, hopefully today 13:28:20 ok, I will do deletion policy fix based on it 13:28:35 cool 13:29:32 reviews 13:29:37 #link https://review.openstack.org/#/q/project:openstack/senlin+status:open 13:30:52 still lots of patches need review 13:31:20 yep, wondering if we can/should swallow them all 13:32:16 hope so. In worst case, leave those ones which are not necessary for m2 release 13:32:21 to next week 13:32:40 okay, just need eyes from everyone 13:32:51 #topic open discussion 13:33:11 Will~ 13:33:18 a link to share with you: http://superuser.openstack.org/articles/what-s-new-under-the-openstack-big-tent-monasca-and-senlin 13:33:38 may and may not need a proxy to check it 13:34:11 I can access it 13:34:22 I can see it. 13:34:27 cool 13:34:28 it was very slow last time I checked it 13:34:33 Cool , now we get more attention. 13:34:38 nice article :) 13:34:42 ghost ... :D 13:35:06 truth is we need more 13:35:08 who made this? 13:35:32 we can use openstack wiki as the platform 13:35:37 Qiming, yep, we need more, both eyes and hands :) 13:36:22 haiwei, Nicole is responsible for publishing this kind of info 13:37:45 another thing, we have got some helps again, on api spec 13:37:46 https://review.openstack.org/#/c/263579/ 13:38:48 nice 13:39:01 thanks for those kind help 13:39:26 great job 13:39:30 I have put your name into corresponding tables here: https://wiki.openstack.org/wiki/CrossProjectLiaisons 13:40:28 ok 13:40:36 ok 13:40:44 got it 13:41:01 speak up if you want to take away some job roles from me, :) 13:41:14 my name still appears here and there 13:42:15 I'll do my best to serve you well, but I am still grateful to the ones who are sharing the workload, :P 13:42:35 okay 13:42:41 sure, will be glad to undertake some jobs 13:42:43 anything else? 13:43:04 no further topic from me:) 13:43:10 I have one 13:43:19 I will be glad to help you :) 13:43:32 have anyone tried to create a node with the latest source? 13:44:02 I tried, but failed 13:44:03 Not yet... 13:44:05 latest resource? 13:44:08 not yet 13:44:08 you mean 13:44:12 There appears to be an error in the Senlin article, it states this meeting occurs "Tuesdays at 13:00 UTC in #senlin on IRC" 13:44:17 oh, my eye... sigh 13:44:23 latest source... 13:44:34 cschulz, oh, ... 13:45:38 ... 13:45:55 Qiming has made a fix a few days ago, make a node's cluster_id '' instead of None if the node is not in a cluster 13:46:27 seems introducing new bugs? 13:46:31 I seems it is related to that fix 13:46:39 I am not sure it is a bug 13:47:16 the unit test passed, but when doing by command line it failed with an DB error 13:47:36 okay, keep digging 13:47:59 see if any traces from senlin-engine 13:48:49 maybe it is related to the default None value or something 13:48:52 https://github.com/openstack/senlin/blob/master/senlin/engine/service.py#L970 13:48:56 this line 13:49:21 that line is good 13:49:41 if make cluster_id equal None, it will pass 13:50:10 I also think current source is OK, but it really failed in my test 13:50:13 maybe line 980 should be checked? 13:50:35 not sure about it 13:50:47 will dig it tomorrow 13:50:49 that fix was necessary 13:50:56 maybe it is my fault 13:51:20 but I'm not 100% sure I have covered all possible places where we are assigning value to node.cluster_id 13:51:46 thx, haiwei 13:51:54 nope 13:52:05 are we done? 13:52:16 it is ok for me 13:52:31 ok for me too 13:52:38 thanks for joining, guys, talk to you next week 13:52:42 #endmeeting