01:00:32 #startmeeting Compass 01:00:33 Meeting started Thu Aug 27 01:00:32 2015 UTC and is due to finish in 60 minutes. The chair is wshao. Information about MeetBot at http://wiki.debian.org/MeetBot. 01:00:35 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 01:00:37 The meeting name has been set to 'compass' 01:00:57 #info Weidong 01:01:16 Let's give a few minutes. 01:01:41 the main agenda is to check on Ansible status. 01:01:43 OK 01:03:43 xichengc: do you know the current status? 01:04:08 From my part, i have the parameters sorted out. 01:04:31 ok. ready to work on the adapter related conf changes? 01:04:31 and will start writing metadata for the new adapter 01:04:41 as well as adding adapters 01:04:53 ok cool. sounds good. 01:04:55 but the adapter confs will still be written in the "old way" 01:06:24 that is fine. 01:06:46 it is more important to have it in the workflow, making it easy to verify the actual deployment 01:06:51 we should wait for baiguoku's update on his part... 01:07:05 ok. I will follow up on that. 01:07:14 will he join us today? 01:07:26 For compass-ci, could you remove the failing job? 01:07:46 it triggers emails 01:07:59 I have already removed the failing periodic jobs 01:08:06 now you only see the successful ones. 01:08:47 those failing jobs are based on old dev branch which was deprecated long time ago. 01:08:51 i see. 01:09:51 Is compass-ci all good now? with logs available? 01:10:01 i mean, available through public IPs 01:10:04 yes 01:10:20 all right, cool. 01:10:24 http://12.234.32.108/logs/ 01:10:44 you need to give status server and log server two domain names 01:10:52 I will add that. 01:11:14 i also customized the zuul status page a little bit 01:11:18 with compass logo and everything 01:11:26 http://12.234.32.107/ 01:11:36 nice. 01:11:59 make sure everything is git-trackable 01:12:07 including the zuul config. 01:12:30 yes, all the changes i made have been added to compass-core/project-config 01:13:14 meimei1: do you know the status of ansible (for kilo) testing status? 01:14:32 the deployment of all-in-one is failure 01:14:35 #info chenshuai 01:14:47 said by xuhan 01:14:52 ok. 01:15:10 I am not very familiar 01:15:12 do you know what the errors are? 01:15:20 download galera db failed 01:15:22 you mean the upstream os-ansible-deployment? 01:15:26 yes 01:16:33 is this the only problem? 01:16:35 For this type of errors, you could email us in the openstack-dev mailing list (with [Compass] prefix). or unicast to me. 01:17:50 i know Chef allows setting of a local repo, or we could do this at the OS installation step. add the local repo with high precedence 01:18:06 previously problems has been resolved 01:18:20 xichengc: could you help on this offline? 01:18:34 what do you mean by "chef allows setting of a local repo"? 01:19:00 you meant apt repo? 01:19:03 for example? 01:19:14 osad will git clone all openstack repo, at a slower speed. 01:19:15 chef provider allows custom repo setting, as I recall. But in Compass, we config this at the OS level. 01:19:40 I see. 01:20:00 i remember osad has the repo settings in conf files right? 01:20:08 or hardcoded in tasks? 01:20:38 baigk: xichengc's question is for you 01:20:58 or xuhan 01:22:04 ok. let's address the error after the meeting. 01:22:38 I do not have anything else on the agenda. anything else? 01:23:07 I do have something less prioritized though... 01:23:41 sure 01:24:15 if we have time, we should consider giving our 3rd testing suites(zuul, jenkins, gearman) the ability to avoid racing problems. 01:24:45 for example, change A + master head works, change B+ master head works, but A+B+master head doesn't 01:25:21 zuul actually has the ability to run test against all pending changes to determine the potential failures. 01:25:27 i see. I am not familiar with this. please write in an email? 01:25:41 but scalability could be a problem if we have many contributors. 01:25:47 sure. 01:26:17 xichengc: yes, it's in defaults/repo_packages/openstack_clients.yml 01:26:27 thanks baigk 01:27:08 we can discuss offline about this. 01:29:29 baigk 01:29:44 can we have a short phone call after this? 01:30:14 some tasks are very slow , like "download lxc","Build all known python packages requirements" and "Create a build report for all known packages within a release" 01:30:23 ok 01:30:43 I will call you on espace 01:31:02 anything other issues/questions/topics? 01:31:12 especially from Shanghai team? 01:31:58 countdown 5 01:32:00 countdown 4 01:32:02 countdown 3 01:32:05 countdown 2 01:32:07 countdown 1 01:32:09 #endmeeting 01:32:25 Thank you all. 01:32:42 Though I don't think I can end it. 01:32:45 weidong is back on.. 01:32:55 Done? 01:32:59 wshao_ please #endmeeting 01:33:02 i am not allowed to 01:36:44 #endmeeting 01:36:54 Will do it later 01:43:07 #endmeeting