15:16:29 #startmeeting openstack_ansible_meeting 15:16:30 Meeting started Tue Jul 10 15:16:29 2018 UTC and is due to finish in 60 minutes. The chair is cloudnull. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:16:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:16:34 The meeting name has been set to 'openstack_ansible_meeting' 15:17:20 #topic roll call 15:17:26 o/ 15:17:39 o/ 15:17:45 i thought meeting was in one hour... 15:18:12 o7 15:18:36 I have it now on my calendar though I'd be happy to be wrong :) 15:18:44 \o 15:19:03 "Join us on IRC, in #openstack-ansible, on Tuesday at 16:00 UTC, every tuesday." utc is now 15:19 15:19:41 * cloudnull may be living in the past 15:20:35 we good to continue now or should be pull the eject and reconvene when I can tell time? 15:20:59 we can continue now i don't mind 15:21:57 anyone else? 15:22:26 ok 15:22:26 #topic bug triage 15:22:46 lets start with the issue spots brought up just a min ago 15:22:57 https://bugs.launchpad.net/openstack-ansible/+bug/1778098 15:22:57 Launchpad bug 1778098 in openstack-ansible "os_horizon role fails if horizon_custom_themes is specified" [Low,Confirmed] - Assigned to Amy Marrich (amy-marrich) 15:24:43 spotz so what do you need from us on that ? 15:24:52 anything that we can help with? 15:25:06 So from reading the comments, I'm not sure it'd just a doc (me) update 15:27:02 As it's low I can attempt the code change but it'll take me a while to get to and figure it but I'll be travelling for the next week 15:30:04 is Dmitriy R. around ? 15:30:12 * cloudnull was reading the issue 15:30:24 looks like there's a couple approaches being talked about 15:31:10 guilhermesp -cc ^ 15:34:37 cloudnull: Yeah so from reading the added research it does seem we might need to make a change AND then doc it 15:34:37 hey cloudnull and everyone. I was having lunch :D 15:34:41 let me read 15:34:47 No lunch for you guilhermesp:) 15:34:53 sorry we started the meeting early 15:35:16 i think that's not the point, here is UTC -3 hahah 15:36:14 spotz: I can continue to reproduce this bug if you can't at the moment 15:36:28 what I did was creating a new variable in horizon_role to point a custom_theme_path 15:36:45 was there a change to the roles that you needed to make it work ? 15:37:09 opps IRC lag :) 15:37:15 do you have a diff of the changes? 15:37:35 maybe we could get that into the role and then spotz can doc the update? 15:37:36 just a sec, need to access my test vm 15:37:45 thanks! 15:38:54 i defnetly need a new laptop, sloooow 15:39:38 new X1 :d 15:40:02 cloudnull, are you buying for all of us ? 15:40:29 good idea :) 15:40:42 cloudnull: going to paste the diffs now, T-10 15:40:56 https://i.imgflip.com/2dqpvx.jpg 15:41:10 excellent 15:41:28 guilhermesp we'll take a look at that change and see what needs to be implemented and doc'd 15:41:45 http://paste.openstack.org/show/725423/ 15:41:54 for now, lets move on 15:41:58 https://bugs.launchpad.net/openstack-ansible/+bug/1780733 15:41:58 Launchpad bug 1780733 in openstack-ansible "Neutron rootwrap error" [Undecided,New] 15:42:07 after meeting I give some more details 15:42:14 thanks guilhermesp :D 15:42:31 jrosser did you track that issue down ? 15:43:02 something that we need to change in the rootwrap config ? 15:43:27 maybe something we can fix by pulling those configs out of the venv ? 15:45:53 oh right yes 15:46:10 it sounded like something similar had popped up for cinder 15:46:22 but i don't know about that or what the fix was 15:46:56 I wonder if this is an ocata specific issue? 15:47:06 anyone else seeing anything similar ? 15:47:11 d34dh0r53 mgariepy ? 15:47:20 ocata? 15:47:52 I have not seen that, the bug says queens though 15:47:56 's/ocata/queens/g' 15:47:59 haven't seen it yet 15:48:11 was reading a different issue and was confusing myself 15:48:13 this happens absolutely reliably if i run tempest 15:48:34 what is the process that it's trying to kill ? 15:48:43 hmm, which log is that? 15:48:54 we may be seeing that in our gates during the tempest runs 15:49:23 d34dh0r53 mind digging into that and seeing if you're seeing the same ? 15:49:28 well if theres anyone who knows about this stuff i can dig more details and put them onto the bug 15:49:36 i just need pointing at what info is useful 15:49:43 jrosser: yes, please cloudnull: sure 15:52:41 cool! so marked triaged, assigned to d34dh0r53 - lets see if we can confirm this issue today. 15:52:53 cloudnull: will do 15:53:10 thanks! 15:53:13 next 15:53:16 #link https://bugs.launchpad.net/openstack-ansible/+bug/1780385 15:53:16 Launchpad bug 1780385 in openstack-ansible "[os_heat][pike] heat-container-agent fails to communicate with Keystone" [Undecided,New] 15:54:41 looks like we need to instruct heat to use the public VIP ? 15:55:15 or maybe we just need to doc it? 15:56:15 seems sensible to me that the service use the internal VIP by default however in the event that's unreachable by some services setting the admin/public vip may be needed. 15:56:30 thoughts? 15:57:43 surely it has to be the external vip, as you shouldnt really be talking to the internal one from a VM? 15:58:47 i think that may be a magnum specific limitation? 15:58:52 DimGR 15:58:53 ? 15:59:15 yo 15:59:39 DimGR could you look at https://bugs.launchpad.net/openstack-ansible/+bug/1780385 and see if that makes sense . 15:59:39 Launchpad bug 1780385 in openstack-ansible "[os_heat][pike] heat-container-agent fails to communicate with Keystone" [Undecided,New] 15:59:45 meetime? 15:59:50 looking 16:00:01 it seems magnum + heat related 16:00:09 prometheanfire: cloudnull is living in the past so we started early 16:00:22 but it may just be we have the heat config set incorrectly 16:00:29 prometheanfire: which makes you late ;) 16:01:48 boooo 16:02:08 heheh 16:06:36 lets put a pin in that for now . 16:06:56 DimGR if you have a chance to have a look at that and provide some guidance it'd be greatly appreciated. 16:07:13 https://bugs.launchpad.net/openstack-ansible/+bug/1780011 16:07:13 Launchpad bug 1780011 in openstack-ansible "OSA Queens: libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory" [Undecided,New] 16:07:17 i should not call this a bug but more like a config issue . Endpoints in heat.conf should contain accesible hostnames for heat to work 16:07:23 ok i will update the ticket 16:07:54 thanks DimGR ! 16:08:13 spotz if have a moment mind taking a crack at documenting that limitation ? 16:08:19 or anyone for that matter? 16:08:45 cloudnull: If details are in there I can give it a shot before I leave town 16:09:00 as for the libvirt issue. I suspect we just need to make sure that nova-compute is started after libvirt which we can do in our systemd template. 16:09:16 spotz no rush, but it'd be great to close that out 16:09:59 i'll take 1780011 16:11:08 ok thats all the new Undecided issues. 16:11:33 #topic open discussion 16:11:39 could you take a look at https://bugs.launchpad.net/openstack-ansible/+bug/1657742 16:11:39 Launchpad bug 1657742 in openstack-ansible "Add default roles used by Barbican" [Undecided,New] - Assigned to Vadim Kuznetsov (vakuznet) 16:12:06 vakuznet sure. 16:12:11 lets have a look 16:13:17 vakuznet looks like those role just need to be added to the os_barbican tasks 16:13:30 yes 16:13:34 do you have some cycles to get a patch up? 16:13:51 ready to commit 16:14:44 looks like we just need to have the additional roles add here https://github.com/openstack/openstack-ansible-os_barbican/blob/master/defaults/main.yml#L58-L59 16:15:23 if you could have a review prep'd for that we can get that bug closed out and we can backport the fix to the needed branches. 16:16:29 cloudnull a bit more then that 16:16:45 ok? 16:17:39 I've confirmed the bug. 16:17:55 vakuznet if you have a pr for that I'd be happy to help review it 16:18:08 or if you have questions I'd be happy to help investigate. 16:18:17 ok thanks 16:18:35 anyone else have an airing of grievances or something they want to talk about? 16:19:06 cloudnull if you have a moment, can you swing by https://review.openstack.org/#/c/577240/ 16:19:22 oh course :) 16:19:39 thanks, it will help with visibility 16:19:39 done 16:19:45 gracias 16:19:57 will take a few to merge but that should be in soon-ish 16:20:05 ok welp. thats all folks 16:20:18 thanks for hanging out, sorry I couldn't tell time 16:20:40 #endmeeting