16:00:09 #startmeeting OpenStack Ansible Meeting 16:00:10 Meeting started Thu Apr 9 16:00:09 2015 UTC and is due to finish in 60 minutes. The chair is b3rnard0. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:11 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:14 The meeting name has been set to 'openstack_ansible_meeting' 16:00:26 #chair cloudnull 16:00:27 Current chairs: b3rnard0 cloudnull 16:00:37 #topic Agenda & rollcall 16:00:47 #link https://wiki.openstack.org/wiki/Meetings/openstack-ansible#Agenda_for_next_meeting 16:01:03 o/ 16:01:10 presente 16:01:10 @' 16:01:47 hi 16:02:02 Hi 16:03:03 hi 16:03:24 i didn't see any action items from last week. anything we want to discuss action items wise? 16:04:02 I guess one action item would be starting the blueprint around EC2 and S3 api support in juno 16:05:00 BjoernT do you have one ready for review ? 16:05:06 uh BjoernT I'm confused - juno already has s3/ec2 api support as good as it gets 16:05:28 s3 middleware is missing in osad 16:05:39 BjoernT are you talking about swift3? 16:05:42 ec2 works if we fix keystone_es2_url 16:05:52 right swift3 16:06:39 IE https://github.com/stackforge/swift3 16:07:54 hmm, so for kilo+ we'd probably have to include s3/ec2 support from outside the standard swift/nova trees - that's a whole different ball-game 16:08:20 doable if the resource allocation is worth the effort 16:09:45 let's get through the rest of the agenda before we go down this rabbit hole 16:09:54 yes 16:10:06 cloudnull: do we want to talk blueprints? 16:10:33 sure 16:10:52 #topic Blueprints 16:11:01 #link https://blueprints.launchpad.net/openstack-ansible/+spec/implement-ceilometer 16:11:20 the one in progress is for ceilometer. alex is gone for the week so he's not here to talk about progress on that at this time. 16:11:44 odyssey4me: we also have https://review.openstack.org/#/c/168976/ 16:11:50 any chance we can scratch the details from launchpad? 16:12:01 leave the detail in the spec and only the summary in launchpad 16:13:08 whats that ? 16:13:30 the ceilometer blueprint - it has all the details in both launchpad and the submitted spec 16:13:46 it's just very verbose 16:13:57 no matter - re: https://blueprints.launchpad.net/openstack-ansible/+spec/tunable-openstack-configuration 16:14:33 ah yes we can pair that down 16:15:10 http://docs-draft.openstack.org/76/168976/1/check/gate-os-ansible-deployment-specs-docs/845ff17//doc/build/html/specs/kilo/tunable-openstack-configuration.html 16:16:50 I have implemented a PoC in a review which seems to have disappeared which is a WIP, but illustrates the basic point and is covered in the complementary spec in https://review.openstack.org/168976 16:17:40 cloudnull fed back that the dict merge approach taken is less friendly, predictable and maintainable than the approach in https://review.openstack.org/168104 16:18:14 I have yet to look into that in detail. My time's focused elsewhere and until I'm done with kilo-related commitments this will stall. 16:18:45 ah, there's the WIP (second page of reviews) :o https://review.openstack.org/165683 16:19:25 ok so lets table this for now. 16:19:31 I'm game for the alternative method, it's just a matter of time and priority right now. 16:19:59 if someone wants to do the analysis and patch the spec and do a test review then that'd be great. 16:20:25 I think we all agree on the objective, it's only the method in question. 16:21:04 so does anyone here have cycles and want to run with that for now? 16:22:10 ok then. 16:22:15 :) 16:22:31 so on the topic of blueprints BjoernT : what say you? 16:23:12 nah, no time, lol 16:23:51 minimal-kilo should probably be changed to implemented 16:24:00 speak now or forever hold your piece. 16:24:18 woof 16:24:26 glance, heat, horizon kilofication should be superceded by master-kilofication? 16:24:58 yes they should 16:25:28 it would seem to me that build-facts-archive and dynamic-inventory-lib could dovetail nicely 16:26:25 both, however, are a stretch in terms of the osad manifesto - which we have yet to get into the details of 16:26:27 yes. that could probably go hand in hand 16:27:09 I think both are useful to deployers, but they stray from the idea of deploying openstack and go more into the realm of running ansible. 16:28:29 this is true, however in the case of inventory i think that is essential to OSAD. and presently our inventory system works but it could be a whole lot better 16:28:55 but itll be a bit of a balance 16:29:44 cloudnull I agree - the inventory library is on the perimeter and would add much value if we did that. 16:30:17 storing an inventory of what's out there, however, seems like a better fit for a deployer's CMDB which the library can consume 16:31:04 +1 16:33:14 so barring anything else on bp's b3rnard0 lets move on 16:33:38 next up 16:33:43 #topic Milestones 16:33:53 anything specific there to be discussed? 16:34:08 i know we still have 3 bugs outstanding for 10.1.3 16:34:08 we're actively working on kilo 16:34:50 b3rnard0 which bugs? 16:35:12 #link https://launchpad.net/openstack-ansible/+milestone/10.1.3 16:35:17 andymccr did the work for the glance / swift chunking https://bugs.launchpad.net/openstack-ansible/+bug/1436647 16:35:19 Launchpad bug 1436647 in openstack-ansible trunk "Make chunk size configurable for swift-backed glance" [Medium,In progress] - Assigned to Andy McCrae (andrew-mccrae) 16:35:25 and thats merged. 16:35:57 yup 16:36:24 how about https://bugs.launchpad.net/bugs/1438118 16:36:25 Launchpad bug 1438118 in openstack-ansible juno "some v10 Juno playbooks containing a wrong host pattern" [Low,Triaged] - Assigned to Serge van Ginderachter (svg) 16:36:28 also merged 16:37:13 svg worked on that and its merged. 16:37:22 last one is https://bugs.launchpad.net/bugs/1430872 16:37:23 Launchpad bug 1430872 in openstack-ansible trunk "Add token dhcp_domain to dhcp_agent.ini template" [Medium,Confirmed] - Assigned to Jesse Pretorius (jesse-pretorius) 16:37:47 the listed status in the list is throwing me off :-P 16:37:47 ok, that's on me - I haven't had a chance to actively work on it 16:38:14 is that a hard requirement for 10.1.3 or can we push out to 10.1.4 16:38:32 I took a look a day or two ago and realised that it would involve a new template and a bit more than a simple nova.conf entry, so I pushed it out. 16:38:41 so yes, good question b3rnard0 16:39:02 i think it should be a requirement for 10.1.3 its been an ask for a while now. 16:39:33 looking at you BjoernT 16:39:43 it's not hard work, but it introduces a fair change to how dnsmasq is used and could put risk into the 10.1.3 release 16:39:59 if we've committed to doing a 10.1.3 release tomorrow, is that still doable or can we get another volunteer, err, active contributor 16:40:12 I would recommend shifting it with my apologies - I can prep the change, but we should hold it until after 10.1.3 16:40:41 ok. lets push it 16:40:52 push it real good? 16:40:58 we patch it already manually so 10.1.3 is good 16:40:59 done. 16:42:45 so b3rnard0 moving on 16:43:08 okie dokie, anything else milestones wise or should we go to open discussion 16:43:32 #topic Open discussion 16:43:40 https://bugs.launchpad.net/openstack-ansible/+bug/1442239 16:43:40 Launchpad bug 1442239 in openstack-ansible "Commit c5d488059d9407f1b9b96552159ffc298c8dc547 is invalidating sshd_config" [Undecided,New] 16:44:21 BjoernT: we have a couple of items on the agenda for open discussion. cloudnull, any of those that need to be discussed? 16:44:32 ok 16:44:42 wow, I thought that was long gone - we had a bug about that ages ago 16:45:25 * Do we have defined criteria for core team members? This would be useful to discuss prior to bringing more in. (palendae) 16:45:48 i believe palendae is at a conference 16:46:09 true, but to address this we should be using https://wiki.openstack.org/wiki/Governance/Approved/CoreDevProcess 16:46:44 cloudnull I believe we've discussed that, although perhaps something should be added to the contributing file in the repo to clarify for the future? 16:47:09 as for Reach agreement regarding osprofiler middleware across openstack services during Kilofication. Available: Yes/no and config strategy? (stevelle) - i think the decision was itll be available but disabled. 16:47:26 odyssey4me maybe we add a point about that 16:47:48 and the last one Clarify the meaning of the Triaged status and decide on how we are going to use that status. [1] - (d34dh0r53) 16:47:58 d34dh0r53 ^ 16:48:41 ahh, didn't realize that was still on the agenda. 16:49:00 d34dh0r53 you cannot escape the agenda 16:50:27 https://wiki.openstack.org/wiki/Bugs#Status according to this the triaged status should be used when "The bug comments contain a full analysis on how to properly fix the issue" which we are not really doing. Just want clarification as to how we're using the Triaged status 16:50:56 I'm +1 on that definition. 16:51:08 as am I 16:51:13 +1 16:51:28 so motion carries . 16:51:52 so in our triage meeting we will set bugs to confirmed once they've had the once over (unless of course it contains the appropriate information) 16:51:55 we should update the contributing with a brub on states 16:52:00 #info d34dh0r53: https://wiki.openstack.org/wiki/Bugs#Status according to this the triaged status should be used when "The bug comments contain a full analysis on how to properly fix the issue" which we are not really doing. Just want clarification as to how we're using the Triaged status 16:52:22 yeh we should be confirming/invalidating bugs in triage meeting, not necessarily resolving them. 16:52:31 and prioritizing 16:52:34 agreed 16:52:58 so BjoernT back to your bug(s). 16:53:14 i just have this shiny new one https://bugs.launchpad.net/openstack-ansible/+bug/1442239 16:53:15 Launchpad bug 1442239 in openstack-ansible "Commit c5d488059d9407f1b9b96552159ffc298c8dc547 is invalidating sshd_config" [Undecided,New] 16:54:03 cloudnull agreed that we should add a blurb on the launchpad states and what they mean to us 16:54:36 we should also define the purpose of the two community meetings - it would help to focus them 16:55:26 odyssey4me +1 16:55:39 on to that bug, this is kinda a duplicate of stuff we encountered ages back, eg: https://bugs.launchpad.net/openstack-ansible/+bug/1404343 and perhaps it's a duplication of https://bugs.launchpad.net/openstack-ansible/+bug/1416626 16:55:40 Launchpad bug 1404343 in openstack-ansible trunk "External CI failures due to SSH issues" [High,Fix committed] - Assigned to Hugh Saunders (hughsaunders) 16:55:42 Launchpad bug 1416626 in openstack-ansible trunk "invalid sshd_config entry after ssh_config.yml task runs" [Low,Confirmed] - Assigned to Miguel Grinberg (miguelgrinberg) 16:56:41 it may be a regression which we need to take care of, but it seems to me that https://bugs.launchpad.net/openstack-ansible/+bug/1442239 is a duplicate of https://bugs.launchpad.net/openstack-ansible/+bug/1416626 16:56:42 Launchpad bug 1442239 in openstack-ansible "Commit c5d488059d9407f1b9b96552159ffc298c8dc547 is invalidating sshd_config" [Undecided,New] 16:56:44 Launchpad bug 1416626 in openstack-ansible trunk "invalid sshd_config entry after ssh_config.yml task runs" [Low,Confirmed] - Assigned to Miguel Grinberg (miguelgrinberg) 16:57:11 1404343 is the cause for 1416626 so not duplicate imho 16:58:25 1416626 seems like the same thing. 16:58:31 so its a bug in the ansible module 16:59:15 as I recall, prior to the master refactor we actually did this in a patch - so we have a regression and need to bring if forward 17:00:05 we are out of time. lets take this convo back to the channel and or the ML. 17:00:14 andymccr: yes your are right 17:00:28 thanks everyone 17:00:35 #endmeeting