22:05:06 <jeblair> #startmeeting zuul
22:05:07 <openstack> Meeting started Mon Jan 23 22:05:06 2017 UTC and is due to finish in 60 minutes.  The chair is jeblair. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:05:08 <fungi> oh, yeah, meeting
22:05:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:05:11 <openstack> The meeting name has been set to 'zuul'
22:05:15 <jeblair> #link agenda https://wiki.openstack.org/wiki/Meetings/Zuul
22:05:27 <jeblair> #link previous meeting http://eavesdrop.openstack.org/meetings/zuul/2017/zuul.2017-01-09-22.03.html
22:05:53 <jeblair> #topic Announcements
22:06:43 <jeblair> I'd like to remind everyone to please frequently review zuulv3 related changes, and to use a bookmark in gerrit or gertty dashboard
22:07:03 <jeblair> the search query you should use is: "status:open (branch:feature/zuulv3 OR topic:zuulv3 OR branch:feature/gearman-zk-shim)"
22:07:23 <jeblair> #info use gerrit search query:  "status:open (branch:feature/zuulv3 OR topic:zuulv3 OR branch:feature/gearman-zk-shim)"
22:08:02 <jeblair> that's updated with the latest based on mordred's zk shim work
22:08:09 <jeblair> #topic Actions from last meeting
22:08:17 <jeblair> mordred describe plan for zuulv2->nodepoolv3 shim in mailing list post
22:08:20 <jeblair> that is completed
22:08:48 <jeblair> the branch exists
22:09:06 <jeblair> and morded has a change to it that alters the .gitreview and the README.rst files: 421478
22:09:47 <jeblair> #topic Status updates: Nodepool Zookeeper work
22:10:13 <jeblair> aside from that change, i don't think mordred has anything further on the zk shim
22:10:52 <jeblair> i've started working on the zuul-nodepool integration job
22:11:04 <jeblair> not quite running yet, but should have something soon
22:11:31 <pabelanger> I wouldn't mind trying an experimental job for some ansible things I've been working on for integration testing, if we are open to that
22:12:17 <pabelanger> bascially bootstraps a node with zuul / nodepool / zookeeper, makes sure things are running with a simple image build right now
22:13:01 <jeblair> pabelanger: well, when i proposed this job, i wanted to keep things very simple and fast, so i'm running nodepool with a fake image so we don't have to test the image building
22:13:25 <jeblair> pabelanger: (nodepool has the devstack test which covers image building; zuul doesn't need to test that)
22:13:27 <pabelanger> jeblair: sure, the same could be done with the playbooks
22:14:10 <jeblair> pabelanger: ok, how about this -- i proceed with my "just spawn the daemon in a shell script" approach until we get it working, then look at making it nicer with playbooks
22:14:25 <pabelanger> sure, I don't want to stop progress
22:15:31 <jeblair> i think the deployment tooling we will need for this will be very simple, so we can get it in place soon.  but i know we're going to want more deployment tooling for other stuff later, so i think it would be fine to use it in this job as long as we aren't making the job itself too complex
22:16:44 <Shrews> I'm getting really close to having nodepool actually satisfy requests. Then we can work on re-enabling tests after that.
22:16:56 <Shrews> (again)
22:17:23 <jeblair> then it sounds like we might be pretty close to needing this job!  :)
22:18:38 <jeblair> #topic Status updates: Devstack-gate roles refactoring
22:18:49 <jeblair> anyone have anything to discuss about this?
22:18:59 <rcarrillocruz> my update is
22:19:04 <rcarrillocruz> that i will have updates soon
22:19:13 <rcarrillocruz> sorry for been away lately
22:19:30 <rcarrillocruz> new job + new city + house still not reeady to live in is proving to be stressful :/
22:19:40 <rcarrillocruz> clarkb: ^
22:20:33 <jeblair> rcarrillocruz: hang in there :)
22:20:38 <rcarrillocruz> ++
22:21:10 <jeblair> #topic Status updates: Zuul test enablement
22:21:51 <jeblair> there are some stale changes for this... i've been poking people privately to remind them.  hopefully they should get updated soon.  otherwise, i don't think there's much progress on this front recently?
22:23:14 <jeblair> without SpamapS around, i'm going to skip the progress summary
22:23:21 <jeblair> #topic Open Discussion
22:23:44 <jeblair> I've started work on the "actually run ansible playbooks" part of the work
22:23:52 <jeblair> which is probably one of our last big hand-wavey things
22:24:12 <fungi> hopefully a lot of that is liftable from master?
22:24:17 <Shrews> \o/
22:24:18 <jeblair> fungi: none of it
22:24:25 <fungi> or is there a major divergence in how that will work
22:24:29 <fungi> yikes
22:24:42 <jeblair> *running* ansible isn't the hard part, that's what comes from master
22:24:53 <jeblair> the hard part is running native ansible playbooks defined in zuul repos
22:25:01 <jeblair> rather than auto-generated playbooks from jjb
22:25:07 <fungi> ahh, yep that would count as novel still ;)
22:25:26 <jeblair> so this is the opportunity for the folks who have said "ansible, oh cool, i'd like to help with that" to help with that.  :)
22:25:48 <jeblair> please follow those changes and review them.  there will be many places for people to jump in and help.
22:25:53 <jeblair> but it will require engagement.
22:26:41 <jeblair> #link run native ansible playbooks https://review.openstack.org/423550
22:26:45 <jeblair> it starts with that change ^
22:26:56 * Shrews avoids a lame marriage joke
22:28:13 <jeblair> anyone have anything else?
22:28:27 <Shrews> nope
22:29:16 <fungi> the change to mark the nodepool zk work as implemented merged earlier today
22:29:24 <fungi> (and there was much rejoicing)
22:29:42 <Shrews> fungi: oh? review #?
22:29:51 <fungi> the zk-based image builds that is
22:30:23 <fungi> #link https://review.openstack.org/424229 Mark nodepool image workers specs as complete
22:30:38 <jeblair> oh, and that reminds me, i put this change up on the infra agenda tomorrow -- it adds some info to the v3 spec that slipped through the cracks:
22:30:54 <jeblair> #link https://review.openstack.org/410442 Update zuulv3 spec to include job repo information
22:31:56 <jeblair> hopefully that's not too surprising to anyone
22:32:27 <rcarrillocruz> ah, that's neat
22:32:34 <rcarrillocruz> and makes a lot of sense
22:32:56 * mordred waves at people, having just gotten back from dinner
22:33:40 <fungi> that'll nicely simplify devstack-gate too if i'm understanding correctly
22:34:23 * Shrews throws a schnitzel at mordred
22:34:23 <jeblair> fungi: exactly, it's a comprehensive replacement for $PROJECTS, and goes in the job definition
22:35:08 <fungi> Shrews: a wet wad of kartoffelsalat would probably work better
22:35:46 * mordred throws an epic amount of Schweinhaxl at Shrews
22:37:09 <jeblair> thanks everyone!  this seems like as good a time as any to
22:37:11 <jeblair> #endmeeting