19:02:42 <SpamapS> #startmeeting tripleo
19:02:42 <openstack> Meeting started Tue Nov 18 19:02:42 2014 UTC and is due to finish in 60 minutes.  The chair is SpamapS. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:02:43 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:02:46 <openstack> The meeting name has been set to 'tripleo'
19:03:21 <SpamapS> #topic agenda
19:03:21 <SpamapS> * bugs
19:03:21 <SpamapS> * reviews
19:03:22 <SpamapS> * Projects needing releases
19:03:22 <SpamapS> * CD Cloud status
19:03:24 <SpamapS> * CI
19:03:26 <SpamapS> * Tuskar
19:03:29 <SpamapS> * Specs
19:03:31 <SpamapS> * open discussion
19:03:41 <SpamapS> Remember that anyone can use the link and info commands, not just the moderator - if you have something worth noting in the meeting minutes feel free to tag it
19:03:47 <SpamapS> #topic bugs
19:04:06 <SpamapS> #link https://bugs.launchpad.net/tripleo/
19:04:06 <SpamapS> #link https://bugs.launchpad.net/diskimage-builder/
19:04:06 <SpamapS> #link https://bugs.launchpad.net/os-refresh-config
19:04:06 <SpamapS> #link https://bugs.launchpad.net/os-apply-config
19:04:06 <SpamapS> #link https://bugs.launchpad.net/os-collect-config
19:04:09 <SpamapS> #link https://bugs.launchpad.net/os-cloud-config
19:04:11 <SpamapS> #link https://bugs.launchpad.net/tuskar
19:04:14 <SpamapS> #link https://bugs.launchpad.net/python-tuskarclient
19:04:23 <SpamapS> I think we should also add os-net-config
19:04:26 <SpamapS> dprince: thoughts?
19:04:53 <GheRivero> o/
19:05:16 <dprince> SpamapS: yes, lets do
19:05:20 <tchaypo> No progress on my critical
19:05:59 <dprince> SpamapS: I already created the project, etc. So just the link should be good I think. May be missing some infra magic/integration though
19:06:20 <SpamapS> #link https://bugs.launchpad.net/os-net-config
19:06:23 <SpamapS> added to the wiki
19:06:46 <SpamapS> dprince: can you take on adding it to the infra magic machine?
19:06:59 <dprince> SpamapS: ack, will do
19:07:28 <SpamapS> #info dprince will make sure os-net-config has infra-mojo
19:08:09 <SpamapS> 5 critical bugs
19:08:15 <SpamapS> we forgot to talk about UID stability in Paris (DOH!)
19:08:53 <SpamapS> https://bugs.launchpad.net/tripleo/+bug/1385346
19:08:56 <uvirtbot> Launchpad bug 1385346 in tripleo "upstart service unreliable after introducion of pipe to logger" [Critical,Fix committed]
19:09:12 <SpamapS> I ran into the expect-fork brokenness for that one last night
19:09:28 <marios_> SpamapS: at least one more of those 5 is fix committed though
19:09:44 <SpamapS> It struck me that we shouldn't be using logger to syslog things.
19:10:02 <SpamapS> Each of these projects has logging configurations that can just call syslog directly for logging.
19:11:12 <SpamapS> marios_: agreed
19:11:17 <SpamapS> ok seems like there's nobody here
19:11:44 <tchaypo> Four weeks running with no/minimal attendance
19:12:00 <tchaypo> Last week all we did was look for 5 old reviews to pursue
19:12:13 <SpamapS> summit gets brutal
19:12:24 <SpamapS> and then we'll have US Thanksgiving next week
19:12:53 <SpamapS> Anyway, I don't see any other bugs that we really need to discuss, so we'll move on.
19:13:14 <SpamapS> #topic reviews
19:13:41 <SpamapS> #info There's a new dashboard linked from https://wiki.openstack.org/wiki/TripleO#Review_team - look for "TripleO Inbox Dashboard"
19:13:44 <SpamapS> #link http://russellbryant.net/openstack-stats/tripleo-openreviews.html
19:13:47 <SpamapS> #link http://russellbryant.net/openstack-stats/tripleo-reviewers-30.txt
19:13:50 <SpamapS> #link http://russellbryant.net/openstack-stats/tripleo-reviewers-90.txt
19:14:54 <dprince> SpamapS: one more bug I'd mention is https://bugs.launchpad.net/tripleo/+bug/1392732
19:14:55 <uvirtbot> Launchpad bug 1392732 in tripleo "locally cached images no longer work w/ devtest" [Critical,In progress]
19:14:56 <SpamapS> If we give everybody 10 days of summit-lag, so 30 reviews of adjustment, we're still a bit behind on the 3-per-day rate
19:15:02 <tchaypo> Down to just one review with >60 reviews on 30 day stats, where
19:15:27 <dprince> SpamapS: without that fixed my devtest has been horribly broken. Honestly, I kind of dislike acquire-images entirely
19:15:54 * dprince doesn't see why downloading cached images ahead of time wasn't sufficient.
19:16:06 <SpamapS> dprince: I agree and I don't know why we have acquire-images but I hadn't looked deep into it.
19:16:32 <dprince> I feel like we are going the wrong direction with devtest_* scripts in this regard... making things more complex and adding extra args to the devtest_ scripts themselves
19:16:57 <dprince> In other words while I have proposed a fix... I kind of think we might should just revert it entirely
19:17:27 <dprince> SpamapS: anyways, maybe I'll post a review and we can discuss there. Carry on with the reviews...
19:17:49 <GheRivero> +1 to revert it. It needs some polish
19:18:12 <dprince> GheRivero: ack, thanks
19:18:36 <SpamapS> dprince: sounds good, and I do think that we maybe need to have a discussion about it.
19:19:01 <SpamapS> Anyway, while our reviewers have not been as busy, I think we're keeping our heads above water at least.
19:20:02 <SpamapS> We do need to keep working on those older reviews.
19:20:40 <marios_> SpamapS: we called out some last week and assigned them
19:20:52 <SpamapS> awesome
19:21:07 <SpamapS> I would like to move on though.
19:21:21 <marios_> sure
19:21:23 <SpamapS> #topic Projects needing releases
19:21:51 <SpamapS> We've been landing stuff, so seems like releases would be good.
19:21:54 <marios_> i havent done this for a while
19:22:08 <marios_> unless someone new would like to
19:22:20 <SpamapS> I was supposed to add greghaynes to the team that is allowed to
19:22:25 <marios_> in which case i can volunteer to help them
19:22:32 <SpamapS> but he's not here to say yes ;)
19:22:49 <marios_> SpamapS: i can ping him tomorrow my am
19:23:16 <greghaynes> O/
19:23:21 <SpamapS> greghaynes: want to do releases?
19:23:24 <greghaynes> Yep!
19:23:31 <marios_> cool
19:23:39 <SpamapS> marios_: want to show him how?
19:23:39 <greghaynes> also, apparently my calendar time is wrong
19:23:44 <greghaynes> SpamapS: done it before :)
19:23:46 <SpamapS> greghaynes: DST ;)
19:23:54 <SpamapS> greghaynes: ooohh ok then boom
19:24:01 <SpamapS> #info greghaynes to do releases
19:24:02 <marios_> greghaynes: wiki pretty good but feel free to ping me
19:24:05 <marios_> SpamapS: sure
19:24:25 <SpamapS> #topic CD Cloud status
19:24:35 <SpamapS> So I take this section to mean the various CI regions.
19:24:44 <SpamapS> but we don't CD them, so.. :-P
19:24:52 <SpamapS> tchaypo: anyway, HP2?
19:25:16 <SpamapS> he may have gone to dinner
19:25:22 <tchaypo> Sigh
19:25:48 <tchaypo> I broke the undercloud. Getting it going again. Would like to talk to you later
19:25:54 <SpamapS> Please do
19:26:07 <tchaypo> Right now the undercloud nodes can't get heat metadata
19:26:30 <SpamapS> tchaypo: ok, you're still in UTC+1 right?
19:26:40 <tchaypo> Yes
19:26:48 <SpamapS> tchaypo: ping me when you want to chat, don't want to keep you from dinner. :)
19:26:57 <SpamapS> Any other thoughts?
19:27:00 <tchaypo> Ok
19:27:42 <SpamapS> #topic CI
19:27:57 <SpamapS> #link http://goodsquishy.com/downloads/s_tripleo-jobs.html
19:28:27 <greghaynes> We were having an intermittent fail where nodes would go into deleting state... did that get fixed?
19:28:37 <greghaynes> Seems a lot more green
19:28:56 <SpamapS> Yeah I've been seeing passes.
19:29:21 <SpamapS> I would like to discuss the possibility of discussing testing the various update methods.
19:29:41 <greghaynes> meta
19:30:08 <SpamapS> I'm aware of three: 1) Try to get Heat to do rolling rebuilds (non-functional AFAIK) 2) tripleo-ansible, used by Helion product, doesn't work upstream, and 3) Update using package repos -- don't know where that lives.
19:30:58 <SpamapS> I feel like we need to test _at least_ updating from master to the tested commit.
19:31:12 <SpamapS> And we should also consider testing updating from stable to the tested commit once that works.
19:31:24 <SpamapS> Thoughts?
19:31:31 <slagle> if we add ci jobs that use the delorean packages, we could test some package updates
19:31:40 <SpamapS> +1
19:31:43 <TheJulia> Fundimentally it is a workflow, so the ansible code (of course I'm biased) seems like the simplest path to integrate
19:31:50 <greghaynes> well, do we have any update system that actually works upstream?
19:31:59 <greghaynes> seems like a prereq to testing an update system
19:32:06 <SpamapS> greghaynes: tripleo-ansible should be straight forward to fix for upstream.
19:32:10 <dprince> Yes, I'm using delorean locally so I'd love to have it in CI... and it is an upgrade option too
19:32:21 <slagle> greghaynes: the package updates work upstream
19:32:27 * dprince has some local patches to help use Delorean repos...
19:32:31 <greghaynes> ok, sweet, so theres two things to test :)
19:32:31 <slagle> it isn't really a "thing"
19:32:37 <slagle> you just run "yum update" :)
19:32:39 <TheJulia> SpamapS: we almost have it ready to rock and roll for upstream, maybe a few days tops
19:32:39 <SpamapS> slagle: they're not upstream in that they're not part of upstream's CI yet though.
19:33:07 <greghaynes> Another question is where should those live?
19:33:07 <SpamapS> ok so do we have capacity to run 2 more jobs per change?
19:33:15 <greghaynes> yea, that :)
19:33:16 <slagle> SpamapS: right, so yea, in general i agree, we should test upstream what the downstreams are doing
19:33:31 <SpamapS> One for delorean package updates, and one for tripleo-ansible?
19:33:34 <slagle> if those are the solutions that we're finding will work for us, we should represent that upstream
19:34:04 <greghaynes> derekh is usually the source of truth on our capacity...
19:34:16 <SpamapS> I feel like we shouldn't add tests until we have HP2 up
19:34:28 <dprince> greghaynes: if capacity is an issue we could combine it with one of the existing jobs too
19:34:54 <SpamapS> dprince: the problem with combining is that we need to do an install with _master_ not the tested change.
19:35:09 <SpamapS> But, we could definitely re-use master's images
19:35:19 <SpamapS> so that might be a fairly quick process.
19:35:31 <greghaynes> https://docs.google.com/spreadsheets/d/1LuK4FaG4TJFRwho7bcq6CcgY_7oaGnF-0E6kcK4QoQc/edit?usp=sharing is the list of CI jobs we made at the last mid cycle
19:35:34 <SpamapS> (that is, if we published master images) :)
19:37:05 <SpamapS> anyway, that is spec-worthy
19:37:08 <greghaynes> looks like a couple of the jobs are already marked 'update', IIRC this was the plan for those
19:37:27 <SpamapS> #link https://docs.google.com/spreadsheets/d/1LuK4FaG4TJFRwho7bcq6CcgY_7oaGnF-0E6kcK4QoQc/edit?usp=sharing
19:37:27 <dprince> As a first step even getting these new CI tests into tripleo-experimental might be a good check too
19:37:40 <SpamapS> dprince: true, that helps work out the kinks
19:38:06 <SpamapS> moving on?
19:38:21 <dprince> yes
19:38:28 <SpamapS> #topic Tuskar
19:38:31 <SpamapS> anybody here?
19:38:51 <SpamapS> maybe we can skip and come back if they o/
19:39:05 <SpamapS> #topic Specs
19:39:06 <slagle> i'm sitting in a meeting with jdob :)
19:39:12 <slagle> he says there's nothing
19:39:16 <SpamapS> slagle: excellent thanks
19:40:15 <SpamapS> I think tripleo-review-standards is ready for publish
19:41:02 <slagle> should we move all specs under juno/ that didn't actually make juno to a kilo/ dir?
19:41:31 <greghaynes> hrm
19:41:34 <SpamapS> slagle: that's an interesting question. Certainly we shouldn't approve anything still targetted at juno. :)
19:41:50 <SpamapS> But if they missed implementation, no I think we leave them there.
19:42:30 <SpamapS> https://review.openstack.org/#/c/97939/ , remove-mergepy, seems like one where some of it did happen, and some didn't, despite the fact it wasn't actually approved and published.
19:42:51 <SpamapS> So for that, I think we just approve it where it is, and track whats left as a kilo BP
19:43:04 <slagle> my configurable /mnt/state one got approved recently, but it was under juno/
19:43:23 <slagle> so maybe we just move that one to under kilo?
19:43:24 <SpamapS> slagle: yeah c'est la vie. Maybe we can put up a patch to rename?
19:43:29 <slagle> k :)
19:43:52 <SpamapS> also if I understand right, once a spec is approved, we need to create a BP yes?
19:43:59 <SpamapS> do we need a BP for tripleo-review-standards ?
19:44:27 <SpamapS> I don't think we do.
19:44:31 <SpamapS> As the spec itself is the documentation.
19:44:45 <slagle> i think it said in the spec, that there wouldn't be a bp
19:44:53 <SpamapS> right it does
19:45:12 <SpamapS> +A'd
19:45:55 <SpamapS> With that I think we can move on.
19:46:25 <SpamapS> #topic open discussion
19:46:51 <slagle> are there any more solid plans for a midcycle?
19:47:17 <SpamapS> Ah yes
19:47:18 <slagle> for budget planning questions, we are getting asked if there is a plan, etc.
19:47:29 <SpamapS> so we're waiting for a go-ahead from our office management.
19:47:48 <SpamapS> https://etherpad.openstack.org/p/tripleo-kilo-midcycle-meetup
19:47:53 <SpamapS> #link https://etherpad.openstack.org/p/tripleo-kilo-midcycle-meetup
19:48:08 <SpamapS> wait doh
19:48:12 <SpamapS> that must not be the one
19:48:29 <greghaynes> hrm? It is I think
19:48:31 <slagle> https://etherpad.openstack.org/p/kilo-tripleo-midcycle-meetup
19:48:37 <slagle> #link https://etherpad.openstack.org/p/kilo-tripleo-midcycle-meetup
19:48:40 <greghaynes> oh, right
19:48:54 <SpamapS> #link https://etherpad.openstack.org/p/kilo-tripleo-midcycle-meetup
19:49:13 <SpamapS> lag fail
19:51:09 <greghaynes> haha
19:51:40 <SpamapS> greghaynes: you sent out the original email, can you follow up on that same thread?
19:51:46 <greghaynes> Yep
19:51:52 <SpamapS> cool, I'll ping Omri
19:52:12 <SpamapS> slagle: we've had no other venue offers or suggestions, so Seattle seems most likely.
19:52:30 <SpamapS> Also I want to make something clear about the mid-cycle, from my perspective:
19:52:53 <SpamapS> They are optional opportunities to get together and hash out the problems and immediate plans for completing the kilo work.
19:52:58 <SpamapS> Nobody should feel like they _must_ be there.
19:54:45 <SpamapS> I think it's worth the time away from family, but I just want to make it clear: we need you happy and working hard all the days more than we need you at the mid-cycle days. :)
19:55:01 <SpamapS> anyway, that is all.. if nobody else has anything, I think we can adjourn.
19:56:04 * greghaynes waves
19:56:08 <marios_> night night tripleo
19:56:08 <SpamapS> alrighty then
19:56:10 <tchaypo> Bye
19:56:12 <SpamapS> #endmeeting