19:01:43 <jeblair> #startmeeting infra 19:01:44 <openstack> Meeting started Tue Jun 9 19:01:43 2015 UTC and is due to finish in 60 minutes. The chair is jeblair. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:45 <pleia2> o/ 19:01:45 * morganfainberg lets infra take the channel 19:01:46 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:01:48 <openstack> The meeting name has been set to 'infra' 19:01:49 <nibalizer> o/ 19:01:51 <Clint> o/ 19:01:52 <mrmartin> o/ 19:01:53 <crinkle> o/ 19:01:54 <asselin> o/ 19:01:55 <jeblair> #link agenda https://wiki.openstack.org/wiki/Meetings/InfraTeamMeeting#Agenda_for_next_meeting 19:01:55 <jeblair> #link previous meeting http://eavesdrop.openstack.org/meetings/infra/2015/infra.2015-06-02-19.01.html 19:01:57 <GheRivero> o/ 19:01:59 <yolanda> o/ 19:02:01 <dtantsur> o/ 19:02:06 <taron> o/ 19:02:23 <mordred> o/ 19:02:27 <jeblair> #topic Specs approval: Infra-cloud (jeblair, SpamapS) 19:02:35 <jeblair> #info infra cloud spec was approved 19:02:36 <jeblair> #link infra cloud spec http://specs.openstack.org/openstack-infra/infra-specs/specs/infra-cloud.html 19:02:46 <jeblair> the action on this now moves to the system documentation which is here: 19:02:50 <jeblair> #link infra cloud documentation https://review.openstack.org/180796 19:03:03 <jeblair> so people interested in how we actually deploy that should review that 19:03:10 <ruagair> o/ 19:03:36 <jeblair> #topic Specs approval: Host OpenStack Apps Catalog Service (docaedo, fungi) 19:03:41 <jeblair> #info apps.o.o spec was approved 19:03:41 <jeblair> #link apps.o.o spec http://specs.openstack.org/openstack-infra/infra-specs/specs/apps-site.html 19:03:54 <jeblair> that's two down! :) 19:04:00 <fungi> awesome. guess now i'm on the hook to finish that ;) 19:04:11 <ttx> o/ 19:04:37 <pabelanger> o/ 19:04:41 <jeblair> fungi: i have subscribed to your repo :) 19:04:47 <jeblair> #topic Specs approval: Shade (mordred) 19:04:48 <fungi> and my newsletter? 19:04:55 <jeblair> #link shade spec https://review.openstack.org/187312 19:05:19 <mordred> approve my spec everyone 19:05:24 <mordred> otherwise I'll be sad 19:05:26 <jeblair> this looks like it has buy-in from folks in this area of interest 19:05:38 <jeblair> anyone think it needs more time or shall we open voting on it? 19:06:14 <jeblair> #info voting on https://review.openstack.org/187312 open until 2015-06-11 19:00 UTC 19:06:21 <jeblair> #topic Specs approval: Puppet apply (mordred) 19:06:26 <jeblair> #link puppet apply spec https://review.openstack.org/165217 19:06:32 <mordred> approve my spec everyone 19:06:33 <mordred> otherwise I'll be sad 19:06:39 <jeblair> next up, a spec from mordred! 19:06:49 <anteaya> mordred: can you just add that to the commit message? 19:07:04 <mordred> anteaya: I do have that ability 19:07:11 <anteaya> you do 19:07:30 <jeblair> this one also seems ready to me; it's something we've talked about, and nearly implemented, for a long time 19:07:32 <mordred> as a note, the ansible puppet module has been merged into upstream ansible 19:07:40 <nibalizer> jeblair: agree 19:07:49 <jeblair> #info voting on https://review.openstack.org/165217 open until 2015-06-11 19:00 UTC 19:07:56 <jeblair> #topic Specs approval: Puppet 4 preparation and testing (mordred) 19:08:01 <mordred> approve my spec everyone 19:08:02 <jeblair> #link puppet4 spec https://review.openstack.org/#/c/175223 19:08:03 <mordred> otherwise I'll be sad 19:08:26 <anteaya> mordred specbot 19:08:44 <mordred> this is the last time I'll have 5 specs up for voting at the same time, I promise 19:08:51 <fungi> yikes. puppet 4 already? 19:08:54 <anteaya> I don't believe you 19:08:57 <dougwig> next time it'll be 10 19:08:57 <mordred> fungi: it's coming 19:09:01 <mordred> dougwig: ++ 19:09:01 <nibalizer> fungi: just setting up the testing 19:09:07 <nibalizer> we dont have to pull the trigger 19:09:20 <fungi> nibalizer: more just amazed that puppet 4 already exists 19:09:31 <mordred> fungi: semver solves everything 19:09:49 <nibalizer> your view is probably skewed a bit because we were on 3 for entirely too long 19:09:56 <nibalizer> jeblair: I think this spec is ready 19:10:05 <jeblair> s/were/have been/ :) 19:10:11 <jeblair> (which i guess is the point :) 19:10:14 <mordred> s/3/2/ 19:10:15 <jeblair> nibalizer: agree 19:10:15 <nibalizer> s/3/2/ 19:10:23 <anteaya> dougwig: that is more like it 19:10:33 <jeblair> #info voting on https://review.openstack.org/175223 open until 2015-06-11 19:00 UTC 19:10:36 <nibalizer> wooo 19:10:42 <jeblair> #topic Specs approval: Maniphest bug tracking (mordred) 19:10:44 <jeblair> #link maniphest spec https://review.openstack.org/188196 19:10:49 <mordred> approve my spec everyone 19:10:51 <mordred> otherwise I'll be sad 19:10:56 <pleia2> mordred is a spec machine 19:11:09 <fungi> also he has a great command of the up arrow in his irc client 19:11:17 <anteaya> ha ha ha 19:11:19 <mordred> fungi: I have an up arrow? 19:11:23 <jeblair> ttx: ^ you might want to take a look at this one 19:12:16 <jeblair> but i think this comes almost directly from the summit session, so i think it's probably ready for voting 19:12:19 <jeblair> anyone disagree ^? 19:12:33 <fungi> seems ready 19:12:53 <mordred> I made one informational change in the last revision in response to yolanda's question 19:13:16 <mordred> which is worth pointing out - which is the that I do not think we want to run both things - so we would be migrating everything in storyboard, not just infra projects 19:13:35 <jeblair> *nod* 19:13:45 <jeblair> there aren't too many of those 19:13:54 <jeblair> #info voting on https://review.openstack.org/188196 open until 2015-06-11 19:00 UTC 19:13:57 <jeblair> #topic Specs approval: Puppet Functional Testing (nibalizer, crinkle) 19:14:04 <jeblair> #link puppet functional testing spec https://review.openstack.org/178887 19:14:04 <crinkle> o/ 19:14:38 <jeblair> zaro does not seem to be here 19:15:04 <jeblair> he -1d it with questions, but i think they were asking for more information, not expressing significant disagreement or potential problems 19:15:12 <nibalizer> This is largely a summary of what was discussed at the summit, with cleanups from crinkle 19:15:20 <nibalizer> well more like a rewrite from crinkle 19:15:37 <crinkle> ah i can take a look at zaro's questions 19:16:33 <jeblair> nibalizer: answered them as comments; i don't actually think they need a new revision, so i'm inclined to say let's open voting as-is 19:16:34 <crinkle> looks like nibalizer answered them 19:16:47 <mordred> ++ 19:16:49 <nibalizer> jeblair: i agree, but I'm biased :) 19:17:03 <jeblair> #info voting on https://review.openstack.org/178887 open until 2015-06-11 19:00 UTC 19:17:07 <jeblair> #topic Specs approval: Host a code search service (taron, fungi, pleia2) 19:17:08 <nibalizer> woooo 19:17:11 <jeblair> #link code search spec https://review.openstack.org/188574 19:17:19 <fungi> if you don't approve my spec, i won't be sad. i was planning to spend this morning reviewing everyone else's open specs and then reality got in the way (again) 19:17:31 <taron> heh 19:17:50 <fungi> this is the project taron, our outreachy intern, is going to be working on this summer 19:17:55 <jeblair> fungi: but mordred might be sad 19:17:56 <mordred> this may be the first spec we've had in infra which suggests that someone write go 19:18:09 <fungi> a frightening thought indeed 19:18:15 * mordred opens the barn door to show everyone 27 sad goats 19:18:20 <pleia2> hah 19:18:41 * mordred closes the door again satisfied that people know the consequences 19:18:43 <fungi> mordred: fwiw, that go suggestion was courtesy of jeblair ;) 19:18:51 <jeblair> mordred's new programming language is goat 19:19:04 <mordred> jeblair: WHY IS THAT NOT A LANGUAGE????? 19:19:08 * fungi assumed this was just an "at" replacement written in go 19:19:09 <anteaya> it will show up everywhere 19:19:10 <taron> I was about to ask the same thing 19:19:12 <anteaya> eat the flowers 19:19:20 <anteaya> be in the neighbour's yard 19:19:26 <jeblair> #info voting on https://review.openstack.org/188574 open until 2015-06-11 19:00 UTC 19:19:41 <mordred> anteaya: eat the neighbour's yard 19:19:42 <jeblair> that was fun! 19:19:45 <jeblair> #topic Schedule Project Renames 19:19:46 <mordred> jeblair: whee! 19:19:47 <anteaya> mordred: that too 19:20:09 <jeblair> i think we mumbled something about doing this on friday as we were ejected from the room last week 19:20:20 <fungi> we have entirely too many of these 19:20:27 <fungi> but i agree friday would work for me 19:20:27 <jeblair> we have a metric boatload 19:20:32 <anteaya> well this friday worked better for me and fungi than last friday 19:20:32 <mordred> yah. and I still haven't tested the playlist :( 19:20:38 <mordred> playbook 19:20:39 <mordred> whatever 19:20:39 <pleia2> I think the puppet openstack stuff needs to move too, and it's not yet on the list 19:20:47 <crinkle> EmilienM: ^ 19:20:48 <mordred> yah. I beleive I saw a patch for that 19:20:48 <anteaya> playbill 19:20:48 <nibalizer> mordred: lets spin up a spotify playlist for these ops 19:20:55 <jeblair> mordred: do you think you might have something for us by then? 19:20:57 <fungi> i wonder where we could get a metric boat for these 19:21:14 <jeblair> mordred: or should we proceed as we have in the past? 19:21:21 <anteaya> fungi: not the states 19:21:24 <jeblair> mordred: and will you be around? 19:21:29 <mordred> jeblair: I mean, let's plan to proceed as we have in the past - but I'l also try to have something for us 19:21:30 <pleia2> I'm out on friday for family medical thing, but you can go on without me 19:21:33 <mordred> jeblair: I will be in berlin 19:21:38 <mordred> jeblair: I do not know if that counts as around 19:21:59 <ttx> jeblair: adding to my review queue 19:22:07 <fungi> mordred: i'm happy to put together yet another cut-n-paste batch for this round if the playbook is not ready for prime time 19:22:27 <anteaya> well this friday looks like just fungi and jeblair (you good for friday?) for roots 19:22:33 <anteaya> and a huge list 19:22:43 <mordred> I mean, I'll be around if we can do it earlier in the day 19:22:44 <anteaya> pleia2: is next friday better for you? 19:22:56 <mordred> if we do it later in the day, I'll be, you know, drunk - but could still be around 19:23:23 <pleia2> anteaya: yeah, I'll be around 19:23:33 <anteaya> pleia2: so the 19th is better for you? 19:23:38 <anteaya> mordred: the 19th? 19:23:58 <anteaya> I'm around on the 19th but not root so kind of useless 19:24:04 <jeblair> we normally do it around 2200 utc 19:24:13 <jeblair> mordred: what time would be best for you this friday? 19:24:24 <fungi> either of the next two fridays are wide open for me. i'll just be here fixing (or breaking) things 19:24:31 <mordred> anteaya: I will be around on the 19th 19:24:36 <jeblair> (i think 2200 utc counts as vaguely "late" in berlin) 19:24:44 <mordred> jeblair: I mean, I don't have a lot on my calendar for friday 19:24:48 <mordred> so I have no idea 19:25:04 <jeblair> i will be sprinting on friday the 19th 19:25:16 <mordred> problem is - the times that would be good for me are times when it's bad to have an outage 19:25:19 <jeblair> (project-team guide virtual sprint) 19:25:20 <anteaya> oh yeah, the sprint 19:25:26 <mordred> ahyes 19:25:47 <anteaya> okay well if it is just jeblair and fungi this friday I'm around to do what I can to help 19:25:48 <jeblair> i mean, we might decide to call it quits by 2200 on the 19th anyway 19:25:55 <ttx> jeblair: I plan to start doing things on the Thursday fwiw 19:26:34 <fungi> i think we've gotten the project renames sufficiently streamlined that we can manage it with just two people in a reasonable amount of time 19:26:39 <jeblair> fungi: yup 19:26:44 <ttx> but yeah, I definitely won't be around anymore at 2200 utc 19:26:50 <jeblair> ttx: good plan :) 19:27:18 <jeblair> let's go ahead and do this friday, the 12th, at 2200 and we'll see what happens with mordred and his playbook 19:27:31 <mordred> sounds like a children's book 19:27:31 <jeblair> (i think we would love for either or both to show up, as they are able) 19:27:44 <anteaya> mordred: like where the wild things are 19:27:53 <jeblair> fungi: sound good ^ ? 19:28:12 <fungi> wfm 19:28:18 <jeblair> #agreed project renames friday june 12 at 2200 utc 19:28:32 <jeblair> i'll send the announcement 19:28:34 <pleia2> I'll try to pop in if my husband is recovering ok, but no promises :) 19:28:42 <anteaya> pleia2: thanks 19:28:48 <jeblair> #action jeblair send announcement for project renames friday june 12 at 2200 utc 19:28:50 <anteaya> pleia2: hope things go well for him 19:28:58 <pleia2> anteaya: thanks 19:29:19 <jeblair> #topic Priority Efforts (Downstream Puppet) 19:29:48 <asselin> hi, I sent an e-mail about the proposed virtual sprint for common-ci 19:30:01 <yolanda> i started work on nodepool.conf either on puppet module, and in nodepool itself 19:30:03 <jeblair> #info Virtual Sprint scheduled: Wednesday July 8 1600 UTC Sprint will run for 48 hours 19:30:11 <nibalizer> woot 19:30:12 <jeblair> #link https://wiki.openstack.org/wiki/VirtualSprints#OpenStack_Common-CI_Solution 19:30:15 <asselin> no objections stated, so we should be good witht he date time agreed in the last meeting 19:30:31 <pleia2> turns out I will be around for that (other engagement has been postponed), anteaya you too? 19:30:31 <nibalizer> beaker-rspec tests are comming 19:30:45 <anteaya> pleia2: yes I expect to be available 19:30:49 <nibalizer> they're almost in place for openstackci and then I will fungi them out to all the other modules 19:30:56 <jeblair> nibalizer: which will land first, the spec or the tests? :) 19:30:59 <anteaya> yes do fungi them out 19:31:05 * fungi has just been mordreding, fwiw 19:31:18 <anteaya> fungi: eating the neighbour's yards? 19:31:24 <jeblair> anteaya: drinking 19:31:30 <anteaya> ha ha ha 19:31:31 <nibalizer> jeblair: dunno 19:31:31 * mordred hands fungi a mostly unused implement 19:31:59 <fungi> well, i meant following in the tradition of scripting mass change batches of mostly trivial/repetitive stuff 19:32:05 <nibalizer> we talked a bit with pabelanger and pleia2, testing will focus on infra's use case i.e. ubuntu but our friends who use centos are welcome to add support and testing to the modules for that 19:32:13 <fungi> but drinking too, yes 19:32:27 <cody-somerville> :] 19:32:43 <jeblair> nibalizer: sounds good 19:32:51 <jeblair> tests help make that kind of support viable 19:33:06 <jeblair> (it has not been in the past, we will just instabreak things) 19:33:13 <pabelanger> nibalizer, jeblair Yup, I can help out where ever possible. 19:33:33 <zaro> o/ 19:33:52 <nibalizer> thats all I got 19:33:58 <jeblair> cool, thanks 19:34:00 <nibalizer> excited for the sprint 19:34:05 <jeblair> #topic Open discussion 19:34:08 <yolanda> jeblair, saw you unlocked a pair of my changes, cool 19:34:19 <yolanda> i just did a quick rebase but i need to look carefully 19:34:27 <jeblair> yolanda: yeah, the blocking change merged a long time ago, i just missed those, sorry 19:34:44 <yolanda> no problem 19:35:19 <pabelanger> I wanted to know if creating a spec for grafana was worth while. Outside the CI dashboard spec that is going on now 19:35:50 <asselin> ci dashboard spec might get abandoned.... 19:36:00 <pabelanger> since a few people, nibalizer, timrc are doing stuff downstream with it 19:36:10 <pabelanger> asselin, I didn't know that 19:36:32 <timrc> What would the grafana spec include? 19:36:38 <nibalizer> pabelanger: eh, i think getting a grafana up and having it driven by git might be enough 19:36:55 <timrc> One of the things I really want are more data points. I think we need a spec for data gathering and analysis in general :) 19:37:04 <jeblair> pabelanger: i think with the current scope of "run grafana, use grafyaml to translate static yaml files to grafana configs" probably doesn't need a spec unless you want to. if it gets any more complicated than that, we probably should. 19:37:04 <anteaya> pabelanger: look at the logs for monday's third party meeting 19:37:11 <asselin> pabelanger, take a look at the latest comments: https://review.openstack.org/#/c/135170/ 19:37:17 <anteaya> pabelanger: it provides some context 19:37:26 <pabelanger> timrc, mostly just importing it into -infra, a server that runs it and allows people to setup graph points 19:38:07 <pabelanger> jeblair, Ya, that is what I was mostly asking. If a spec was required, then I'd do it. But if we just wanted to stand up a server and run grafyaml, then that is good too 19:38:11 <asselin> an alternative working solution is being proposed. Perhaps the spec will be to just use this: http://ec2-54-67-102-119.us-west-1.compute.amazonaws.com:5000/?project=openstack%2Fcinder&user=&timeframe=24 19:38:31 <timrc> I'm ambivalent on a spec. A spec would maybe open the floor for more ideas from other people that maybe haven't listened in on the conversations we've had on #openstack-infra. 19:38:44 <asselin> pabelanger, currently hosted here: https://git.openstack.org/cgit/stackforge/third-party-ci-tools/ 19:38:52 <fungi> pabelanger: i think we've been leaning toward specs for any implementation requiring new servers, but stuff running on existing servers without substantial new complexity seems fair game for no spec 19:39:34 <pabelanger> okay, so no spec. I'm cool with that. I'll focus on the puppet-grafana module this week and ping people for reviews 19:39:47 <pabelanger> while will be timrc and nibalizer :) 19:39:53 <pabelanger> which* 19:40:02 <nibalizer> :P 19:40:05 <timrc> pabelanger, Awesome. We already have grafana deployed in HP so if you need any assistance there, I'd be happy to help. 19:40:23 <timrc> pabelanger, Also did you solve automating datasources and orgs? 19:40:40 <pabelanger> The only other thing I had was around SteveK's packaging work. Anybody work local with him, or contact. I've tried pinging him a few times, but nothing back 19:41:01 <pabelanger> timrc, nothing official yet, that's the last piece for me 19:41:14 <pleia2> pabelanger: he's in .au, so ping in US night are best 19:41:41 <timrc> pabelanger, I'll take this offline... there are some other gotchas I stumbled across when deploying this into HP :) 19:41:48 <pabelanger> pleia2, okay, perhaps a ML post is better. 19:41:50 <pleia2> pabelanger: he also does respond to email, which I resort to sometimes when I don't want to stay late 19:41:59 * pleia2 nods 19:42:03 <pabelanger> was mostly looking for an update from him, to help get the infra packaging going again 19:42:13 <pabelanger> since I have some spare cycles to work on it 19:42:21 <pabelanger> timrc, roger 19:42:30 <pleia2> cool 19:42:52 <pabelanger> That is all from me 19:43:54 <jeblair> anteaya, asselin: whatever comes out of the third-party ci dashboard area, we should have a spec for it; even if it's "run patrick's thing" 19:44:05 <anteaya> jeblair: understood 19:44:14 <asselin> jeblair, ok 19:44:14 <anteaya> asselin: are we able to create that spec yet? 19:44:23 <anteaya> asselin: I can work on it with you if you wish 19:44:35 <asselin> anteaya, ok let's discuss offline 19:44:41 <anteaya> asselin: okay 19:45:26 <jeblair> pabelanger, mordred: is there a packaging spec yet? 19:45:39 <jeblair> ah 19:45:40 <jeblair> https://review.openstack.org/179713 19:45:55 <mordred> yah 19:46:07 <pabelanger> ya 19:46:13 <pabelanger> that is the one I was referring too 19:46:16 <jeblair> ok, so that's where the conversation is 19:46:46 <mordred> I should respond to zigo 19:46:51 <jeblair> anyone have anything els? 19:47:14 <mordred> I think there is a difference between "to bypass distros" and "that need to exist sometimes because of the operational need to have a package that is not currently in the distros" 19:47:25 <mordred> it's a difference of intent 19:47:28 <mordred> not of substance 19:47:35 <jeblair> yeah 'bypass' doesn't really cover that subtlety 19:48:32 <jeblair> okay, we can let the tc get started early! 19:48:36 <jeblair> thanks everyone! 19:48:38 <zaro> mordred: iirc you said you were working on spec for testing gerrit upgrades? 19:50:10 <jeblair> to be continued in infra channel ^ 19:50:13 <jeblair> #endmeeting