15:00:03 #startmeeting puppet-openstack 15:00:04 Meeting started Tue Feb 23 15:00:03 2016 UTC and is due to finish in 60 minutes. The chair is EmilienM. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:09 The meeting name has been set to 'puppet_openstack' 15:00:13 #link agenda https://etherpad.openstack.org/p/puppet-openstack-weekly-meeting-20160223 15:00:15 o/ 15:00:29 hi 15:00:36 o/ 15:00:49 good morning 15:00:54 #topic Review past action items 15:00:57 o/ 15:01:03 EmilienM to push a mitaka beta tag: need to be postponed due to CI issues 15:01:22 #action EmilienM to push a mitaka beta tag once our CI is stable 15:01:29 mfisch and chem to initiate a blueprint about admin_token_auth deprecation 15:01:34 mfisch, chhavi: postponed? 15:01:37 yeah 15:01:42 no action, probably none this week 15:01:44 <_ody> o/ 15:01:51 ok 15:01:56 <_ody> lurking for a bit. Need to get kid ready for school. 15:02:08 #action mfisch & chem postpone blueprint about admin_token_auth deprecation 15:02:30 #topic mailing list 15:02:32 mfisch: o/ 15:02:44 yo 15:02:56 ok im up 15:03:08 hello 15:03:13 first is mailing list... its possible to subscribe to openstack-dev and just to the [puppet] tag 15:03:22 and I'd like everyone to do so, especially if you are core 15:03:29 or if you want to become 15:03:44 when you subscribe you can choose the topics you want 15:03:50 but yeah, it's a basic thing to subscribe & read e-mails if you want to get involved 15:04:07 did the old mailing list get removed entirely? If not might be worth posting that there also 15:04:26 clayton: nobody is using it for long time, I guess we can close it 15:04:34 yes lets close iut 15:04:38 I'm not admin though, we might need to ask Hunner or _ody 15:04:44 action? 15:04:50 should probably post an update about how to subscribe to openstack-dev tags there before closing it though 15:04:58 #action _ody / Hunner to close puppetlabs-openstack mailing list 15:04:58 so people wondering why it's so quiet know what they should do 15:05:06 its been almost a year since we used it 15:05:10 * _ody will make it happen 15:05:29 right, _ody can send a last email about that on this ML 15:05:31 #action _ody to close old ML and tell people to subscribe to the new one 15:05:59 mfisch: thx for pointing that out 15:06:17 the excuse "oh e-mails... so much traffic" does not work anymore ;-) 15:06:41 #topic keystone::python deprecation 15:06:43 mfisch: again ! 15:06:47 yay me 15:06:47 ok 15:06:50 we have some reviews 15:06:52 I think it's important to give people good instructions on how to subscribe to the tag vs just tell people to do it 15:06:59 it's not obvious 15:07:16 thank you 15:07:19 clayton: right, does OpenStack provide an official guide? 15:07:33 there is https://wiki.openstack.org/wiki/Mailing_Lists 15:07:37 not sure, when I looked a year ago I couldn't find one, but that doesn't mean it didn't exist 15:07:47 I found an explanation from fungi I think that said how to do it 15:07:55 crinkle explained it to me once 15:08:03 but once I found it it, its easy 15:08:24 I think https://wiki.openstack.org/wiki/Puppet/Mailing_lists is clear 15:08:25 #undo 15:08:27 Removing item from minutes: 15:08:47 specially https://wiki.openstack.org/wiki/Mailing_Lists#Future_Development 15:09:01 that explains it 15:09:06 it's well explained here, I don't where is trouble. 15:09:13 bkero: did you read it? 15:09:20 I did 15:09:23 cool 15:09:30 do we have more questions about ML? 15:09:36 thanks for finding that link 15:10:04 cool 15:10:10 so keystone::python deprecation 15:10:14 ok 15:10:37 so I found an ancient class called keystone::python and all it does is install keystoneclient 15:10:42 just like keystone::client 15:10:44 we have 4 commits: https://review.openstack.org/#/c/282065 - https://review.openstack.org/283376 - https://review.openstack.org/283377 and https://review.openstack.org/283378 15:10:56 for some reason things like glance and neutron are pulling it in 15:11:00 references from like 2013 15:11:10 it's part of legacy code iirc 15:11:13 so I'd like to make this class do nothing except warn and remove all the warnings 15:11:18 then remove it completely in N 15:11:31 I dont see any reason why glance::api needs to install the keystoneclient 15:11:34 (for example) 15:11:38 agreed 15:11:54 right 15:12:05 it's up to packaging dependencies to solve this problem, I think 15:12:09 so we can discuss more in the reviews 15:12:28 thats it for that topic 15:12:33 they are all lgtm, except the puppet-keystone 15:12:38 you +2d it? 15:12:44 no, I -1'ed 15:12:45 ah denis did 15:12:53 yep I see I will fix 15:12:56 you're right 15:13:11 we have a last topic in the agenda 15:13:15 me agin! 15:13:18 #topic r10k path and testing 15:13:21 mfisch: hello ! 15:13:28 he's hot 15:13:41 I brought this up on the ML but it's been frustrating that I can't run unit tests easily since r10k is not installed properly 15:14:00 I'd been just pushing rveiews and letting jenkins find issues which is a waste of time 15:14:00 that's something we need to adress, indeed 15:14:20 I dont honestly know much about how those deps are installed 15:14:25 isnt r10k listed as a requirement? 15:14:26 it's weird you're the only one to complain until now 15:14:30 well, it's not in your path, it is installed properly 15:14:36 I've had the same issue. 15:14:46 its not that it its not in my path 15:14:47 it happens if you set GEM_HOME before running bundler 15:14:52 its not in the path the code wants it to be in 15:15:01 usually /foo/1.9.1/bin 15:15:18 I run a script to execute testing: https://gist.github.com/EmilienM/61d590a89c44364d0436 15:15:20 I don't think it happens if you use gemsets or some such 15:15:21 all I run is bundle install && bundle exec rake spec 15:15:43 EmilienM: do you have r10k installed on the test system? 15:15:52 bottom line it should be easy and work out of the box, I dont know enough about gems to discuss more intelligently than that 15:16:08 clayton: no, it installs it everytime in my temp dir 15:16:19 mfisch: you're right, it should work out of the box 15:16:28 I'm not sure how it's finding the r10k binary then, since I think all the tooling assumes its in your path 15:16:42 Is there a more sane way? 15:16:45 the tooling wants it to be somewhere pretty specific 15:16:56 it's not just running "r10k" its specifying a full path from what I saw 15:17:44 maybe I'm misremembering then. I did have the same issue a while back 15:17:57 I thought I added it to my path temporarily as a work around 15:18:18 clayton: I ended up symlinking the installed copy from /usr/local/bin into where it was looking 15:18:22 should we just file a bug for this and move on? 15:18:31 can we create a bug with all of the details like steps to reproduce? some of the problem is that it doesn't happen to everyone 15:18:32 yes, please file a bug 15:18:43 seems like an environment thing 15:18:48 which project? 15:19:23 the one you like 15:19:26 it affects all modules 15:19:35 let's start with keystone module 15:19:38 ok 15:19:52 thanks 15:20:06 #action mfisch to file bug on r10k issue during spec tests 15:20:06 #action mfisch to create a bug about r10k path & testing 15:20:09 #undo 15:20:09 Removing item from minutes: 15:20:14 anything else? 15:20:23 no, thanks 15:20:27 #topic open discussion 15:21:28 if you have outstanding bugs / patches, go ahead 15:21:34 mwhahaha: where are we with puppet-glance CI? 15:21:58 not sure i haven't had a chance to dig into https://bugs.launchpad.net/puppet-glance/+bug/1548515 much 15:21:58 Launchpad bug 1548515 in puppet-glance "glance acceptance failing after removal of keystone auth token" [Critical,Confirmed] 15:22:33 i'm going to try and take a look later today if i get time but for now glance is broke 15:23:09 mwhahaha: I'll look too 15:23:39 I just got some inputs from jamespage on #openstack-dev about updates in Ubuntu that will likely break puppet-neutron 15:23:51 specially in config file management & also packaging renames... 15:24:20 anything else for today? 15:24:39 short & effective. Have a nice week all 15:24:45 Thanks guys 15:24:46 #endmeeting