20:00:46 <j^2> #startmeeting openstack-chef 20:00:47 <openstack> Meeting started Thu Aug 14 20:00:46 2014 UTC and is due to finish in 60 minutes. The chair is j^2. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:00:50 <openstack> The meeting name has been set to 'openstack_chef' 20:00:54 <j^2> Hey everyone! 20:01:05 <j^2> Here’s our Thursday IRC meeting 20:01:48 <j^2> Anyone around to check in and say hi? 20:02:07 <markvan> j^2: Howdy 20:02:25 <j^2> #topic General Discussion 20:02:41 <j^2> Is there anything specific anyone would like to bring up? 20:02:54 <j^2> I have 3 quick topics, but i’m opening it up to everyone frist 20:03:58 <carlp> I don't have anything 20:04:03 <carlp> so, what do you have? 20:04:09 <j^2> heh, that works :P 20:04:32 <j^2> #topic moving IRC meeting to friday 20:05:05 <markvan> friday morn is fine with me 20:05:28 <j^2> So Ionuț would like to have the IRC meeting moved to friday 20:06:23 <carlp> ok 20:08:05 <j^2> cool, so any other concerns or questions? obviously it’ll start next week not tomorrow beause that’s just redundant 20:09:08 <j^2> ok, cool, next topic 20:09:11 <mattray> don't conflict with the #chef-hacking meeting 20:09:36 <j^2> yeah we shouldn’t, it’s an hour before ours to the #chef-hacking :) 20:09:52 <j^2> #topic the gate jobs for both master and stable 20:10:08 <j^2> markvan: you have made some movement? issues? can you explain it for the group? 20:11:22 <j^2> #info mattray suggests we don’t conflict with the #chef-hacking meeting 20:12:33 <markvan> yup, the gate issue was merged just a bit ago. We are trying to setup the gate to allow for Berks 3.x to be used in Master, but 3.x deprecated some of the 2.x commands, so we needed to put in a check. 20:13:10 <j^2> is this something we can do programaticly per branch or per repo? where does this fix exactl? 20:13:12 <markvan> My first attempt failed, but with jklare's help, we're now back on track 20:14:02 <markvan> So, in working on this, I noticed our current gates are setup to cover all our branches, which means any change could break the stable branches. 20:14:12 <j^2> agreed 20:14:37 <j^2> #info our current gates cover all the branches which could break stable branches 20:14:37 <markvan> I asked out on infra, and it seems like we can have separate gate jobs for Master and then for Stables. 20:14:52 <j^2> oh nice 20:15:08 <markvan> I'll take a look into some examples and put together a blueprint to cover that change. 20:15:24 <j^2> a spec first right? ;) 20:15:39 <markvan> The other issue with messing with gates is that it's hard if not impossilbe to test it completely before pushing up a patch. 20:15:45 <markvan> yup spec 20:16:08 <j^2> #action markvan will write up a spec to split the branches gating jumps into master and stable 20:16:21 <markvan> we had our first patches merged in over a week....yah! 20:16:56 <j^2> btw, the meeting commands are here: https://wiki.debian.org/MeetBot if anyone else wants to leverage them 20:17:11 <j^2> markvan: nice! 20:17:24 <markvan> And i would like to change the gate job for bundler to add the --retry option to allow for temp outages with installing gems. 20:17:41 <j^2> that seems like a great idea 20:18:22 <markvan> And with this latest gate update, jklare can continue to move toward getting our cookbooks up to using Berks 3.x. 20:20:08 <j^2> awesome, that’s great news 20:20:25 <openstackgerrit> Mark Vanderwiel proposed a change to stackforge/cookbook-openstack-dashboard: Update horizon local settings conf files for Juno https://review.openstack.org/112050 20:20:42 <markvan> I think that the summary from gate world 20:21:32 <j^2> awesome works for me 20:21:54 <openstackgerrit> Mark Vanderwiel proposed a change to stackforge/cookbook-openstack-dashboard: Need to be able to enable firewall and vpn in dashboard https://review.openstack.org/108493 20:21:57 <j^2> #topic General Discussion/General Questions or Concerns 20:22:02 <carlp> Yes, great work Mark! 20:22:29 <j^2> I’m opening the floor up to anyone or anything we would like to discuss, please don’t hesitate 20:23:00 <markvan> I have several Juno pathch out there that are trying to get the cookbook up to speed with latest Juno conf file changes. My intent was to get them in early in the cycle to set the stage for Juno. 20:23:37 <carlp> I'll try and spend some time reviewing it this weekend 20:23:38 <j^2> makes sense, this is to fix most/all of the conf file changes? or just get a “worknig viable product?" 20:23:47 <carlp> I'm swapmed until then :( 20:25:07 <markvan> Was trying to cover all the conf changes I could find. In most cases the actual code changes are small, BUT I did try to sync with the latest conf file comments to make it easer to inspect nodes after converges. 20:25:40 <j^2> are you creating specs for the specific changes too? 20:26:03 <j^2> specs in the sense of tests 20:26:11 <carlp> A suggestion: If the options don't exist in a config file, defaults are taken. Maybe we should just strip out all the options we don't specifically set and go from there? 20:26:13 <markvan> Just straight up bugs for these initial changes....since there's very little logic changing. 20:26:14 <j^2> i don’t like how we have the word specs mean too things 20:26:37 <carlp> I know it means not seeing all the options in the files, but we could also move the default files to .orig or something... 20:27:30 <markvan> carlp: the new issue is that some of the base project now don't include sample conf files, so nothing to look at after an install. 20:28:00 <markvan> So, I did the gen sample conf and went from there, since that fairly easy to do once a release 20:28:14 <carlp> markvan: right, but putting a .orig or .sample file next to the generated one should be a lot less eork - no? 20:29:05 <j^2> and also that’ll give the ability to diff against them too if all the sudden something popped up 20:29:07 <carlp> That way we don't have to touch it at all, just update the .sample once a release and put a note in the generated one to check there for all the options 20:29:57 <markvan> k, I see your point. 20:30:33 <carlp> I'm not saying we have to, just a thought... 20:30:36 <markvan> I guess that would only really apply to the major conf files, as most of the others are fairly small and don't change much. 20:30:50 <carlp> yeah, agreed 20:31:07 <carlp> Hell, nova.conf has over 600 options in it. I bet we change like 20 20:31:34 <markvan> yeah, and then they deprecate and/or move them to other sections 20:31:56 <j^2> #info update the .sample once a release and put a note in the generated one to check there for all the options, that would only really apply to the major conf files, as most of the others are fairly small and don't change much. 20:32:49 <markvan> probably need a spec on this idea, to allow others to chime in. But I think it's a good item to try to tackle in the future. 20:32:53 <xmltok> so it looks like launchpad has my ICLA now 20:33:19 <carlp> xmltok: yay! 20:33:29 <j^2> xmltok: nice! 20:33:55 <xmltok> yes, nice, it shows stuff signed from back in august so i guess it worked back then it was just crashing when rendering it back to me 20:34:16 <j^2> awesome, that’s great news 20:34:19 <markvan> btw, I have been working on forking that conf sample generator to instead generate Chef attributes and templates, that way we could have every option gen'd automatically (with a few exception.) 20:35:13 <markvan> At least that might help with trying to keep in sync with the base. 20:37:12 <markvan> New topic? The Versioning spec? We should try to get that initial one out and make mods as needed later. 20:37:14 <j^2> markvan: wouldnt that have some overhead to it? i don’t think that has a long term advantage. greated attributes aren’t much better i think it would cause some confustion? 20:38:48 <j^2> granted a sample generator should be a much larger conversation 20:38:49 <markvan> j^2: Yeah, still working out the details, but would be nice to be able to branch to "k" release and know ahead of time what conf values changed. 20:39:03 <j^2> maybe a spec ;) 20:39:36 <markvan> yup, but seems like something to explore to see if it could help us out as the base grows. 20:40:21 <j^2> sounds good 20:40:33 <j^2> #topic Versioning spec 20:40:49 <markvan> The other issue is that Chef templates don't handle [sections] very well right now. but that's another story. 20:40:57 <j^2> ok so we’ve attacked this spec every which way 20:41:25 <markvan> For Versioning spec, I took a shot with a bunch of updates....would like some feedback if the general idea is sound. 20:41:55 <j^2> https://review.openstack.org/#/c/110763/ is still the review? 20:42:39 <markvan> https://review.openstack.org/#/c/110728/ is the Versioning one. 20:42:45 <j^2> ah opps 20:42:48 <j^2> i’m glad i asked 20:43:11 <j^2> ha! i got my integration one conused 20:43:47 <j^2> #info https://review.openstack.org/#/c/110728/ is the Versioning spec review 20:44:32 <openstackgerrit> Mark Vanderwiel proposed a change to stackforge/cookbook-openstack-dashboard: Need to be able to enable firewall and vpn in dashboard https://review.openstack.org/108493 20:45:04 <markvan> The net of the spec is, lock down Stable branches with locks and versions for every change, and for Master, allow some minor changes to just require a ChangeLog entry. 20:45:31 <j^2> markvan: you’ve made some great changes with this. I need to read it deeper but man great job 20:45:35 <j^2> !m markvan 20:45:35 <[d__d]> You're doing good work, markvan! 20:45:35 <openstack> j^2: Error: "m" is not a valid command. 20:46:01 <j^2> i should convert that ! m command to os-chef-bot 20:46:24 <j^2> oh on that note, os-chef-bot now has this 20:46:26 <j^2> os-chef-bot: help 20:46:26 <os-chef-bot> please click here: http://tranquil-anchorage-9510.herokuapp.com/os-chef-bot/help 20:46:36 <j^2> you can click on it and recieve the commands he can do 20:46:53 <markvan> nice, I forget the cmds all the time.... 20:47:00 <j^2> also, being that we have the openstack bot in this room, we are being logged in two locations 20:47:09 <j^2> the offical openstack one, and botbot.me 20:47:29 <j^2> http://eavesdrop.openstack.org/irclogs/%23openstack-chef/ 20:47:36 <markvan> cool 20:47:41 <j^2> there is some redundancy to this, so i’d like to have a discussion about it 20:47:49 <j^2> the offical openstack is well openstack 20:48:08 <j^2> botbot.me is more the “opensource” channel logger, and has searching 20:48:13 <j^2> openstack doesnt 20:48:21 <j^2> i’m not sure _what_ is the correct answer 20:48:43 <j^2> though theres no harm in having them both 20:48:51 <j^2> i’d just ilke to through it out there to see what people thing 20:48:53 <j^2> think* 20:49:38 <markvan> I probably need reivew the output from each...then I can give better feedback. both for now sounds like no issue 20:49:52 <j^2> agreed 20:49:59 <j^2> but i wanted to at least have it on the radar 20:50:25 <markvan> yup, as we improve our meetings...this will become more important 20:50:34 <j^2> that’s the goal :) 20:50:55 <j^2> i have to admit i am starting to dig the irc style 20:51:14 <j^2> we are all here anyway 20:51:29 <markvan> yeah, it's not to bad...if I coudl only type better 20:51:49 <markvan> next topic, the VPN spec https://review.openstack.org/#/c/111145/ 20:52:10 <j^2> #topic VPN spec 20:52:55 <markvan> This VPN spec has had some good comments and been updated to a point where I think it's pretty good. 20:53:16 <markvan> Just trying to make other aware of that if they're interested. 20:53:47 <markvan> #info VPN spec https://review.openstack.org/#/c/111145/ Please review if interested 20:55:13 <j^2> awesome 20:55:30 <j^2> it seems like a great idea, but man, SSL :-/ i have trouble with basic networking :’( 20:56:14 <markvan> yeah, luckily it looks like this spec will be fairly straight forward recipe coding 20:57:09 <j^2> back to the conversation about integration testing :P 20:57:17 <markvan> sure 20:57:23 <j^2> #topic integration testing 20:57:53 <j^2> any thoughts on this? i personally have a POC when that glance issue is resolved 20:59:08 <j^2> galstrom_zzz: gave me a couple people to ping about the rackspace test suites, honestly i havent yet 20:59:41 <j^2> #action j^2 should ping the RS guys about the slaved jenkins machines 20:59:42 <markvan> I would really like to see this happen. Whatever I can to help, just ask 20:59:54 <j^2> fix glance ;) 21:00:20 <markvan> yeah, I'm have it on my list to ping the core for that and ask what's going on. 21:01:23 <markvan> Seems like Nova added the auto covert to utf8 long ago...why with that exmaple should it take the other projects so long 21:08:00 <xmltok> oh you know what would be nice from a consumer of the cookbooks? tagged version numbers in the git repos 21:08:20 <j^2> cool everyone, thanks for being for this hour, again, we are going to move this to friday’s. I’ll send out some data to the ML 21:08:23 <j^2> #endmeeting