16:00:16 <jklare> #startmeeting openstack-chef
16:00:17 <openstack> Meeting started Mon Nov 16 16:00:16 2015 UTC and is due to finish in 60 minutes.  The chair is jklare. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:21 <openstack> The meeting name has been set to 'openstack_chef'
16:00:32 <jklare> hi everyone
16:02:33 <jklare> i will wait for 10min for people to join and close the meeting after that if nobody does
16:09:19 <sc`> o/
16:09:35 <jklare> hi sc`
16:09:44 <sc`> howdy
16:09:51 <jklare> anything you want to talk about?
16:10:19 <sc`> if j^2 were around, perhaps his alternate packaging idea
16:10:29 <j^2> yep, i’m here
16:10:34 <jklare> cool
16:10:35 <j^2> sorry, caught up in a meeting
16:10:41 <jklare> no worries
16:10:48 <sc`> oh right. forgot the double duty D:
16:10:57 <j^2> :(
16:11:13 <j^2> do we have an agenda for this meeting?
16:11:17 <j^2> i forget
16:11:21 <jklare> nope
16:11:24 <j^2> ah
16:11:43 <j^2> ok, well….
16:11:58 <jklare> i think we do not need an agenda for 1-4 people :)
16:12:10 <j^2> yeah i guess that makes sense
16:12:29 <jklare> so, sc` said you had an alternative packaging idea j^2 ?
16:12:38 <j^2> yeah giftwrap
16:12:54 <j^2> https://github.com/blueboxgroup/giftwrap/blob/master/README.md
16:13:01 <sc`> i was looking at that a bit. haven't played with it yet
16:13:17 <j^2> basically, that was the conversation i was having with the bluebox guys when we were in Dev Lounge jklare
16:13:43 <j^2> it’ll uncouple us from upstream pkgs so we can just “configure” the code
16:13:59 <jklare> you are talking about giftwrap?
16:14:03 <j^2> yep
16:14:07 <jklare> https://github.com/blueboxgroup/giftwrap
16:14:14 <j^2> yep the readme is above
16:14:29 <jklare> ah, did not see that
16:14:51 <jklare> i should read before writing....
16:14:52 <jklare> sry
16:14:56 <j^2> ha!
16:15:17 <jklare> so, yeah i think this looks nice, but for now i have no idea how to properly make use of that inside of our cookbooks
16:15:17 <j^2> so this _could_ be an answer for our pkging problems
16:15:30 <j^2> yeah same here, but i wanted to see what y’all thought before going any deeper
16:16:05 <jklare> if we want to decouple from upstream packaging, we should write this down into a spec  i think
16:16:13 <sc`> all we do is point at the repo endpoints
16:16:25 <jklare> right
16:16:29 <sc`> how the packages get there don't matter to us at this point
16:16:43 <sc`> but yeah, should go in a spec
16:16:45 <jklare> that means if we want to do that, we need to host these packages somewhere after building them
16:16:49 <j^2> jklare: totally,  but before I run with that i wanted to take any major issues i haven’t thought of
16:17:11 <j^2> jklare: i don’t think so, i think the pkgs might be built on the fly?
16:17:20 <j^2> but i’m not 100% sure
16:17:29 <sc`> it's infrastructure we have to maintain and support
16:18:07 <jklare> if we build the packages on the fly, we need to install docker, run the build, get the package, uninstall docker and then install the package
16:18:32 <jklare> i think thats not a very good option
16:18:35 <j^2> true
16:18:40 <sc`> if we build the packages on the fly, where will this be done? would we need to get infra resources?
16:18:54 <j^2> yeah i’m not sure, i haven’t ran through the whole thing yet
16:18:58 <sc`> seems like a heavyweight operation
16:19:08 <j^2> there has to be a place we can host the pkgs though, there has to be
16:19:21 <jklare> giftwrap basically starts docker containers to automate the build process of packages
16:19:31 <j^2> it doesnt have to use docker
16:19:33 <sc`> good for folks like us ($job) but maybe not so much unless we're able to get some kind of resources to throw at this
16:19:40 <jklare> the great thing about giftwrap is the why how its figures out the py dependencies
16:19:48 <j^2> yep, that’s the gist
16:22:46 <jklare> and now for something completely different: we cut stable/liberty last week i think and merged most of the necessary repos
16:23:05 <jklare> :)
16:23:29 <sc`> a great turnaround on shipping liberty after how long it took to get kilo out the door
16:23:36 <jklare> i pushed up the bash i used, since i hated to do the steps described in our wiki manually
16:23:46 <jklare> https://wiki.openstack.org/wiki/OpenstackChefStablebranchCreateNotes#Steps_to_create_stable.2F_branch
16:24:05 <j^2> ROCK ON man, that’s bad ass
16:24:11 <j^2> yeah doing it by hand is bullshit
16:24:22 <jklare> :)
16:24:51 <sc`> jklare: that's awesome. makes things much easier
16:26:07 <jklare> my collegue calbers started the work on mitaka with refactoring the compute cookbook and i think he has nearly finished step 1) removing all the default attributes
16:26:23 <jklare> sadly he is on holiday this week, so we will not see the push before next week
16:26:48 <jklare> i wanted to start looking on the networking cookbook tomorrow and maybe i can push step 1 and 2 this week
16:27:16 <jklare> depending on the amount of 'end of the year' stuff that pops up in my calendar
16:27:18 <jklare> :)
16:28:32 <j^2> :D
16:28:35 <jklare> i also -2ed 2 patches this morning that wanted to add some more configuration attributes by commenting on them, that this can go into stable/liberty, but not into master until we have refactored everything
16:29:15 <j^2> makes sense,
16:29:25 <sc`> yup
16:29:52 <jklare> i think there might be some more patches like this and imo we should move most of them to stable/liberty to allow people to work on the cookbooks for liberty while we refactor for mitaka
16:30:18 <jklare> not sure if we should do the moving or just leave a comment like i did
16:30:47 <sc`> i agree. probably at least a communication to the ml
16:32:00 <sc`> might be worthwhile to just do the moving, unless there's suddenly a huge influx of changes
16:32:11 <jklare> agreed
16:32:41 <sc`> for the most part, just education and awareness will be enough
16:32:51 <jklare> ok
16:32:59 <jklare> just wanted to mention it
16:33:27 <j^2> yep, good that you did, but all in all it’s just lets continue on with what we’re doing?
16:40:18 <sc`> so next week we can expect for master to start being broken?
16:41:08 <j^2> i think so
16:45:37 <jklare> yep
16:49:46 <sc`> should definitely heads-up the ml since we're expecting master to be broken, just to be proactive and get ahead of anyone running into known broken things
16:51:32 <jklare> sounds good
16:55:41 <jklare> anything else we should talk about?
16:56:21 <sc`> i'm good. keeping an eye on rdo packaging efforts in the meantime
16:58:58 <jklare> ok, thanks for attending
16:59:04 <jklare> #endmeeting