15:33:06 <markvan> #startmeeting Openstack-Chef
15:33:07 <openstack> Meeting started Thu Feb 12 15:33:06 2015 UTC and is due to finish in 60 minutes.  The chair is markvan. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:33:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:33:11 <openstack> The meeting name has been set to 'openstack_chef'
15:34:34 <markvan> Hi, it's meeting time again.   j^2 is out, so I'm subbing.
15:35:11 <markvan> Anybody around?  kinda quiet
15:35:26 <krtyy> Here!
15:36:06 <markvan> hi krtyy, any topics you want to chat about?
15:36:42 <krtyy> Nope, I'm good. My last change is waiting in the queue. I've got a list of things for the next release but it's still a work in progress.
15:37:41 <markvan> k, since its just u and me I'll keep this short.  I had 3 topics to give some updates on: mysql, juno patch progress, bootstrap
15:37:43 <j^2> markvan: did you see the Gemfile PR on the testing stack?
15:38:26 <j^2> Daddy has been put in the corner by the nurses, so I have a sec :)
15:39:03 <markvan> yeah, but have not looked at it much.  off hand I don't think that's a direction we want to go, opens up to many env issues.  and chef dk is easy to install and use
15:39:17 <markvan> #topic Gemfile for testing stack
15:41:40 <markvan> I think that would confuse the issue for the testing suite.  The only reaason I thought we might need a Gemfile is to handle some issue with our specifc cookbooks, not trying to worry about other cookbooks.
15:42:27 <markvan> I do like the run_command updates to the rake file, we should grabat least those, but I think we should wait on this gem stuff and not clutter that repo.
15:42:58 <markvan> Also, using "chef exec" it's easy to isolate chef dk from other ruby gem and bundles.
15:43:41 <markvan> k, nm on the rake file changes, I see he's just hooking in to bundler.
15:43:52 <markvan> I'll post my opinion out there
15:44:07 <j^2> Yep that's really it. Its actually pretty benign
15:45:54 <sc`> what is needed in a gemfile that chefdk doesn't provide? just curious
15:47:18 <markvan> Right now, using the ChefDK for standing up our testing suite, no other gems are needed.
15:48:29 <sc`> git rm Gemfile;git comit -m "we don't need no stinkin' gemfiles";git push origin master
15:48:32 <sc`> *cough*
15:48:45 <sc`> that'd have been funnier without the typo
15:48:58 <markvan> But, right now, since we still have gem files within our individual cookbooks, if you also used ChefDK for dev work, you would go down the normal bundler/gemfile path to do unit,lint,style testing.
15:48:58 <markvan> It could be that in kilo we decide to even remove the Gemfiles from all our cookbooks and rely on a level of ChefDK instead, would simplify our lifes.
15:49:01 <markvan> lives
15:50:13 <sc`> I do like that folks are centralizing around chefdk. I introduced it when I joined my team at $job, but we're still doing gem-based stuff for the most part
15:51:14 <markvan> yeah, it's still kinda new and getting past those initial bugs that hold folks up.  ver 0.4 is looking pretty good for me right now
15:52:01 <sc`> I need to 'splode my personal laptop environment and redo all the rvm whatsits to make chefdk play nicely again. it's thoroughly broken
15:52:48 <sc`> knife even segfaults :D
15:53:30 <markvan> yeah, I've always tried to avoid rvm churn, and just use the isolated chefdk setup.
15:54:20 <markvan> j^2: so, your call, if it's explained a bit better in the readme, and is not the default/main testing patch, I guess I don't mind allowing folks to have the gemfile option in the testing suite.
15:54:29 <markvan> path
15:56:20 <markvan> moving on
15:56:24 <markvan> #topic mysql
15:57:07 <markvan> I see we have a couple new patches for putting in mariadb support.   This is with the existing  5.x level of the mysql cookbook.
15:57:36 <markvan> I'm also about to push up patches to take mysql to the v6 level, which will break these mariadb patches a bit.
15:57:49 <markvan> Any opiinion on which order these should go in?
15:58:40 <markvan> Looking at my patch, one of the bigchanges is that now the mysql attributes are gone, instead they are provider variables, so we need to provide our own set of mysql attributes.
15:58:56 <markvan> #link https://review.openstack.org/#/c/155222/
15:59:04 <sc`> bring mysql up to v6 and refactor the patches, I'd say. something's going to break one way or the other
15:59:18 <markvan> #link https://review.openstack.org/#/c/155220/
15:59:52 <markvan> Yeah, that's what I had in mind, since mysql is our main path right now for juno and ubuntu 14.
16:00:37 <markvan> I'll try to get my patches up yet today and make comments in the mariadb ones to reflect this.
16:01:15 <markvan> So, far the changes to the mysql is not that bad, just the shuffling around wtih those attributes worries me a bit
16:01:54 <sc`> that sounds backward breaking, but I'm coffee-1 right now
16:03:08 <markvan> yup, mysql cookbook took care of that already, they removed all those attributes which we used and overridden.  So, I'm just trying to put something similar in place to support what types of values we had before.
16:03:32 <sc`> gotcha
16:03:34 <markvan> But I moved the name space from [mysql]   to [openstack][mysql]
16:03:58 <markvan> since the mysql namespace is gone now
16:04:22 <sc`> makes sense
16:05:15 <markvan> I'll finish and push, then let the comments fly, maybe folks want complete backward compat, and we could temp put in that old name space to ust prime the new ones, but make them deprecated
16:06:07 <markvan> I don't buy into much for backward compat, usually not worth the efforts for the speed of change to the base openstack
16:06:56 <sc`> understandable. I'm an old school FreeBSD guy, so backwards compatibility is all we do :p
16:07:36 <markvan> #topic juno patch "star" list
16:07:58 <markvan> Just a quick update on the Juno patches that I think are waiting for reviews:
16:08:00 <markvan> url: https://review.openstack.org/154634
16:08:00 <markvan> url: https://review.openstack.org/150559
16:08:00 <markvan> url: https://review.openstack.org/153695
16:08:00 <markvan> url: https://review.openstack.org/154639
16:08:01 <markvan> url: https://review.openstack.org/154641
16:08:03 <markvan> url: https://review.openstack.org/154976
16:08:05 <markvan> url: https://review.openstack.org/154887
16:08:07 <markvan> url: https://review.openstack.org/154883
16:08:09 <markvan> url: https://review.openstack.org/154801
16:08:12 <markvan> url: https://review.openstack.org/154647
16:08:14 <markvan> url: https://review.openstack.org/151002
16:08:16 <markvan> url: https://review.openstack.org/153376
16:09:05 <markvan> These plus the mysql ones are needed for Juno RC1.  If you think I missed one, speak up.   There are other nice patche/fixes out there, but none are stopping a Juno release
16:09:53 <markvan> #topic bootstrap
16:10:55 <markvan> Just a quick update on bootstrap progress.  The infra patch went in
16:10:57 <markvan> #link https://review.openstack.org/#/c/154207/
16:11:55 <markvan> And this patch to Common will be the start for the cookbook side of this:
16:11:58 <markvan> #link https://review.openstack.org/#/c/154120/
16:13:03 <markvan> This is work for the new kilo test gates which will be based upon this bootstrap and the rakefiles.  And hopefully not to long after the CI testing with our testing suite
16:14:16 <markvan> #testing suite and repo
16:14:22 <markvan> #topic testing suite and repo
16:15:37 <markvan> Just an update on the testing suite and repo.  The plan is to branch off the current repo:
16:15:37 <markvan> https://github.com/stackforge/openstack-chef-repo
16:15:37 <markvan> and replace it with the testing suite based upon chef provisioning and chef dk.
16:15:37 <markvan> https://github.com/jjasghar/chef-openstack-testing-stack
16:16:02 <markvan> I see these 3 patches to cleanup the "old" repo, I they are fine for that:
16:16:02 <markvan> https://review.openstack.org/#/q/owner:%22Pierre+Rambaud%22+status:open,n,z
16:16:29 <markvan> I have already comment in them about the up coming switch to the new testing suite
16:16:52 <openstackgerrit> ZongKai LI proposed stackforge/cookbook-openstack-network: Enable Distributed Virtual Router  https://review.openstack.org/151083
16:16:54 <j^2> I think we can just do an infra patch to pull upstream or something
16:17:30 <markvan> j^2: not sure what you mean there? pull what from upstream
16:18:21 <j^2> Change the upstream to the master on my github and then I can deprecate my repo
16:18:39 <j^2> I'm pretty sure I can
16:19:05 <markvan> ah, I see, rather then just a new patch set to the repo that replaces all the files?
16:19:22 <j^2> Yeah wouldn't that be easier?
16:19:36 <j^2> I really haven't dug that deep honestly
16:20:09 <markvan> there are that many files in there, I think a patch would be easist and done quickly.
16:20:42 <markvan> The key is branching off the existing (after those outstanding 3 patches get in)
16:21:45 <markvan> #topic general discussion
16:21:57 <markvan> anything else folks want to bringup?
16:22:21 <krtyy> I'm good.
16:24:27 <markvan> thx all, I'll be updating the ML discussion on Juno for Monday's meeting.
16:24:30 <markvan> #endmeeting