14:31:10 #startmeeting Openstack-Chef 14:31:11 Meeting started Thu Mar 26 14:31:10 2015 UTC and is due to finish in 60 minutes. The chair is j^2. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:31:12 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:31:15 The meeting name has been set to 'openstack_chef' 14:31:17 hey everyone! 14:31:34 o/ 14:32:43 hello 14:33:38 o/ 14:33:49 :) 14:34:09 with the blueprint movement, i think it's safe to say we have a first pass at what we'd like done for kilo 14:34:22 \o/ 14:34:30 Did you get my emails? 14:34:56 unrelated, i've been talking with coworkers about upstreaming things that make sense, starting small to get over the learning curve 14:35:33 Looks like Sahara is a no go unless someone here wants to step up and lead. Also Pierre does not have time to finish those tests. I think @markvan said that he might finish them up. 14:35:47 cmluciano: cool, that works 14:35:57 sc`: nice! 14:36:26 yup, I'll talke a look at the tests at some point, I'll update the blueprint to take effect 14:37:24 Ma Wen Cheng proposed stackforge/cookbook-openstack-image: Use 'identity_uri' to configure admin URI for kilo https://review.openstack.org/168015 14:39:45 j^2: one change already came out of those discussions, too 14:40:53 we need to continue with the reviews like always 14:41:13 @core please if you have time check the patches and +2 or +WF where you can 14:41:13 @j^2 @markvan @mattray @wenchma @jklare @cmluciano please if you have time check the patches and +2 or +WF where you can 14:42:37 as we get more aligned to how things work and should work wrt code style and project goals, more patches should bubble up 14:42:41 Yeah, for patches, as you can see, with the moveto kilo, there are a bunch of cleanups in progress. I wish I could share my "starred" list easily. 14:43:15 For example, there are several repo related ones that are ready. 14:43:28 https://review.openstack.org/#/q/status:open+project:stackforge/openstack-chef-repo,n,z 14:44:56 And to hightlight a few others that are ready: https://review.openstack.org/#/q/status:open++branch:master+topic:bug/1433405,n,z most of the test is already in. 14:45:31 And one critical patch that currently is breaking tests is: https://review.openstack.org/#/c/167891/ 14:46:12 though i know all new work goes into master, is stable/icehouse completely frozen to changes? i don't have anything in mind, but just to satisfy curiosity 14:46:53 sc`: more or less, yeah. 14:47:04 unless you find crazy breaking changes, we can backport 14:47:09 we try to do n-1 14:47:10 support 14:47:11 ok, cool 14:47:24 i thought that was the case 14:47:26 Checkout the wiki for branch status. Usually we have last stable release open, and in fact we have already patched stable/juno once: https://review.openstack.org/#/c/164788/ 14:47:43 https://wiki.openstack.org/wiki/Chef/GettingStarted at bottom 14:48:12 for some reason my bookmark for that disappeared 14:48:22 bookmarked again, thanks 14:49:13 yeah, trying to make the wiki a good place for general info, any comments on it are welcome! 14:51:19 oh, so next week 14:51:23 it's ChefConf 14:51:46 I can't run the meetings on monday or thursday being unavailable 14:51:56 I will also be away 14:52:21 I can help with running the meetings 14:52:38 i'll be at chefconf as well, with a contingent of my colleagues 14:52:46 nice! 14:52:58 markvan: sounds good 14:53:02 one of our guys is presenting, too 14:53:25 def all should meetup 14:53:34 seems like something that needs to happen 14:53:39 Just need @markvan 14:53:43 * markvan wow, i have to push harder to make chefconf travel happen... 14:53:47 heh 14:54:34 markvan: ibm is suppsode to be there in force too 14:54:48 maybe you can offer you services to fill in :) 14:57:13 bummer, yeah would like to meet some of those ibmers in person... 14:57:31 i think it's all do to the POWER work 14:57:41 i could be wrong though 14:58:54 ah, ok Power is always a fun topic 15:00:08 cool, any one would like to take the floor for a topic? 15:01:23 I have a few... 15:01:28 go ahead 15:01:35 i don't know if this is necessarily a whole topic in and of itself, but how do i keep up with changes in the pipeline? i want to start shifting my efforts more in the direction of working off of master, and want a better grasp of how it's currently done 15:01:56 #topic new "service" role https://review.openstack.org/#/q/status:open++branch:master+topic:bug/1436050,n,z 15:03:26 Just a quick update on this one. We have been using a "service" user for a long time with the default "admin" role. Now keystone has changed to by default have a new "service" role. So, these patches are to use that new role, we should notice no differences in our tests. 15:04:06 awesome. the least privilege that can be given, the better 15:04:42 yup, exactly, and you will see in the keystone policy.json, that the service role is now being used to separate admin from service a bit 15:05:17 #topic remove api-paste.ini https://review.openstack.org/#/q/status:open+branch:master+topic:bug/1433152,n,z 15:05:41 moving things out of admin will make my security guys very happy 15:05:56 Another quick update. As i was trying to get conf's sync'd with latest kilo changes, I noticed that many of the api-paste.ini were out of date. 15:06:46 Also, we had a tempate for them that had no logic or templates, so I think it best to just remove these templates until such a time as we truely need them, that will reduce our release to release sync work. 15:07:05 (we did the same for the policy files in icehouse) 15:07:26 the api-paste.ini files are shipped with the packages straight from the base openstack source. 15:08:12 yeah, 15:08:24 there is no need duplciate them 15:08:28 makes sense 15:08:52 k, cool. 15:09:01 #topic identity_uri https://review.openstack.org/#/q/status:open+branch:master+topic:bp/identity-uri,n,z 15:10:04 I also started on this blueprint. After reviewing the existsing code, I decided to make the recipes clean by creating a new helper method in Common since the identity_uri is slightly different from the endpoint. 15:10:23 identity_uri affects all components 15:10:29 I have updated the blueprint and put out the starting patches, I think they are fairly straight forward. 15:11:06 yup, I started with block-storage as an example to go with the common patch I did. Once those are done, I will push up all the others 15:11:15 need the openstack-common lib patch merged first 15:11:26 for other patches jenkins pass 15:11:41 yes 15:12:26 btw, I working on a new version of the testing repo patch script to allow multiple patches to be pulled in, like in this case where a Common + component patch is needed. Will make it very easy to test these out. 15:12:49 looks better that add a method identity_uri_transform 15:13:31 yeah, I think that's the right design for what the cookbooks are doing now with regards to the other public uri, auth_uri. 15:16:24 sc`: did you want to discuss how to keep up with changes in the pipeline? 15:29:16 j^2: my last topic was about reviews, since were knee deep in kilo changes, we need a bit more focus on reviews to keep the back log down to a reasonable size. I would like to suggest again that we come up with some "Office hours" where at least the cores go at the reviews 15:30:27 i think the challange is to get everyone available 15:30:29 maybe twice a week for an hour or so? And allow folks to come in an help explain their review needs/coments/qyuestions 15:31:19 it’s a great idea in theory; we’d have to make it Wednesday, i’m betting 15:31:21 yeah, but we could try maybe a morning session and an evening session? with at least 2 cores present to make it work... 15:32:22 Maybe can poll on the ML to get more feedback on that... 15:32:27 yeah 15:32:36 I'll start that thread. 15:34:01 btw, I've been meaning to ask....if there a way to have our ML updates (summary/title line) posted here, like is done with the patches? 15:34:44 might make it easier to link the ML to a wider audience here in the irc... 15:38:51 yeah 15:38:57 so i think there’s a way 15:39:02 but it’s not great not atm 15:39:08 #endmeeting