15:00:28 #startmeeting akanda 15:00:29 Meeting started Mon Apr 27 15:00:28 2015 UTC and is due to finish in 60 minutes. The chair is sarob. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:30 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:34 The meeting name has been set to 'akanda' 15:00:41 roll call 15:00:44 o/ ! 15:01:18 \o/ 15:03:01 im pinging ryan and david 15:03:10 o/ 15:03:28 pinged 15:03:33 :) 15:05:17 o/ 15:05:22 cool 15:05:39 #link https://wiki.openstack.org/wiki/Meetings/akanda#Agenda 15:05:49 pretty simple this morning 15:05:59 lets review the bp 15:06:11 #topic review blueprints 15:06:20 #link https://blueprints.launchpad.net/akanda 15:06:42 oh 15:06:56 WELCOME Adam to the akanda team!!!! 15:07:15 thanks, sarob :) 15:07:39 hey adam! 15:07:44 fancy irc welcome 15:08:33 so i took what was in trello 15:08:39 :D o/ 15:08:52 summarized a few of them, like oslo fix ups 15:09:10 and here you go 15:09:49 im going to merge specs dir 15:10:00 for more details 15:10:22 details are good 15:10:28 is there anything major missing from the bp 15:10:51 assuming the title says enuf to understand 15:10:56 what its about 15:11:02 probably something for image building? 15:11:17 hmm, right 15:11:37 that got munged 15:11:39 before Mark left, he was working on getting a process in place for easily building and uploading appliance images with the same tooling infra uses 15:11:49 DIB? 15:11:53 yes 15:11:56 (yep) 15:12:10 I'm also using DIB to build the image for the ingix thing we're doing 15:12:17 DreamHost folks have been using veewee *shudder* 15:12:19 +1 to DIB 15:12:32 we probably need a BP for building nightly + stable images from source 15:13:05 a lot of the work is done, but it's been on hiatus since Mark left 15:13:10 the ci-updates blueprints mentions it 15:13:27 yeah, i jammed them together 15:13:36 k, that works 15:13:40 it's probably higher priority, though 15:13:47 davidlenwell got stuck on it because we don't have public images currently 15:14:03 ryanpetrello_: so you saw my messages from last night ? 15:14:08 yep :) 15:14:12 nice .. 15:14:14 what about post-commit image builds? the ironic team does something similar with their deploy ramdisk images, that uploads new builds to tarballs.openstack.org after each commit merges to relevant repos 15:14:33 adam_g: +1 15:14:46 thats how we should handle it 15:14:51 yea, at this point, *anything* would work 15:14:59 I currently develop with the veewee-based image we use for DHC production 15:15:18 adam_g +1 15:16:17 so you guys want to start adding some details straight into the bp 15:16:29 I think thats wise .. 15:16:31 i will patch the akanda repo for specs 15:16:42 sounds like a plan.. 15:16:55 so we can start creating specs for the bp real details 15:17:16 im not going to worry about tox spec jobs for right now 15:17:42 #action sarob patch akanda repo for specs 15:17:57 #action team update existing bp with details 15:18:07 so what about timing 15:18:39 do we, can we branch kilo before the summit? 15:19:07 what in this list needs to drop out? 15:19:43 I feel like rug-ha is pretty aggressive 15:20:03 given that I don't think we actually have a working akanda + kilo in-tree 15:20:06 (at all) 15:20:19 some of it's partially implemented in Mark's forks, some of it's still in his brain 15:20:39 but it's possible that he's made more progress on that front that I've been aware of 15:21:01 I hope he hasn't .. he is supposed to be on a honey moon 15:21:07 ;) 15:21:19 "we" need to spread the joy and get 15:21:28 the rest of us working on the bits 15:22:05 I'll be of more help on this by the time he gets back.. finally getting a working dev env going.. hopfully finishing the final bits of that todayw ith ryans help. 15:22:09 yea, that's probably first going to require actually getting you all spun up on dev environments with this stuff 15:22:18 that ^ 15:22:42 brb, gotta grab a power adapter 15:23:42 so lets assume that the trello dev sprint was correct 15:23:49 that i moved over into bp 15:23:53 for now 15:24:08 we as a team focus on documenting the bp 15:24:26 breaking them down into smaller, patchable pieces 15:25:05 being that david is just getting in the flow 15:25:14 and adam is just starting 15:25:24 we have a bit a catch up to do 15:25:39 are we using specs to per review those details before they go into the BP? or just going straight into the BP for now? 15:25:41 I can work with adam to get him up to speed faster than I did 15:25:45 *peer 15:25:57 davidlenwell: good plan 15:26:16 adam_g: start with the bp for right now 15:26:32 ill patch specs right after this meeting so 15:26:39 k 15:26:47 you can start braindumping into there 15:26:50 I think some items will need specs before we should expect anyone to know what to do wit hthem 15:27:00 davidlenwell: yeah 15:27:18 we dont quite know where to start yet 15:27:38 although getting the nightly images builds going is a good one to start 15:27:42 a lot can be derived from context .. but the less guessing the better for me anyways 15:27:52 yeah 15:28:14 sarob: I think what adam was suggesting is that we forgott nightly builds and move to a post commit image build 15:28:23 that way we always have images that are up to date 15:28:39 davidlenwell: yeah, i already merged the two in my head 15:28:44 which I think is smart 15:28:50 very smarty 15:29:10 IMO image builds should be the artifact of successful devstack-tempest jobs run in the gate, so if we can get devstack jobs running against patches we are most of the way there.. is there a status on the devstack integration at this point? 15:29:22 +1 15:29:26 I was discussing this very thing w/ Mark 15:29:36 re: image build promotion 15:29:47 so no pending patches in review 15:29:57 clean slate! 15:29:59 i was planning on hacking the devstack support as part of my first deep dive into akanda/setting up a dev env 15:30:00 so there's ansible recipes to build out a devstack 15:30:08 but currently I'm using the DH image build (which I mentioned earlier) 15:30:12 adam_g: 15:30:17 ah 15:30:18 Mark and I were working on getting a recipe to build out the appliance w/ dib before he left 15:30:22 and it had a few small kinks left 15:30:27 (that he was working on) 15:30:44 so we should probably prioritize that, a job that can build an appliance image from source + devstack + tempest 15:30:48 he did mention that to me before he left when I brought up dib for the ingix image 15:31:01 cool. i think a devstack plugin would be a good idea.. we can host it in one of the akanda repos and enable it in devstack via localrc 15:31:11 yep 15:31:15 that support's in Kilo, right? 15:31:19 the link to a plugin that devstack will download? 15:31:20 yup 15:31:23 I know hew as going that route, too 15:31:37 the stuff I was working on was just getting a devstack + appliance source job for Juno 15:31:41 when does mark return? 15:31:43 because as I mentioned, Kilo doesn't really work yet 15:31:45 (good question) 15:31:47 next week? 15:32:03 i think tomorrow 15:32:08 oh, okay 15:32:36 sooo, thats all i have for today 15:32:53 we need to document the bp 15:33:05 and create some specs for the biggins 15:33:26 #topic any other business? 15:34:26 any thing else meeting wise? 15:34:48 I'd like the express my exictement for our growing team! welcome adam_g! 15:35:07 davidlenwell: +1 15:36:11 okay, lets call this one 15:36:13 thanks guys, excited to be here :) 15:36:27 :) 15:37:03 #endmeeting