20:02:08 #startmeeting 20:02:08 Meeting started Tue Jul 29 20:02:08 2014 UTC. The chair is jzb. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:02:08 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:02:43 i'll do a quick status update: I now have tree content sync'd to http://dl.fedoraproject.org/pub/alt/fedora-atomic/repo/ 20:02:45 OK, walters do you have any agenda items / blocker this week to start? 20:02:49 morning 20:02:51 ran a lorax compose today 20:02:54 kernel panics on startup 20:03:10 i've been investigating mirrorlist/metalink 20:03:52 so for agenda; i'd really like to know if there's been any issues from running rpm-ostree as part of the fedora composes 20:03:53 * mattdm waves 20:04:11 #chair agrimm dgilmore walters dustymabe rangerpb nirik stickster 20:04:11 Current chairs: agrimm dgilmore dustymabe jzb nirik rangerpb stickster walters 20:04:16 did it work? did it fail? somewhere in between but the tooling was not documented well enough/sucked? 20:04:17 #chair mattdm 20:04:17 Current chairs: agrimm dgilmore dustymabe jzb mattdm nirik rangerpb stickster walters 20:04:35 walters: was that for anyone in particular? 20:04:40 * stickster here 20:05:01 jzb, dgilmore if he were here mainly, though I don't know if anyone else can work on that? 20:05:10 dgilmore: I think that's for you 20:05:14 *jinx 20:05:37 rangerpb: did you get to that when you checked in with dgilmore ? 20:06:37 yeah, he's working through the compose process, but being in cz and then flock next week -- he's a bit tied up 20:06:57 he has a few things to add yet -- ill circle back with him on that stuff 20:07:10 i will also drop installer and cloud images in here http://dl.fedoraproject.org/pub/alt/fedora-atomic/testing/ 20:07:18 rangerpb: OK. any chance we could do a sprint or something in cz on this? 20:07:19 I know he was upgrading builders to newer koji/pykickstart for docker stuff. 20:07:25 stickster: ^^ ? 20:07:51 i'd love to particpate virtually in a sprint if i can 20:07:52 walters: you're not at Flock, is that right? 20:07:55 jzb: ...possibly? It depends on who's needed. 20:08:15 stickster: I think at a minimum, dgilmore, nirik, and walters 20:08:22 right, and unfortunately it's mainly only the Brno people from my team going 20:08:24 but walters and other folks probably know better than I. 20:09:06 nirik will be in Brno with us on Mon + Tue before Flock, I don't think I have a lock on all that time :-) 20:09:23 jzb, that'd be a great idea ... he also pointed me at ->https://git.fedorahosted.org/cgit/releng/tree/scripts/run-pungi 20:09:30 nirik: We talked about infrastructure talk for Brno folks on Tuesday midday, but other than that -- maybe Monday? 20:09:43 could be. 20:09:56 should we carry this to the infrastructure list? 20:10:11 I'm at y'all's disposal to try to facilitate any of this that I can. 20:10:13 ah ok 20:11:03 * walters saves that link 20:12:09 I think we need to hear back from dgilmore what is outstanding, then perhaps a session monday to try and hash those items out? 20:12:17 jzb: nirik: walters: It'll be a little more challenging to sprint with 6 hrs difference btw team members 20:12:30 yeah 20:13:04 jzb: We have a Fedora meet & greet scheduled after hours with Brno teams, but maybe a few hours after that would be possible too, as long as people take it easy on the beer :-) 20:13:23 (in addition to Mon afternoon Brno time/morning USA time) 20:13:47 stickster: are you suggesting our folks can't work after six or ten beers? I'm not sure that's in evidence. ;-) 20:13:51 hahaha 20:14:02 I can't but I can play OK ;-) 20:14:07 i have been thinking maybe we can do more frequent but shorter meetings? 20:14:13 *nod 20:14:20 Ok, well, let's table the sprint thing for the mailing list? 20:14:22 walters: sure 20:14:59 * agrimm realizes he's missing a meeting, reads scrollback 20:15:52 walters: let's check in again Thursday morning, maybe? 20:16:06 or we can try to sync via the mailing list a bit more too 20:16:07 so for this meeting, i'm happy with the alt/ mirroring at the moment, SSL alone is a huge leap 20:16:26 walters: all thats working ok for you? 20:16:31 there's other things to talk about potentially like ansible automation for the current compose server; mostly blocks on me learning ansible i guess 20:16:49 nirik, i was able to crontab it so i'm OK with that 20:16:59 walters: if i can learn it you can probably learn it in like 12 seconds :) 20:17:07 walters: oh, I noticed that server is selinux disabled... does it need to be? ;) 20:18:20 nirik, https://bugzilla.redhat.com/show_bug.cgi?id=812798 20:18:22 * nirik is happy to help anyone at all with ansible stuff. 20:19:03 walters: ok. :( 20:19:47 ironically it has to be disabled on the compose server in order to allow it to work in the generated images 20:20:01 walters: We have a number of folks helping make ansible playbooks, so I imagine it'll be easy to find help there 20:20:03 fair enough as long as it's known, etc 20:21:06 walters: have you run into anything new that's a problem since last time? 20:21:10 rbergeron, heh =) yeah i will try 20:22:19 jzb, ah...i can't think of anything, it's now mostly blocked on random f21 bugs and the atomic side issues 20:22:38 the thing we can only figure out once dgilmore has stuff going is generating cloud images in ImageFactory in Koji 20:23:26 agrimm was trying to run it directly outside of koji on the current atomic01.qa box, which i could imagine shipping testing images with 20:25:02 this bit of going for koji -> rpm-ostree -> koji -> cloud image is pretty weird but i don't have anything better 20:25:41 anyways so we can probably break now unless people had other things until we can get a meeting with dgilmore in too 20:25:42 walters: so I imagine that ImageFactory will be used to build *all* cloud images? 20:26:06 or are there some that will fall outside of that process? 20:26:31 dustymabe, yeah. The thing that anaconda does well (or at least canonically) is partitioning and such, and the rpm-ostree-toolbox code was always only intended as a stopgap until anaconda ostree support landed 20:26:55 walters: will it still be possible to build images w/o koji? 20:27:13 jbrooks, yeah 20:28:06 jbrooks, does standalone Oz or imageFactory count as "outside koji". or are you specifically asking for rpm-ostree-toolbox to stick around? 20:28:20 mattdm: last time you asked if anyone would like to help with Docker image creation. Do you have any more details on that? 20:28:32 http://meetbot.fedoraproject.org/atomic/2014-07-21/atomic.2014-07-21-15.03.log.html 20:28:38 agrimm: no, those are fine -- so far I've only had success w/ the toolbox 20:28:59 jbrooks, we'll make you successful with imagefactory soon. :) 20:29:06 excellent 20:29:30 dustymabe, he just walked out - i think dgilmore was doing some prep work for this in updating pykickstart 20:29:35 as nirik mentioned 20:30:20 walters: ok. I was just wondering exactly what Docker images are getting generated? 20:30:33 are they example application? 20:30:34 dustymabe, only a base image AFAIK 20:30:38 hey - do we have any other items specific to the meeting, we're at 30 minutes right now 20:30:53 we can keep talking but I want to hit any #action items pretty quickly 20:30:55 i'm OK to break and follow up with anything else on list 20:31:04 any disagreement? 20:31:08 try for another quick meeting this week? 20:31:24 yeah... can we do early morning Thursday or Friday? 20:31:31 or maybe we just need to find a time with dgilmore 20:31:39 walters: let's do that. 20:31:42 k 20:31:58 OK - thanks everybody! 20:32:11 #endmeeting