17:03:13 <jbrooks> #startmeeting fedora_atomic_wg
17:03:13 <zodbot> Meeting started Wed Jul 26 17:03:13 2017 UTC.  The chair is jbrooks. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:03:13 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:03:13 <zodbot> The meeting name has been set to 'fedora_atomic_wg'
17:03:19 <jbrooks> #topic Roll Call
17:03:24 <jbrooks> .fas jasonbrooks
17:03:25 <dustymabe> .hello dustymabe
17:03:25 <zodbot> jbrooks: jasonbrooks 'Jason Brooks' <JBROOKS@REDHAT.COM>
17:03:28 <zodbot> dustymabe: dustymabe 'Dusty Mabe' <dustymabe@redhat.com>
17:03:35 <miabbott> .hello miabbott
17:03:36 <jberkus> .hello jberkus
17:03:37 <zodbot> miabbott: miabbott 'Micah Abbott' <miabbott@redhat.com>
17:03:39 <zodbot> jberkus: jberkus 'Josh Berkus' <josh@agliodbs.com>
17:04:45 <jbrooks> #chair dustymabe miabbott jberkus
17:04:45 <zodbot> Current chairs: dustymabe jberkus jbrooks miabbott
17:04:50 <jbrooks> A select group
17:05:10 <jbrooks> #topic action items from last meeting
17:05:34 <jbrooks> jberkus to follow up with Fedora infra about moving pa.io MLs to  Fedora infra
17:05:56 <jbrooks> jberkus to write blog post for fedora magazine/pa.io on changes
17:06:09 <jbrooks> jberkus to re-create issues in new repo
17:06:19 <jbrooks> How'd you fare on those, Josh
17:06:23 <jberkus> in process.  the Fedora Magazine mailing list had a bit of a meltdown last week
17:07:05 <sayan> .hello sayanchowdhury
17:07:05 <zodbot> sayan: sayanchowdhury 'Sayan Chowdhury' <sayan.chowdhury2012@gmail.com>
17:07:06 <dustymabe> jberkus: meltdown?
17:07:16 <puiterwijk> Helloish
17:07:21 <jbrooks> #chair sayan
17:07:21 <zodbot> Current chairs: dustymabe jberkus jbrooks miabbott sayan
17:07:23 <jberkus> yeah, the mailing list is the first step in pubication, and it wasn't working for 2 days
17:07:24 <jzb> .hellomynameis jzb
17:07:25 <zodbot> jzb: jzb 'Joe Brockmeier' <jzb@redhat.com>
17:07:32 <dustymabe> #chair puiterwijk
17:07:32 <zodbot> Current chairs: dustymabe jberkus jbrooks miabbott puiterwijk sayan
17:07:48 <jbrooks> #chair jzb puiterwijk
17:07:48 <zodbot> Current chairs: dustymabe jberkus jbrooks jzb miabbott puiterwijk sayan
17:08:06 <jbrooks> jberkus, should we re-action any of those?
17:08:12 <kushal> .hellomynameis kushal
17:08:14 <zodbot> kushal: kushal 'Kushal Das' <mail@kushaldas.in>
17:08:27 <jberkus> yes, although moving issues is jberkus & strigazi, since he's moved a bunch
17:08:30 <dustymabe> and also make the action description be more verbose
17:08:42 <jberkus> as a warning, I'll miss next week's meeting
17:09:08 <jbrooks> #action follow up with Fedora infra about moving pa.io MLs to fedora infra
17:09:13 <sayan> #chair kushal
17:09:13 <zodbot> Current chairs: dustymabe jberkus jbrooks jzb kushal miabbott puiterwijk sayan
17:09:14 <jberkus> for moving the lists, we brought it up on atomic communtiy meeting on Monday, and it was approved
17:09:15 <jbrooks> dang
17:09:20 <jbrooks> cool
17:09:22 <jbrooks> #undo
17:09:22 <zodbot> Removing item from minutes: ACTION by jbrooks at 17:09:08 : follow up with Fedora infra about moving pa.io MLs to fedora infra
17:09:39 <puiterwijk> jbrooks: pa.io?
17:09:59 <puiterwijk> I'm going to guess not literally http://pa.io, as that seems... an ad website?
17:10:03 <jbrooks> #action jberkus strigazi to continue moving kube issues to new kube-sig tracker
17:10:10 <jbrooks> puiterwijk, projectatomic.io
17:10:17 <puiterwijk> Ahh. Ack
17:10:19 <sayan> projectatomic.io
17:10:34 <puiterwijk> Sorry, when I read "pa", my mind goes to "pulseaudio" for now
17:10:42 <puiterwijk> jbrooks: just ping me after the meeting and we can work things out
17:10:48 <jbrooks> #action jberkus to write blog post for fedora magazine/pa.io on changes
17:11:06 <jbrooks> Dusty's actions
17:11:08 <jbrooks> dustymabe to get a new fedora calendar entry for the Fedora Atomic
17:11:09 <jbrooks> Working Group meetings
17:11:16 <jbrooks> dusty to put in place updating of last f25 ref automated via bodhi
17:11:22 <jbrooks> dustymabe to write blog post about last f25 updates
17:11:31 <jbrooks> dusty to write a comprehensive blog post on updating f25->f26 and will  include update about f25 updates ref
17:11:41 <jbrooks> dustymabe to create kubernetes-sig pagure repo and notify interested parties
17:11:52 <dustymabe> #info atomic calendar created https://apps.fedoraproject.org/calendar/atomic/
17:11:54 <jbrooks> dustymabe jberkus to write doc on docs vfad with all resources linked
17:12:15 <jberkus> dustymabe: do you have time to work on that today?
17:12:39 <dustymabe> #info patch sent to infra for f25 atomic ref https://pagure.io/atomic-wg/issue/303#comment-450097
17:12:53 <dustymabe> still waiting on infra team to review/merge the patch
17:13:21 <dustymabe> #info kubernetes sig pagure repo created https://pagure.io/atomic/kubernetes-sig
17:13:22 <puiterwijk> dustymabe: that was merged about an hour ago, just before today's mash
17:13:39 <jbrooks> On the upgrade directions, I'll send you that blurb about kube today -- I just figured out the etcd migration bit this am
17:13:40 <puiterwijk> #link http://infrastructure.fedoraproject.org/cgit/ansible.git/commit/?id=39b719c
17:13:41 <dustymabe> puiterwijk: cool didn't see an email from anyone about it (though I haven't checked in an hour)
17:13:43 <jbrooks> dustymabe, ^
17:13:54 <dustymabe> jbrooks: thanks
17:13:58 <puiterwijk> dustymabe: didn't send an email yet, sorry
17:14:07 <dustymabe> #action dusty to write a comprehensive blog post on updating f25->f26 and will  include update about f25 updates ref
17:14:23 <dustymabe> #action dustymabe jberkus to write doc on docs vfad with all resources linked
17:14:32 <jbrooks> Cool
17:14:41 <jbrooks> ksinny to update wiki
17:14:43 <dustymabe> jberkus: yeah I can try to work on that today with you
17:14:47 <jbrooks> What was that one about?
17:14:54 <dustymabe> jbrooks: ksinny did that
17:15:05 <jbrooks> Sweet
17:15:06 <dustymabe> it was related to the mailing list/IRC changes
17:15:20 <jbrooks> Then there was kubernetes-sig to discuss version tagging for kube system containers
17:15:37 <jbrooks> Which, I don't think we did this -- I'll take an action to start a discrete ticket about it
17:15:48 <dustymabe> jbrooks: +1
17:15:56 <jbrooks> #action jbrooks to create kube-sig ticket about versions discussion
17:16:04 <jbrooks> OK, to the ML!
17:16:10 <jbrooks> I mean issue tracker
17:16:21 <dustymabe> jbrooks: :) - looks like we don't have any meeting tickets for today
17:16:28 <dustymabe> but I do know jberkus has something and I do as well
17:16:28 <jbrooks> Oh yeah?
17:16:29 <jbrooks> Nice
17:16:33 <jbrooks> OK
17:16:39 <jbrooks> #topic open floor
17:16:48 <jberkus> btw, we need to get something in there for updating atomic wiki page and pagure homepage to explain where to contact us ...
17:16:48 <jbrooks> jberkus, go ahead, then dusty
17:17:04 <jberkus> dusty first, I'm double-timing until :30
17:17:04 <dustymabe> jbrooks: ksinny updated the atomic wiki page
17:17:09 <jberkus> ah, ok!
17:17:35 <dustymabe> ok i'll go
17:17:46 <jbrooks> shoot
17:18:21 <dustymabe> #info atomic qcow/vagrant images from yesterday's release are set to update from the wrong ref (the updates ref) - we are fixing this, but we need to communicate this out to users
17:18:46 <dustymabe> we need to let people know how to switch back to the right ref
17:19:42 <jbrooks> dustymabe, your upgrades post?
17:19:43 <dustymabe> i'm guessing email - but I can also open an issue that we can point people to in the future
17:19:48 <jbrooks> And email
17:20:09 <dustymabe> jbrooks: yeah - but maybe in the upgrades post I'll just mention it and link to the issue, rather than that being the focus of the post
17:20:17 <jbrooks> Sure
17:20:18 <dustymabe> actually. no
17:20:27 <dustymabe> probably shoulnd't mention it in the upgrades post at all
17:20:34 <dustymabe> because this is only an issue with 'new media'
17:20:42 <dustymabe> if someone upgrades they won't have a problem
17:21:02 <dustymabe> so probably not relevant for an upgrades post, right?
17:21:19 <jbrooks> I guess not
17:21:28 <dustymabe> either way:
17:21:48 <dustymabe> #action dustymabe to communicate out issue with ref in atomic cloud images
17:21:58 <jbrooks> Sounds good
17:22:04 <jbrooks> jberkus, You have an item as well?
17:22:07 <dustymabe> ok jberkus has something for open floor but he is in another meeting for the next 8 minutes
17:22:12 <jbrooks> Ah
17:22:15 <dustymabe> does anyone else have something they would like to bring up?
17:22:20 <jberkus> sorry ...
17:22:25 <jbrooks> no prob
17:22:58 <jberkus> well, laggy discussion: docker in a container
17:23:16 <dustymabe> i.e. as a system container?
17:23:21 <jberkus> I wanted to check is anyone is working on this actively
17:23:22 <jberkus> yes
17:23:27 <jbrooks> The problem w/ it is our versioning
17:23:38 <jberkus> ??
17:24:07 <jbrooks> If you can't say, I want version X of this container, then there's little point in being able to run a different version of docker
17:24:27 <jberkus> so, why can't we?
17:24:33 <jbrooks> The container image will always be a the same version as what we ship in the ostree
17:24:46 <jbrooks> We need to solve the manual versioning vs auto versioning deal
17:25:11 <jbrooks> And really, we could use releases of rawhide images, too
17:25:25 <jberkus> I thought that was already solved?  IIRC, maintainers can put in a manual version if they want
17:25:40 <jbrooks> Is that where it settled?
17:25:40 <jberkus> hmmm, no maxamillion
17:25:43 <jberkus> yes
17:25:55 <jberkus> ideally, we want auto versions for docker too
17:26:00 <jberkus> but this would be a short-term fix
17:26:09 <jberkus> the other question is getting the packages for more advance docker
17:26:10 <jbrooks> OK, so we can do that, so ppl can run older dockers if they want
17:26:22 <jbrooks> No solution for newer-than-stable dockers
17:26:33 <jbrooks> Because we aren't releasing non-stable images
17:26:35 <jberkus> baby steps then
17:26:53 <dustymabe> jbrooks: could people pull from the candidate registry?
17:27:05 <dustymabe> i don't really know what that means, but figure I should ask
17:27:06 <jbrooks> Yeah, I guess that works for most images
17:27:11 <jbrooks> Not my kube ones though
17:27:27 <jbrooks> Because the build system itself can't see the candidate registry
17:27:44 <dustymabe> ahh, I see- so basically because your images are 2nd order images
17:27:46 <jbrooks> But I guess that's not a problem for the docker pkg, so that's actually a good use case
17:28:18 <jbrooks> It's annoying, because 1.7 is the current latest kube
17:28:22 <jbrooks> We have it in rawhide
17:28:26 <jberkus> FYI, one reason this is coming up now is that the CRI-O people are making system contianers to run CRI-O, and want the ability to NOT have docker on the host
17:28:30 <jbrooks> And system containers would be perfect
17:28:47 <dustymabe> are there any action items we can come up with for this?
17:28:50 <jbrooks> But we can't do it -- I could remake them to not share images, but it'd be wasteful
17:29:08 <jberkus> jbrooks: so we need to bring up rawhide containers, then, but we need maxamillion for that discussion
17:29:10 <jbrooks> jberkus, the new ostree lets us remove
17:29:23 <dustymabe> jberkus: we've already brought it up
17:29:36 <jbrooks> jberkus, That ticket I'm going to create on the kube-sig list will touch on this
17:29:37 <dustymabe> it's under his mound of work, so it will come in time
17:29:46 <jberkus> yah
17:29:55 <dustymabe> jbrooks: i was supposed to create a ticket on the atomic-wg issue tracker for this
17:29:59 <dustymabe> and I never did
17:30:28 <dustymabe> #action dustymabe to create ticket to track RFE for rawhide based containers to be made available from registry.fedoraproject.org
17:30:48 <jbrooks> dustymabe, did we ever get a new docker image in the container registry
17:31:10 <jbrooks> Oh, I see container-engine
17:31:21 <jbrooks> http://pkgs.fedoraproject.org/cgit/container/container-engine.git/
17:31:57 <jbrooks> jberkus, Another cool thing would be to build a f25 copy of that image, so ppl could run an older docker
17:32:06 <jbrooks> 1.13 is technically not supported for kube
17:32:13 <jbrooks> although it seems to work fine
17:32:26 <jbrooks> f25's docker on f26 via system container
17:32:34 <dustymabe> ok we are starting to get detailed - any meeting related stuff surrounding this conversation?
17:32:41 <dustymabe> action items?
17:32:47 <dustymabe> or anyone else with anything for open floor
17:32:57 <dustymabe> walters: I think you guys just found an issue with atomic workstation?
17:33:13 <walters> in it not shipping updates?  yes
17:33:14 <dustymabe> we should probably start tracking those issues in the atomic-wg issue tracker - do you mind opening one?
17:33:18 <jberkus> are we officially calling it Atomic Workstation?
17:33:43 <dustymabe> jberkus: I haven't heard otherwise
17:33:56 <jberkus> lots of folks had objections
17:34:09 <jberkus> the other open question was whether it belongs to us or Workstation team
17:34:09 <puiterwijk> jberkus: I've been told to use that name until a better one emerges
17:34:13 <jberkus> I thought them, but no?
17:34:25 <dustymabe> jberkus: i actually don't know
17:34:44 <puiterwijk> dustymabe: funny thing: the latest rawhide tree won't mount luks devices :).  (haven't had time to debug yet)
17:34:57 <jbrooks> All right, I think we may be ready to close
17:35:03 <dustymabe> i know the people who have been owning it are mostly a small group of individuals
17:35:07 <dustymabe> who are mostly here in this meeting
17:35:21 <dustymabe> jbrooks: +1
17:35:29 <jbrooks> #endmeeting