16:01:48 <jbrooks> #startmeeting CentOS Atomic SIG
16:01:48 <centbot> Meeting started Thu Apr 27 16:01:48 2017 UTC.  The chair is jbrooks. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:48 <centbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:01:51 <jlebon> howdy
16:01:59 <jbrooks> #chair walters jlebon jberkus
16:01:59 <centbot> Current chairs: jberkus jbrooks jlebon walters
16:02:00 <kbsingh> hi guys, am partially here
16:02:08 <jbrooks> #chair kbsingh
16:02:08 <centbot> Current chairs: jberkus jbrooks jlebon kbsingh walters
16:02:20 <jbrooks> #topic downstream release
16:03:29 <jbrooks> We're ready to move on a new release, I have this PR to merge that bumps the version # and does a redo on the signing issue we're seeing w/ the static delta in the repo: https://github.com/CentOS/sig-atomic-buildscripts/pull/268
16:03:30 <walters> it looks like we're still hitting gpg issues? https://github.com/CentOS/sig-atomic-buildscripts/pull/268
16:03:54 <jbrooks> Yeah, we never fixed it -- kbsingh has to do it, I don't have access
16:04:20 <kbsingh> do we need that PR to go in before we do the next build ?
16:04:25 <jbrooks> Yes
16:04:42 <kbsingh> otherwise, the repo build + new media etc, is all setup to run from automation at 23:00 UTC Fri the 28th
16:04:57 <kbsingh> and its going to push the content to buildlogs + images into the right place, on its own
16:05:16 <kbsingh> I've walked this 2 times manually, and have a fair confidence that it will just work
16:05:19 <jbrooks> OK, will the repo go to buildlogs, or to the current location?
16:05:29 <kbsingh> to buildlogs, along with the images
16:05:34 <jbrooks> Because if we're moving to buildlogs, we need to change the release package
16:05:45 <kbsingh> but once +1'd from a few people, anyone on the infra side can push as needed
16:05:56 <jbrooks> That's great
16:06:18 <walters> is there any public information/source code for "automation" ?
16:06:23 <jbrooks> Maybe we can get the release package changed today? Hosts need to know where to get their updates
16:06:25 <kbsingh> i think we should redo the release side, once we are sure we are going to move - at the moment, the buildlogs push is to make the repo's just-built, more accessible than needing me to rsync and scp bits around
16:06:47 <jbrooks> OK, so it'll get built on Fri, and then you'll need to actually move things around
16:06:55 <kbsingh> walters: there should be, but the hostnames etc wont be in there obviosuly. the other part, i am going to just put into the downstream branch
16:07:03 <walters> ok, cool
16:07:17 <kbsingh> jbrooks: *I* wont, stuff gets to buildlogs, then hughesjr or Arrfab can move it as needed, wen needed
16:07:32 <jbrooks> All right, that's good
16:08:55 <jbrooks> OK, anything else on downstream?
16:09:57 <jbrooks> #topic continuous / devel
16:10:07 <jbrooks> jlebon, walters Any items to discuss on this front?
16:10:39 <walters> it's been reliably doing its job, but we haven't made any fundamental improvements
16:10:44 <walters> i have kind of stopped doing the alpha cuts
16:10:45 <jbrooks> I think it's been a little while since we've had an alpha release?
16:10:49 <jbrooks> right
16:11:04 <jbrooks> You figure it's not really needed?
16:11:19 <jberkus> who's the target for alpha-not-continuous?
16:11:28 <walters> we did add a smoketested branch
16:11:39 <jlebon> i do want to enhance the smoketesting so that we also sanity check images
16:11:47 <jlebon> that's https://github.com/CentOS/sig-atomic-buildscripts/issues/229
16:12:12 <walters> jberkus, i was using it personally on my home server for a while (but rebased a while ago to FAH - that's another story)
16:13:28 <walters> broadly speaking i do think there are too many streams =)  so maybe it isn't worth it
16:13:55 <jbrooks> OK, maybe we should edit the wiki page
16:14:05 <jlebon> i do use alpha sometimes for rpm-ostree development. when continuous is completely broken
16:14:08 <jberkus> jbrooks: the wiki page needs to be updated in general
16:14:59 <jbrooks> I've used it when I want to test come containerized kube stuff on a system w/o the components, but I could just as well use another branch
16:15:33 <walters> strawman then: edit the wiki to drop alpha, and point people at smoketested
16:15:52 <jlebon> +1
16:15:53 <jbrooks> jberkus, We should file some issues for needed edits / adds
16:16:00 <jbrooks> +1
16:16:25 <walters> #action walters to redirect cahc/alpha → cahc/smoketested
16:16:35 <jlebon> walters: alternatively, make alpha become smoketested?
16:16:41 <jbrooks> That works, and I can talk about that in the next release blog post for downstream
16:16:51 <walters> jlebon, yeah, it's possible
16:17:03 <jberkus> wait, does smoketested already exist?
16:17:23 <jlebon> jberkus: yup! it's awesome
16:17:56 <jberkus> ok, +1 for alpha to be an alias for smoketested if it's not difficult to do
16:18:01 <jberkus> if it is difficult, then just drop alpha
16:19:11 <jlebon> if there's no opposition to that (committing to alpha directly), i can update jobs
16:19:35 <walters> are you just going to symlink it?  or do write-ref?  (well, we can take this offline)
16:20:14 <jbrooks> Yeah, the nice thing about smoketested as a ref name is that it's somewhat meaningful
16:20:23 <jbrooks> Where it's unclear what alpha is supposed to mean
16:20:34 <walters> let's move details of this to https://github.com/CentOS/sig-atomic-buildscripts/issues/269
16:20:41 <jbrooks> I always see alpha and think: broken
16:20:53 <jbrooks> All right
16:21:03 <jbrooks> #open floor
16:21:09 <jbrooks> Other items this week?
16:21:35 <jbrooks> I had an action from last week to write a thing about contributing / getting involved w/ centos atomic, I still need to do that
16:22:20 <walters> i think i mentioned this in passing before, but i'm planning to do a blog post on how I rebased my home server from centosah to FAH...I'm obviously a bit biased but I think it was a really beautiful demonstration of the value of the rpm-ostree model particularly on bare metal
16:22:33 <jbrooks> Awesome
16:22:38 <jbrooks> Oh, kbsingh What's your take on this sig taking up work on cccp-related stuff?
16:22:48 <jbrooks> Is there already another sig that deals with that?
16:23:08 <jbrooks> We've been doing a lot of work around fedora's image registry in the fedora atomic wg
16:23:26 <jbrooks> It's made me wonder where in the centos community that stuff happens
16:25:05 <rbowen> One of my projects at the moment is attempting to attend all of the SIG meetings and get a broader notion of what happens where in the CentOS community, and map that a bit. And, on the other side, what's not happening that could fit.
16:25:22 <jbrooks> rbowen, welcome, glad to have you
16:25:47 <rbowen> That's really just starting, so I don't have an answer to your question above, but I hope to, eventually.
16:25:48 <rbowen> :)
16:25:52 <jbrooks> Sweet
16:26:17 <jbrooks> All right, if there are no other items to discuss, we'll close
16:26:43 <jbrooks> #endmeeting