17:05:39 <number80> #startmeeting Cloud WG (2014-09-05)
17:05:39 <zodbot> Meeting started Fri Sep  5 17:05:39 2014 UTC.  The chair is number80. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:05:39 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:06:01 <number80> #char jzb mattdm agrimm oddshocks geppetto
17:06:09 <mattdm> hi!
17:06:09 * jsmith joins (a bit late)
17:06:14 <jzb> .hellomynameis jzb
17:06:15 <zodbot> jzb: jzb 'Joe Brockmeier' <jzb@zonker.net>
17:06:19 <geppetto> .hellomynameis james
17:06:20 <oddshocks> number80: jeez, messing up already. tsk tsk ;)
17:06:20 <zodbot> geppetto: james 'James Antill' <james.antill@redhat.com>
17:06:22 <roshi> .hellomyname is roshi
17:06:24 <number80> .hellomynameis hguemar
17:06:25 <zodbot> number80: hguemar 'Haïkel Guémar' <karlthered@gmail.com>
17:06:29 <dustymabe> .hellomynameis dustymabe
17:06:30 <zodbot> dustymabe: dustymabe 'Dusty Mabe' <dustymabe@gmail.com>
17:06:36 <imcleod> .hellomynameis imcleod
17:06:37 <zodbot> imcleod: imcleod 'Ian McLeod' <imcleod@redhat.com>
17:06:39 <oddshocks> .hellomynameis oddshocks
17:06:43 <zodbot> oddshocks: oddshocks 'David Gay' <dgay@redhat.com>
17:06:45 <roshi> .hellomynameis roshi
17:06:46 <zodbot> roshi: roshi 'Mike Ruckman' <mruckman@redhat.com>
17:06:52 <roshi> there, did it right this time
17:07:03 <dustymabe> roshi: :)
17:07:10 <number80> #chair imcleod roshi jsmith
17:07:10 <zodbot> Current chairs: imcleod jsmith number80 roshi
17:07:24 <jsmith> .hellomynameis jsmith
17:07:25 <zodbot> jsmith: jsmith 'Jared Smith' <jsmith.fedora@gmail.com>
17:07:27 <number80> #chair jzb mattdm agrimm oddshocks geppetto
17:07:27 <zodbot> Current chairs: agrimm geppetto imcleod jsmith jzb mattdm number80 oddshocks roshi
17:07:36 <number80> everyone is around ?
17:07:40 <number80> ok
17:07:42 <dustymabe> number80: jzb: Here is the link from last meeting: https://lists.fedoraproject.org/pipermail/cloud/2014-September/004194.html
17:07:52 * adamw around if needed (but still clueless)
17:07:58 <number80> thanks
17:08:09 <number80> adamw: you're welcome to join us :)
17:08:21 <number80> #topic Coordinate F21 / fedora next website
17:08:23 <dgilmore> me is kinda here
17:08:29 <number80> https://fedorahosted.org/cloud/ticket/36
17:08:38 <number80> jzb: since you're the lead here ?
17:08:41 <mattdm> website in staging:
17:08:43 <mattdm> http://stg.fedoraproject.org/en/get-prerelease#cloud
17:09:00 <mattdm> ticket the websites people are using:
17:09:02 <mattdm> https://fedorahosted.org/fedora-websites/ticket/267
17:09:27 <jzb> number80: sorry - point of order
17:09:35 <jzb> last week we agreed we'd start with AIs from the prior meeting
17:09:48 <dgilmore> mattdm: big issue
17:09:50 <number80> jzb: my apologies
17:09:56 <dgilmore> atomic is 64 bit only
17:10:00 <jzb> number80: no harm
17:10:05 <dgilmore> website has both 32 and 64 bit
17:10:11 <roshi> true
17:10:13 <roshi> good catch
17:10:29 <jzb> long story short, I'll get them the info this weekend/Monday
17:10:39 <jzb> we discussed it yesterday in the readiness meeting.
17:11:05 <number80> #info staging website: http://stg.fedoraproject.org/en/get-prerelease#cloud
17:11:18 <jzb> or...
17:11:18 <number80> #info related ticket https://fedorahosted.org/fedora-websites/ticket/267
17:11:23 <jzb> mattdm: did you already provide the info?
17:11:28 <number80> #topic Atomic Image
17:11:35 <number80> https://fedorahosted.org/cloud/ticket/64
17:11:50 <mattdm> jzb I gave them some basic info but not really everything
17:11:50 <number80> (just making sure to have useful minutes)
17:12:28 <mattdm> anyway sorry. on with the topic?
17:13:03 <number80> we're discussing the atomic image right ?
17:13:30 <mattdm> yes but I was replying to something previous still. but i'll hold the rest of that.
17:14:45 <mattdm> what *is* the state of Atomic?
17:15:18 <agrimm> mattdm, it's progressing slowly in walters' absence
17:15:52 <agrimm> probably more movement next week.  I have a working vagrant image & cloud image... we just don't have the full workflow string together
17:16:07 <jzb> agrimm: links for those?
17:16:20 <jzb> or are they somewhere we can bang on?
17:16:37 <oddshocks> mattdm: meetings have been pretty null as far as atomic goes, due to only a couple people being able to show up. mostly they've been cancelled unfortunately
17:17:01 <agrimm> jzb, I will get them to some place you can bang on on Monday.
17:17:14 <agrimm> maybe sooner
17:17:14 <number80> #info agrimm has a working vagrant & cloud image (still figuring out the full workflow)
17:17:42 <mattdm> agrimm can that image be produced via the spin kickstarts in koji?
17:18:11 <imcleod> mattdm: took the words out of my mouth....
17:19:53 <agrimm> mattdm, I actually don't know.  The hold-up was originally related to the atomic mirror situation.
17:20:14 <agrimm> I have been using imagefactory to ensure that the transition from what we have now to koji later is as smooth as possible
17:20:24 <agrimm> but I don't know when we can actually make the switch
17:20:46 <number80> maybe dgilmore has some hints about it ?
17:20:51 <imcleod> agrimm: OK.  If you are already driving it through Factory it should be relatively painless to migrate to koji at some point.  I can help.
17:21:42 * dgilmore notes this sounds like f22 type thinsg
17:21:43 <imcleod> agrimm: dgilmore already has nightly build attempts for Atomic using the kickstart file in the spin-kickstarts repo.
17:21:44 <dgilmore> things
17:22:07 <dgilmore> I need to glue up making a nightly atomic tree
17:23:08 <mattdm> dgilmore once that's happening, putting agrimm's work into the spin kickstart should be okay for F21, not delayed to F22, right?
17:23:17 <mattdm> Or are you saying all of this is missing the F21 deadlines?
17:24:09 <dgilmore> mattdm: adding vagrant images is too late
17:24:28 <dgilmore> we should have been looking at that months ago
17:24:45 <mattdm> dgilmore agreed on vagrant, yes.
17:25:08 <mattdm> I'm just worried about the just normal atomic cloud images
17:25:22 <dgilmore> atomic i have no idea what agrimm is doing, but anything more than the docker host tree ive been working on is too late for f21
17:25:22 <mattdm> And I think imcleod and agrimm were talking about both
17:27:09 <roshi> both of what?
17:27:34 <mattdm> roshi both atomic cloud image and atomic vagrant image
17:27:40 <roshi> vagrant and the docker host? (which is atomic, right?)
17:27:41 <agrimm> dgilmore, I've actually been building the cloud image with the exact same kickstart from spin-kickstarts, so there's no work for you to consume there
17:27:49 <agrimm> The vagrant kickstart is two lines different in the main kickstart, and ten lines of postinstall shell.
17:27:59 <geppetto> dgilmore: So we can't introduce other trees within a release?
17:28:14 <dgilmore> geppetto: no
17:28:17 <mattdm> can we tweak the composition of the tree? (i hope?)
17:28:30 <dgilmore> mattdm: tweak how?
17:28:46 <geppetto> I guess add packages?
17:29:01 <geppetto> I mean, we'd kind of have to … if they come in via. deps.
17:29:29 <jzb> or remove
17:29:37 <mattdm> dgilmore change the json file if need be
17:29:38 <jzb> IIRC we may have a request coming to remove geard
17:29:57 <number80> jzb: related to kubernetes ?
17:29:58 <dgilmore> jzb: we pull the git repo every time we compose
17:30:15 <dgilmore> geard is gone from fedora and it wont be installable anymore
17:31:02 <jzb> number80: the state of geard is best described as "in flux"
17:31:13 <jzb> dgilmore: solves that then
17:31:27 <number80> ack
17:33:34 <mattdm> agrimm -- so, you're saying that what you're using is the same as in spin-kickstarts already, right? any additional tweaks needed?
17:34:10 <mattdm> if not, and I'm understanding correctly, this means that the test compose atomic images should be good to actually test, right?
17:34:22 <agrimm> mattdm, nope, doesn't look like it.
17:34:44 <mattdm> wait was that the good kind of "nope" or the bad one? :)
17:34:48 <dgilmore> I got a tc6 atomic tree installed today
17:34:59 <roshi> speaking of testing, has anyone given any thought to how to test/validate the atomic images?
17:35:32 * mattdm hopes we can pull in some of the wider atomic community to help with that
17:35:36 <agrimm> mattdm, that's "nope, no additional tweaks needed." :)
17:35:42 <mattdm> agrimm: \o/
17:35:48 <jzb> dgilmore: is that grabbable?
17:36:03 <dgilmore> jzb: from koji yes
17:36:05 <number80> awesome :)
17:36:08 <agrimm> roshi, there's going to be about 75% overlap with regular cloud images
17:36:21 <dgilmore> http://koji.fedoraproject.org/koji/taskinfo?taskID=7531371
17:36:37 <dgilmore> the cloud install tree is a mess and naeeds major work
17:36:44 <roshi> fair enough, that's still 25% more testing at least though :)
17:37:31 <mattdm> #info Fedora Atomic TC6 image (should be, you know... testable!) http://koji.fedoraproject.org/koji/taskinfo?taskID=7531374
17:37:34 <agrimm> roshi, which is basically 1) test docker, and 2) test ostree updates and rollbacks
17:37:44 <jzb> woot!
17:37:53 <jzb> agrimm: so we can do 1/2 with this
17:38:04 <mattdm> dgilmore "major work" to make it not ugly or major work to make it function?
17:38:05 <jzb> we'd need a new compose / tree to do update/rollback
17:38:10 * agrimm should have let dgilmore speak first on this topic, it seems
17:38:20 <dgilmore> mattdm: major work to make it function
17:38:27 <imcleod> dgilmore: That tree is defined by this file in spin-kickstarts, correct?  -> https://git.fedorahosted.org/cgit/spin-kickstarts.git/tree/fedora-install-cloud.ks
17:38:34 <dgilmore> imcleod: yes
17:38:44 <imcleod> dgilmore: can you briefly elaborate on the "mess"?
17:38:47 <dgilmore> many packages the cloud install needs are not there
17:38:52 <imcleod> OK.  That's a mess.
17:38:59 <dgilmore> and tehre is no environments in the tree at all
17:39:19 <dgilmore> there should be a cloud-product-environment and the cloud images should use that
17:39:30 <dgilmore> so there is a bunch of comps changes needed
17:39:39 <dgilmore> and likely some kickstart changes needed
17:40:51 <imcleod> OK.  I made a very tentative proposed comps change quite a while ago but then kind of lost the thread.  I can try to pick that up but will need a few pointers to details on the environments setup.  I can take this as an action if I can get a few moments of your time offline after the meeting.
17:40:52 <mattdm> dgilmore is there a ticket for this?
17:41:03 <mattdm> imcleod++
17:41:14 * mattdm is as happy with that as with a ticket :)
17:41:20 <dgilmore> mattdm: no idea, its the same issue basically as the server product had
17:41:30 <dgilmore> but we are much further behind
17:42:30 <imcleod> dgilmore: Is it fair to summarize this as "generate comps, spin kickstart and environments changes as needed to get the required cloud packages into the Cloud spin?"
17:42:45 <dgilmore> mattdm: there is at least one bug in bugzilla
17:43:00 <dgilmore> the packages missing probably doesnt have one yet
17:43:16 <dgilmore> but it will need one filed so we can block alpha on getting it fixed
17:43:25 <dgilmore> imcleod: yes
17:43:40 <dgilmore> imcleod: the environments bit is in comps
17:44:32 <imcleod> dgilmore: OK.  I'll see if I can figure it out.
17:45:16 <dgilmore> imcleod: looking at what the server guys have done this week should help
17:45:23 <mattdm> #action imcleod to generate comps, spin kickstart and environments changes as needed to get the required cloud packages into the Cloud spin
17:45:28 <mattdm> ^^^^ :)
17:45:36 <dgilmore> any groups needed for the cloud-product-environment need to be in the install kickstart
17:46:26 <mattdm> for cloud currently, that's @core and a few random packages
17:47:05 <mattdm> as I understand it, the idea here is to put those random packages into a cloud-product group instead
17:47:12 <dgilmore> mattdm: those random packages need to go in a comps group
17:48:18 <mattdm> okay sounds like the problem is at least well understood
17:48:23 <mattdm> do we have anything else like that?
17:49:12 <dgilmore> just needs done :)
17:49:36 <mattdm> okay, and we've got imcleod tagged for it. :)
17:49:46 <mattdm> any other things we know need done?
17:49:46 <dgilmore> imcleod: when you have thinsg together I am happy to review and make sure they get applied
17:49:51 <dgilmore> I can do some testing also
17:50:01 <jzb> imcleod: I will owe you beers
17:50:12 <dgilmore> straigten out the install tree, make the kickstarts match
17:50:14 <imcleod> dgilmore: Cheers.  I cannot commit to spin-kickstarts so you'll be getting patches from me :-).
17:50:23 <dgilmore> setup automated nightly atomic tree composition
17:50:52 <imcleod> dgilmore: I suspect this is, as mentioned above, adding a comps group for the cloud-enabling packages, adding that group to the tree and then including it in the cloud kickstart.  Am I missing anything obvious?
17:50:55 <dgilmore> mattdm: its getting there
17:51:27 <dgilmore> imcleod: thats about it, we should add an environment to comps so people can select it
17:51:53 <imcleod> dgilmore: K.
17:51:54 <dgilmore> and we should think up ways to document and make it easy for people to make custom images
17:52:10 <mattdm> #help documentation for making custom images would be nice
17:53:11 <mattdm> okay number80 next on the agenda? :)
17:54:23 <number80> let's esee
17:55:28 <number80> #topic next steps on test plan
17:55:36 <number80> https://fedorahosted.org/cloud/ticket/61
17:55:42 <number80> roshi: ?
17:56:03 <roshi> well, I put up drafts of the test plan and test overview a while back
17:56:19 <roshi> https://fedoraproject.org/wiki/User:Roshi/QA/Cloud_Docs
17:56:30 <mattdm> #help everyone look at roshi's test plan!
17:56:35 <roshi> never got much feedback on them
17:56:41 <roshi> so I presume they're good
17:57:03 <roshi> I think we've got a decent list of things to test for the base cloud image (AMI)
17:57:15 <roshi> but nothing for the docker container image or the atomic image
17:57:40 <mattdm> roshi yeah have read through them several times and I think they are good
17:57:48 <number80> #help add test cases for Atomic Host image
17:57:53 <mattdm> number80++!
17:58:03 <jzb> roshi: aren't Atomic == docker host same thing?
17:58:12 <roshi> docker host == atomic aiui
17:58:23 <number80> (as for the docker image, it's now under the umbrella of the Base WG)
17:58:31 <jzb> oh, gotcha. Misread.
17:58:37 <roshi> but we're also talking docker container images (for running inside a container) and talks of the vagrant image too
17:58:45 <roshi> none of which are blocker as far as I know
17:58:56 <roshi> but should still be testing if we're going to say "We made this."
17:59:15 <roshi> number80: docker image being the docker container image or the atomic image?
17:59:19 <jzb> somewhere in beta timeframe we should probably try a test day w/the base folks
17:59:27 <jzb> running the docker image in Atomic image
17:59:30 <mattdm> good idea jzb
17:59:34 <number80> roshi: container one
17:59:50 <dustymabe> jzb: yeah would be great to have an atomic/docker test day
17:59:51 * roshi doesn't remember who the QA contact is in base
18:00:03 <mattdm> roshi looks like the docker image (the one that runs inside the container, not the host) is outside of cloud wg now, so we should pass that on
18:00:21 <number80> roshi: just remember that Atomic => container hosting platform, docker image => the one used by containers
18:00:26 <mattdm> #help we hould plan a test day with the base wg for running the docker base image in atomic
18:00:34 <roshi> nothing to pass on because I haven't written anything for testing it :)
18:00:40 <mattdm> roshi yay!
18:00:43 <mattdm> :)
18:00:53 <mattdm> and the vagrant image is f22 material so nothing to worry about now
18:00:59 <number80> jzb: do you know if there's a test day for Atomic image ?
18:01:03 <dustymabe> mattdm: dgilmore and a few others of us talked about terminology earlier for the "docker image"
18:01:03 <roshi> yeah number80 - just took a while to suss out due to either misunderstanding on my part or equivocation about the image names
18:01:06 <jzb> number80: not yet
18:01:09 <roshi> there isn't now, a test day
18:01:24 <roshi> it would be good to have one
18:01:25 <jzb> #action jzb plan test days for Atomic image, base docker image
18:01:29 <dustymabe> mattdm: would probably be less confusing if we started saying "docker container image"
18:01:43 <number80> jzb: thanks :)
18:01:46 <jzb> well, we could just say "docker image"
18:01:56 <jzb> and "atomic"
18:02:04 <roshi> the biggest issue I talked to mattdm about was documented in the "Ideas on what to test" page on my wiki space
18:02:06 <dustymabe> jzb: seems like enough people are getting confused
18:02:23 <roshi> I did my best to document things we should test since there wasn't really a list to begin with
18:02:27 <dustymabe> jzb: it makes sense to me and you but most people assume 'docker image' can be run as a VM
18:02:37 <jzb> hmm
18:02:38 <jzb> do they?
18:02:50 <mattdm> To add another wrinkle of confusion, we also want to distinguish that this is a docker _base_ image (in docker's terminology).
18:02:50 <roshi> I'm writing a blog post explaining the different images we're making - once I have it fully grokked :)
18:02:51 <dustymabe> seems like it keeps coming up in chats
18:02:53 * dgilmore needs to go seek out food
18:02:54 <jzb> dustymabe: I'm perpetually confused so maybe I just hadn't noticed when others were ;-)
18:03:08 <number80> maybe we could discuss this on the list and have a lexicon on the wiki ?
18:03:12 <roshi> could perhaps be good on the magazine as well
18:03:28 <mattdm> number80: good call. dustymabe can you start a thread on the cloud list?
18:03:33 <roshi> I'll also put it on the wiki - since I feel like there's some confusion for sure
18:03:34 <jzb> ISTR someone is working on a clarification piece.
18:03:41 <dustymabe> mattdm: regarding the terminology ?
18:03:45 <mattdm> dustymabe yeah.
18:03:50 <dustymabe> mattdm: sure
18:04:14 <number80> #action dustymabe start a discussion on terminology (docker/atomic) on the cloud m-l
18:04:15 <mattdm> that way we all know how to properly refer to the bikeshed. :)
18:04:34 <roshi> bicyclic housing unit :p
18:04:34 <number80> well, next topic ?
18:04:39 <number80> lol
18:04:44 <jzb> roshi: +1
18:05:11 <roshi> any comments on the test plan and what to test we can talk in fedora-cloud after the meeting
18:05:23 <dustymabe> mattdm: should I include the base WG on that email since they are responsible for the docker container image?
18:05:23 <roshi> if that works
18:05:32 <roshi> +1 dustymabe
18:05:45 <number80> +1
18:05:57 * dustymabe takes the hint :)
18:06:03 <number80> ok
18:06:14 <number80> #topic Automatic Smoketests on Image Build
18:06:22 <number80> https://fedorahosted.org/cloud/ticket/38
18:06:33 <number80> agrimm, roshi ?
18:06:52 <roshi> we talked a bit about this earlier this week
18:07:12 <agrimm> hmm, we talked about two tasks last friday, and I think I put my whole update in one of them...
18:07:17 * agrimm looks for the other
18:07:38 <mattdm> also i know oddshocks's fedimg service _does_ do a basic boot/ssh test, but I don't know the details
18:07:46 <roshi> for basic smoke tests I was thinking fabric based ssh tests
18:08:15 <roshi> and I think I can get those to run in taskotron, but would need to talk to tflink about it more down the line
18:08:18 <number80> works for me
18:08:29 <roshi> https://fedoraproject.org/wiki/User:Roshi/QA/Cloud_Docs/what_to_test
18:09:28 <roshi> I think most of those things listed there can be turned into automated tests via fabric
18:09:36 <oddshocks> mattdm: yeah, right now I just have it configured to run /bin/true and check the exit code, but fedimg will run an entire shell script and check the exit code
18:09:47 <oddshocks> it's just I haven't put anything in that script besides the basic /bin/true by default
18:09:53 <roshi> I also made some test cases for doing manually
18:10:14 <roshi> at the bottom of the link I just gave
18:12:22 <mattdm> oddshocks do we have somewhere easy (or an easy command to run) so mere mortals can see the state of those tests?
18:13:03 * dustymabe wonders if the FPL is a mere mortal?
18:13:13 <oddshocks> mattdm: sure, there's an image.test fedmsg topic :)
18:13:21 <jzb> dustymabe: no, but there can be only one.
18:13:21 <oddshocks> mattdm: nothing to see at the momennt though, still smoketesting in staging
18:13:31 <oddshocks> mattdm: but yeah you can see if tests have started, completed, or failed
18:13:32 <number80> dustymabe: he needs to eat and sleep at least ;)
18:13:38 <dustymabe> :)
18:13:48 <jzb> number80: I've never seen him sleep.
18:14:00 <oddshocks> mattdm: and of course there's more I can add in the future
18:14:05 <number80> lol
18:14:05 <roshi> I think this discussion would be good to have with the qa-devel group
18:14:22 <roshi> get oddshocks and tflink, kparal all on the same sheet of music
18:14:40 <roshi> since we would want the results with the rest of the results throughout fedora and on resultsdb
18:14:57 <number80> sounds good
18:15:20 * roshi isn't sure when the next qa-devel meeting is
18:15:56 <roshi> for the time being though, we can do manual tests and automagic tests as we have them and report in the wiki that the images are valid once we have images
18:16:20 <number80> roshi: is it the same as the QA meeting ?
18:16:35 * oddshocks nods
18:16:35 <roshi> nope, it's the devel team for taskotron
18:16:43 <number80> ok
18:16:56 * oddshocks apologizes for being semi-unresponsive, on a call
18:17:05 <roshi> I don't see the automagic testing as being a huge deal right now - since we can do these tests manually now
18:17:09 <roshi> we just need to document them
18:17:14 <number80> roshi: it doesn't appear on fedocal, please request that they add it :)
18:17:25 <roshi> for atomic and the docker bace image tests especially
18:17:32 <roshi> might not be one scheduled
18:17:42 <mattdm> #info for the time being we can do manual tests and automagic tests as we have them, and report in the wiki as usual
18:17:43 <roshi> but I'll check in with them
18:18:09 <roshi> I'll be making sure the wiki pages have the tests added to the results page
18:18:19 <roshi> talk to adamw and the qa folks about placement and whatnot
18:19:25 <mattdm> roshi thanks!
18:19:36 <oddshocks> roshi: I'm going to be doing a docs binge this week and there will be a lot more info on how fedimg tests, at least
18:19:47 <number80> good
18:19:57 <roshi> sweet
18:20:04 <roshi> we can link up on that sometime too
18:21:30 <number80> next topic ?
18:22:07 <mattdm> yep
18:22:09 <number80> #topic start communication/collaboration on cloud image updates
18:22:16 <number80> https://fedorahosted.org/cloud/ticket/51
18:22:24 <number80> roshi: again ^^
18:22:51 <roshi> this is a lot of stuff for not even being a proper cloudy person :P
18:23:06 <number80> sure :)
18:23:15 <roshi> this was the other thing agrimm and I talked about
18:23:39 <mattdm> roshi I think you're an *honorary* proper cloudy person by now :)
18:23:44 <roshi> haha
18:23:51 <mattdm> #info agrimm has some good info in the ticket
18:23:52 <roshi> something like that for sure :)
18:24:28 <roshi> I think some discussion should probably happen about where the testing actually *happens* - but that process listed in the ticket is good afaict
18:25:13 <roshi> suggestions on the list of times a update should be created would be good
18:25:20 <roshi> those were what we could come up with
18:26:36 <mattdm> #help more suggestions for cases where an update should be created would be helpful
18:27:08 <roshi> as for the where to run the tests in that process, I think the conversation with qa-devel will be important
18:27:25 <roshi> since that's where all the other testing is going to be moving towards, aiui
18:28:38 <mattdm> +1
18:29:35 <mattdm> do we have any other action to take on this one at this point?
18:29:59 * roshi doesn't have anything he doesn't think
18:30:09 <roshi> more discussion within the ticket should be good
18:30:09 <number80> ok, last point :)
18:30:52 <mattdm> #help more discussion in the ticket please
18:31:32 <number80> #topic Process for determining when and why Docker trusted images need to be rebuilt
18:31:37 <number80> https://fedorahosted.org/cloud/ticket/59
18:31:52 <roshi> this process might be almost covered by the last, right?
18:32:03 <jzb> almost yeah
18:32:07 <number80> yup
18:32:08 <jzb> I've sent a draft to the list for discussion
18:32:15 <jzb> and ... crickets
18:32:26 <jzb> well, sorta - I got a bunch of questions from one person directly
18:32:37 <jzb> asked if it was OK to send ot list, but didn't get a reply
18:32:49 <jzb> but it'd be good to get some feedback from the rest of the WG
18:32:57 * oddshocks nods
18:33:07 <roshi> link to that email?
18:33:08 <dustymabe> jzb: I think from what I read you had said once a month?
18:33:08 <number80> jzb: it should be ok
18:33:26 <jzb> dustymabe: short version: montly or if a security update triggers it
18:33:26 * roshi didn't respond because up until this week I didn't really grok docker at all
18:33:55 <dustymabe> jzb yep. sounds good to me
18:34:25 <jzb> OK
18:34:43 <jzb> lemme go back through and see if any of the other qs need to go to the list and I'll close this out.
18:34:46 <jzb> EOF
18:34:59 <number80> do we skip the open floor ?
18:35:18 <number80> or do you have any other topics ?
18:35:20 <dustymabe> number80: any items from last week that got lost?
18:35:37 <dustymabe> https://lists.fedoraproject.org/pipermail/cloud/2014-September/004194.html
18:35:43 <number80> dustymabe: there is #65 => https://fedorahosted.org/cloud/ticket/65
18:35:51 * roshi has nothing since we alreday talked about the bicyclic housing unit that is naming images in a less confusing way
18:36:05 <number80> but since we're not in charge anymore of the docker image, it may be ok to skip it
18:36:36 <number80> I suggest that we close it ?
18:36:57 <roshi> so the concrete list of the images we want to produce (product and others) is: Base (AMI, etc), Atomic, Docker (handled by Base), right?
18:37:03 <roshi> +1 number80
18:37:12 <dustymabe> cool
18:37:30 <number80> roshi: more or less, yes
18:37:39 <mattdm> number80 I think close for now. this is about a second thing -- images produced _at Docker_ via stuff in fedora-dockerfiles. (eg not releng or "official fedora" deliverables)
18:37:50 <roshi> what's the answer that doesn't require the "more or less" modifier?
18:37:59 <number80> #action number80 close ticket #65
18:38:03 <roshi> :)
18:38:08 <number80> roshi: we have few variants of base ;)
18:38:15 <number80> probably not for F21
18:38:23 <roshi> ok, for F21 then
18:38:34 <roshi> I just haven't seen The List for F21 (tm)
18:38:39 <mattdm> yeah let's just worry about f21 for now :)
18:38:48 <number80> ok
18:38:53 <number80> #topic openfloor
18:39:04 <number80> last call before we end this meeting
18:39:18 <oddshocks> Just a reminder that there's a weekly Atomic meeting in #atomic on Tuesdays
18:40:00 <roshi> so those three are The List for F21?
18:40:03 <number80> oddshocks: the atomic/infra meeting ?
18:40:03 <oddshocks> 20:00 UTC. that's all I got
18:40:34 <oddshocks> number80: infra meeting is Thursdays. but a few infra people do go to the #atomic meeting
18:40:38 <dustymabe> number80: https://apps.fedoraproject.org/calendar/cloud/
18:40:40 <number80> #info Weekly Atomic meeting in #atomic on tuesdays 20:00 UTC
18:40:44 <oddshocks> https://apps.fedoraproject.org/calendar//meeting/738/
18:41:27 <number80> ok
18:41:32 <number80> good
18:41:47 <number80> anything else ?
18:41:57 <jzb> Nope
18:42:04 <mattdm> thanks number80!
18:42:16 <number80> thank you guys for attending and see you next week ! :)
18:42:18 <oddshocks> yeah, thanks :)
18:42:21 <jzb> indeed, thanks number80 !
18:42:25 <roshi> thanks number80 :)
18:42:25 <number80> #endmeeting