16:30:27 #startmeeting fedora_coreos_meeting 16:30:27 Meeting started Wed Feb 27 16:30:27 2019 UTC. 16:30:27 This meeting is logged and archived in a public location. 16:30:27 The chair is rfairley. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:30:27 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:30:27 The meeting name has been set to 'fedora_coreos_meeting' 16:30:35 #topic roll call 16:30:50 .hello2 16:30:50 .hello2 16:30:50 dustymabe: dustymabe 'Dusty Mabe' 16:30:51 .hello2 16:30:53 rfairley: rfairley 'None' 16:30:56 ajeddeloh: ajeddeloh 'Andrew Jeddeloh' 16:31:08 .hello2 16:31:09 dbenoit: dbenoit 'David Benoit' 16:31:09 .hello2 16:31:11 yzhang: yzhang 'Yu Qi Zhang' 16:31:52 #chair dustymabe ajeddeloh dbenoit yzhang 16:31:52 Current chairs: ajeddeloh dbenoit dustymabe rfairley yzhang 16:32:02 .hello2 16:32:03 sanja: sanja 'Sanja Bonic' 16:32:54 .hello2 16:32:55 jlebon: jlebon 'None' 16:33:03 give another min for people to join :) 16:33:14 .hello sinnykumari 16:33:15 #chair sanja jlebon 16:33:15 Current chairs: ajeddeloh dbenoit dustymabe jlebon rfairley sanja yzhang 16:33:15 ksinny: sinnykumari 'Sinny Kumari' 16:33:18 .hello mnguyen 16:33:19 mnguyen_: mnguyen 'Michael Nguyen' 16:33:42 .hello2 16:33:43 slowrie: slowrie 'Stephen Lowrie' 16:34:02 .hello2 16:34:03 lorbus: lorbus 'Christian Glombek' 16:34:47 #chair ksinny mnguyen_ slowrie lorbus 16:34:47 Current chairs: ajeddeloh dbenoit dustymabe jlebon ksinny lorbus mnguyen_ rfairley sanja slowrie yzhang 16:35:14 o/ hi all! 16:35:36 * ajeddeloh waves 16:35:42 #topic Action items from last meeting 16:35:48 o/ 16:36:02 Action Items, by person 16:36:05 ----------------------- 16:36:09 * ajeddeloh 16:36:14 * ajeddeloh to open a few tickets and design doc PRs for discussed 16:36:17 open floor topics on philosophy 16:36:18 .hello2 16:36:19 miabbott: miabbott 'Micah Abbott' 16:36:40 * rfairley needs to find a better way to paste these 16:37:15 ajeddeloh did these get opened? 16:37:18 yeah 16:37:20 just looking in the tracker 16:37:21 https://github.com/coreos/fedora-coreos-tracker/issues/153 16:37:46 Haven't PRd the design doc yet since there hasn't been much discussion on the tracker 16:38:23 # link https://github.com/coreos/fedora-coreos-tracker/issues/153 16:38:31 #link https://github.com/coreos/fedora-coreos-tracker/issues/153 16:38:45 .fas jasonbrooks 16:38:47 jbrooks: jasonbrooks 'Jason Brooks' 16:39:03 #info To discuss more in linked ticket on Including fedora specific behavior 16:39:11 #chair miabbott jbrooks 16:39:11 Current chairs: ajeddeloh dbenoit dustymabe jbrooks jlebon ksinny lorbus miabbott mnguyen_ rfairley sanja slowrie yzhang 16:39:34 good to move to next? 16:39:44 next action item^ 16:40:41 we can discuss more in the linked ticket, then maybe discuss that at the meeting next week 16:40:52 on fedora specific behavior^ 16:41:00 * dustymabe 16:41:01 * dustymabe to create tickets for platform artifact creation 16:41:39 #info dustymabe created tickets #146->#151 for image artifact uploads 16:41:44 that's all 16:42:16 +1 16:42:39 * kaeso 16:42:42 * kaeso to create ticket for node arch-label introspection 16:43:24 kaeso - around? 16:43:36 * rfairley looking in tracker 16:43:42 https://github.com/coreos/fedora-coreos-tracker/issues/154 16:43:53 He might be in PTO 16:43:56 on* 16:43:57 yeah he is out 16:44:06 i think it could be closed now with https://github.com/projectatomic/rpm-ostree/pull/1766, though let's let kaeso take a look 16:44:13 #info kaeso created labeling supported architectures and runtime introspection ticket 16:44:20 #link https://github.com/coreos/fedora-coreos-tracker/issues/154 16:44:31 jlebon: +1 16:45:07 let's move to meeting topics 16:45:21 https://public.etherpad-mozilla.org/p/20190227-FCOS-Meeting 16:45:40 if anyone has any open-floor topics for today, please feel free to add in the link ^ 16:45:59 #topic Keep/Remove Python dependent package: policycoreutils-python 16:46:08 #link https://github.com/coreos/fedora-coreos-tracker/issues/126 16:47:06 We asked Dan on the ticket and he seems positive towards not having policycoreutils-python in FCOS 16:47:20 +1 for trying it 16:47:29 we can use semodule, setsebool to do some selnux policy changes 16:48:23 does anyone think we might be still missing something if we don't ahave policycoreutils-python in FCOS base? 16:49:51 walters: jlebon: thoughts? 16:50:04 i'm good with shipping without it to start 16:50:10 ksinny: i think if dan agrees we can go ahead and experiment 16:50:20 is there another tool that we might need to write for changing file contexts? https://github.com/coreos/fedora-coreos-tracker/issues/126#issuecomment-466185962 16:51:18 chcon should still be there. for modifying patterns, there's /bin/vi :) 16:51:30 I'm +1 to not shipping it, we can always add things later, but not remove them 16:51:31 I think I only use chcon 16:51:56 UX doesn't need to be good for uncommon actions, it just needs to be possible 16:52:11 yeah, I used restorecon and chcon sometimes. That's it 16:53:07 i think we're all agreed 16:53:09 +1 - seems we are in agreement on shipping without policycoreutils-python to start 16:53:35 +1 16:53:50 \o/ 16:53:55 #info FCOS will not have policycoreutils-python in base packages 16:54:30 #agreed We will experimint with initially not shipping policycoreutils-python-utils in the FCOS base system 16:54:53 thanks ksinny :D 16:55:04 :) 16:55:19 I think we can head over to open floor topics now 16:55:35 yeah 16:55:51 #topic Open Floor 16:56:24 #topic tested out bare metal installer in PXE boot setup - seems to be working for BIOS (legacy boot) installs 16:56:31 #link https://github.com/coreos/fedora-coreos-tracker/issues/91#issuecomment-467631190 16:56:42 woo! 16:56:59 dustymabe: want to start on this? 16:57:12 dustymabe++ 16:57:29 ^ update on this for the discussion here I mean 16:57:57 yep. basically i verified PXE installs can happen with bare metal legacy boot using the artifacts from the ISO we are generating 16:57:57 dustymabe++ 16:58:18 #info FCOS PXE installs can happen with bare metal legacy boot using the artifacts from the ISO we are generating 16:59:08 that's all I had 16:59:09 nice! 16:59:11 dustymabe++ 16:59:17 dustymabe++ 16:59:17 yzhang: Karma for dustymabe changed to 10 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 17:00:00 good that bare metal is coming together more and more 17:00:20 * rfairley to next topic 17:00:29 #topic ignition spec 3.0.0 progress 17:00:45 It's coming along nicely I think 17:01:02 not going to quite make the "end of feb" goal we set 17:01:07 #link https://github.com/coreos/ignition/milestone/5 17:01:11 but not _too_ far off 17:01:27 ajeddeloh++ 17:01:28 ^these are quickly going through in the milestone 17:01:29 ajeddeloh++ 17:01:29 ksinny: Karma for ajeddeloh changed to 3 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 17:01:33 i'm actively playing with integrating it with FCOS 17:01:40 ajeddeloh++ 17:01:40 rfairley: Karma for ajeddeloh changed to 4 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 17:01:40 jlebon++ 17:01:40 ajeddeloh: what are the largest work items we have left? 17:01:43 ajeddeloh: Karma for jlebon changed to 6 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 17:01:43 ajeddeloh what major work for 3.0.0 is still outstanding? 17:02:09 Largest ones are config merging rework and validation to reject duplicates 17:02:28 then a bunch of tiny stuff that shouldn't be too bad 17:02:46 then I'll release Ignition 2.0.0-alpha with spec 3.0.0-experimental 17:03:21 ajeddeloh: of those two largest items are they both on you? or is someone else picking up one of them 17:03:23 another major one is sanity checking /var mounts work well as design 17:03:30 yeah 17:03:44 both on me, they're somewhat intertwined 17:03:57 also both blocked on the automatic translation PR 17:04:08 that's part of the integration stuff i'm doing too. but i don't foresee issues 17:04:16 will be sure to review outstanding PRs! 17:04:37 yeah having people to review the PRs recently has been massively helpful. 17:04:38 I'm planning on doing another pass through all the open PRs sometime this afternoon 17:05:49 ajeddeloh: kk regarding 2.0.0-alpha 17:05:56 is that the planned version number? 17:06:08 n.b. we're skipping 1.0 for dumb golang vendoring related issues 17:06:17 0.x -> 2.x 17:07:28 * dustymabe speculates about going straight to 3.x :) 17:07:35 nononononono 17:07:38 0.x -> 3.x (for spec 3.0.0) 17:07:44 certainly easier to remember 17:07:55 Yeah but they wont be tied together in the future 17:07:58 and shouldn't be 17:08:15 are they in the same repo? 17:08:18 Ignition version itself is for API and user faces changes 17:08:28 spec is for well the spec itself 17:08:29 yeah 17:09:19 Versioning them independently has been hugely helpful, it's a bit of confusion initially but worth it 17:10:04 ehh. you're the experts :) 17:10:18 next topic? 17:10:27 example: if we rework the config/ directory so people vendoring ignition need to call a different api (as we plan to do) but don't change the spec, the change will be invisible to Ignition users but not Ignition as a library consumers 17:11:57 +1 - let's make a note of the 2.0.0-alpha version 17:13:09 #info Ignition API version will be vendored with the Ignition config (from 0.x to 2.x and 3.x) 17:13:46 ^ sounds correct? 17:14:02 no 17:14:10 #undo 17:14:10 Removing item from minutes: INFO by rfairley at 17:13:09 : Ignition API version will be vendored with the Ignition config (from 0.x to 2.x and 3.x) 17:14:11 They will be independently versioned 17:14:59 #info Ignition API and config are versioned independently. Changes to the config version need a change to the API version (since new features were added or changed) but not vise versa 17:15:34 ahh, got it :) 17:15:52 moving to next topic 17:16:11 #topic now producing openstack artifacts in pipeline 17:16:45 #info we are now producing openstack artifacts in the FCOS pipeline 17:16:52 I believe strigazi asked for this 17:17:30 https://github.com/coreos/fedora-coreos-pipeline/pull/50 17:17:39 thanks jlebon for the review 17:18:01 EOM 17:18:23 it's really nice how clean and easy to work on the pipeline is so far 17:19:04 jlebon: yeah. there are a few papercuts, but it's mature for it's age IMHO 17:19:54 yeah, agreed 17:20:28 really cool seeing coreos-assembler, fedora-coreos-pipeline, fedora-coreos-config all come together 17:21:09 #endmeeting