15:00:12 <jednorozec> #startmeeting RELENG (2022-04-05)
15:00:13 <zodbot> Meeting started Tue Apr  5 15:00:12 2022 UTC.
15:00:13 <zodbot> This meeting is logged and archived in a public location.
15:00:13 <zodbot> The chair is jednorozec. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:00:13 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:13 <zodbot> The meeting name has been set to 'releng_(2022-04-05)'
15:00:13 <jednorozec> #meetingname releng
15:00:13 <jednorozec> #chair nirik sharkcz pbrobinson pingou phsmoura mboddu dustymabe ksinny jednorozec
15:00:13 <jednorozec> #topic init process
15:00:13 <zodbot> The meeting name has been set to 'releng'
15:00:13 <zodbot> Current chairs: dustymabe jednorozec ksinny mboddu nirik pbrobinson phsmoura pingou sharkcz
15:00:20 <nirik> morning
15:01:36 <jednorozec> morning
15:02:33 <jednorozec> #topic Final Freeze
15:02:48 <jednorozec> today we entered final freeze in Fedora 36 release cycle
15:03:23 <jednorozec> I have noticed that in here https://fedorapeople.org/groups/schedule/f-36/f-36-releng-tasks.html
15:03:39 <jednorozec> there is a task: Set current release to active in Package database
15:04:14 <jednorozec> is this leftover or is releng supposed to do something? should it be bodhi instead of package database?
15:04:16 <nirik> thats wrong
15:04:26 <nirik> thats done after GA...
15:04:45 <nirik> it's the thing that gnome-software looks at to decide if it should offer people the upgrade.
15:04:52 <jednorozec> ahhh
15:04:55 <jednorozec> I see
15:05:02 <nirik> so, it needs to only be done after release
15:05:29 <jednorozec> right, is there some template that this is generated from?
15:05:33 <jednorozec> I can fix it
15:05:45 <jednorozec> but not sure where
15:06:25 <jednorozec> ha got it https://pagure.io/fedora-pgm/schedule/
15:06:27 <mboddu> jednorozec: bcotton has it
15:06:36 <nirik> yeah, file an issue there.
15:06:41 <jednorozec> yup
15:08:58 <jednorozec> done
15:10:14 <jednorozec> ok that should be all for the freeze
15:10:36 <jednorozec> anything else regarding the current release cycle?
15:11:14 <nirik> just the usual I guess... ;)
15:11:23 <nirik> pushing updates, making rc's when ready, etc
15:11:26 <jednorozec> monig on
15:11:37 <jednorozec> related to the release
15:11:40 <jednorozec> .releng 10726
15:11:41 <zodbot> jednorozec: Issue #10726: Bodhi updates for Fedora 36 have wrong karma limits - releng - Pagure.io - https://pagure.io/releng/issue/10726
15:11:59 <jednorozec> SO it seems we are doing the days to stable in bodhi wrong
15:12:51 <nirik> well, it's kinda nitpicky... if we are frozen, does it matter ? but sure...
15:12:51 <jednorozec> here is the policy https://docs.fedoraproject.org/en-US/fesco/Updates_Policy/#karma-requirements
15:13:03 <jednorozec> here is our config https://pagure.io/fedora-infra/ansible/blob/main/f/roles/bodhi2/base/templates/production.ini.j2#_585
15:13:21 <nirik> so, we should change it after freeze starts, not when freeze ends right?
15:13:23 <mboddu> When we unfreeze from beta, we should also update it
15:13:47 <mboddu> While we are in freeze it doesn't matter since we dont push updates
15:14:03 <nirik> right, but thats what we are doing wrong, right?
15:14:13 <mboddu> Yeah, as per policy
15:14:20 <nirik> it will mean possibly less updates going to stable after we unfreeze
15:14:42 <mboddu> Either fix policy or change the bodhi config when freeze starts
15:15:11 <mboddu> Or ignore both and just update the config when we unfreeze
15:15:15 <nirik> I'd say just change it at freeze start... just easier. ;)
15:15:22 <jednorozec> yeah
15:15:24 <mboddu> Yeah, agreed :)
15:15:49 * jednorozec writing comment to the ticket
15:19:07 <jednorozec> done
15:19:09 <jednorozec> moving on
15:20:04 <jednorozec> .releng 10713
15:20:07 <zodbot> jednorozec: Issue #10713: Add ELN to the set of packages to be re-signed at each Branch - releng - Pagure.io - https://pagure.io/releng/issue/10713
15:20:13 <jednorozec> so eln has modules?
15:20:14 <jednorozec> :D
15:20:45 <nirik> yeah, thats what I said too! :)
15:21:29 <jednorozec> so for the docs changes, we should always add rawhide key when branching. is that right?
15:23:50 <nirik> well, 1 week or so before branching we should resign rawhide with the new key... and I guess eln also then...
15:24:03 <nirik> so when we branch we can make a rawhide with the new key
15:24:28 <mboddu> eln modules is tricky
15:25:02 <jednorozec> ok so I will create new issue with the schedule repo, to add this one week before branching.
15:25:19 <mboddu> As there is no sign-module in sigul, we have to use robosig to module signing
15:25:21 <jednorozec> our SOP needs some tweaks as-well
15:25:48 <nirik> I already got bcotton to add it for f37+. ;)
15:25:53 <nirik> but I didn't do the sop end
15:25:54 <jednorozec> weee
15:25:55 <mboddu> And we need to find all the latest NS and sign them rather than signing all of the modules in eln
15:26:05 <jednorozec> nirik, I will fix the SOP
15:26:28 <jednorozec> mboddu, is this somewhere documented?
15:26:44 <mboddu> jednorozec: Nope, I just figured out the problem yesterday
15:28:19 * mboddu bbiab
15:28:32 * jednorozec is looking at sigul
15:28:49 <jednorozec> maybe adding method to sign modules there will be the best solution
15:31:26 <nirik> we probibly need to figure out what to do with sigul long term...
15:31:57 <jednorozec> yeah there is bunch of opened issues
15:32:05 <jednorozec> and few pullrequests
15:33:27 <nirik> and it needs to be re-written to not use nss
15:33:47 <nirik> but thats a bigger topic I think than we have today. ;)
15:33:57 <jednorozec> yeah
15:34:17 <jednorozec> I am actually thinking about looking into it little bit more
15:35:52 <jednorozec> lets move on
15:36:21 <nirik> I wonder if we couldn't look into sigstore.
15:36:30 <jednorozec> link?
15:36:33 * mboddu stays away from sigul unless a re-write is proposed/accepted :)
15:37:18 <nirik> https://www.sigstore.dev it may not do what we need...
15:38:17 <jednorozec> hmm
15:38:45 <jednorozec> hm
15:38:53 <jednorozec> the page has RH logo on it
15:39:06 <nirik> yeah...
15:39:09 <jednorozec> so it might be used somewhere internaly
15:39:20 <nirik> it's more for the source side of things, but perhaps we could leverage it.
15:39:23 <nirik> just a thought.
15:39:35 <jednorozec> yeah, we should explore it
15:40:27 <mboddu> And purdue university :)
15:42:11 <jednorozec> I have nothig else right now.
15:42:30 <jednorozec> So rafine old tickets? Or has anyone of you something to discuss?
15:43:03 <mboddu> Nothing from me
15:43:13 <nirik> nothing I can think of off hand...
15:44:50 <nirik> we could hunt for tickets to close?
15:44:57 <jednorozec> yeah
15:45:11 * jednorozec puts on hunting hat
15:45:51 <nirik> should we close the f36 changes ones? or we do that after release?
15:46:41 <nirik> .releng 10262
15:46:42 <zodbot> nirik: Issue #10262: koji armv7hl F35 libreoffice build fails in %prep - releng - Pagure.io - https://pagure.io/releng/issue/10262
15:46:50 <nirik> I think this is solved now... can be closed?
15:47:08 <jednorozec> +1
15:47:41 <jednorozec> .releng 8579
15:47:42 <zodbot> jednorozec: Issue #8579: Clean out accidentally committed tarballs in rpms/tlog - releng - Pagure.io - https://pagure.io/releng/issue/8579
15:47:47 <nirik> .releng 10383
15:47:55 <zodbot> nirik: Issue #10383: possibly missing SRPMS link in updates tree? - releng - Pagure.io - https://pagure.io/releng/issue/10383
15:48:03 <nirik> I think thats done ^
15:48:47 <jednorozec> yeah
15:48:55 <nirik> did that tlog tar get removed?
15:49:11 <jednorozec> i dont see it there
15:50:32 <nirik> cool.
15:51:28 <nirik> .releng 10401
15:51:29 <zodbot> nirik: Issue #10401: Unable to build datovka on s390x builders, probably out of filesystem space - releng - Pagure.io - https://pagure.io/releng/issue/10401
15:51:39 <nirik> I think this is done. At least all recent builds completed.
15:52:03 <jednorozec> huh I should get datovka :)
15:52:30 <jednorozec> nirik, close it
15:52:40 * nirik looks at what it is.
15:52:51 <jednorozec> heh czech e goverment thing
15:52:54 <nirik> oh, some cz gov thing?
15:53:02 <jednorozec> it your e post box
15:53:14 <jednorozec> so you get taxes and documents in electronic way
15:53:24 <jednorozec> there was no other than IE client for ages
15:53:33 <jednorozec> now it is multiplatform :)
15:53:57 <nirik> huh, cool.
15:54:38 <nirik> so we are low on time now... what was the decision on the f36 changes? keep them until after release? or can we close them now?
15:55:12 <jednorozec> So I would check what actually got in
15:55:19 <jednorozec> some of them may be defered to f37
15:55:58 <nirik> ok
15:56:48 <jednorozec> I will check them and close all that landed until next meeting
15:57:02 <nirik> oh, just saw https://bugzilla.redhat.com/show_bug.cgi?id=2071913
15:57:16 <nirik> whats the status of new openh624 repos? they were sent a while back?
15:58:17 <mboddu> I need to run it by jednorozec but f36 beta came in between and then my HR mumbo jumbo happened
15:58:22 <mboddu> That was on me, sorry
15:58:39 <mboddu> jednorozec: Are you free tomorrow? And can you throw in a meeting invite?
15:58:48 <jednorozec> mboddu, ack
15:59:29 <nirik> we did get f37 builds last time too right?
16:00:31 <mboddu> Yup, they were updated
16:01:21 <jednorozec> mboddu, send I may be late few minutes will update you tmrw
16:01:34 <jednorozec> and we are out of time
16:01:40 <jednorozec> #endmeeting