15:00:27 <jednorozec> #startmeeting RELENG (2021-10-12)
15:00:27 <zodbot> Meeting started Tue Oct 12 15:00:27 2021 UTC.
15:00:27 <zodbot> This meeting is logged and archived in a public location.
15:00:27 <zodbot> The chair is jednorozec. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:00:27 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:27 <zodbot> The meeting name has been set to 'releng_(2021-10-12)'
15:00:28 <jednorozec> #meetingname releng
15:00:28 <zodbot> The meeting name has been set to 'releng'
15:00:28 <jednorozec> #chair nirik sharkcz pbrobinson pingou mboddu dustymabe ksinny jednorozec
15:00:28 <jednorozec> #topic init process
15:00:28 <zodbot> Current chairs: dustymabe jednorozec ksinny mboddu nirik pbrobinson pingou sharkcz
15:00:51 <nirik> morning
15:03:26 <mboddu> Morning
15:03:59 * mboddu is sorta here
15:04:11 <jednorozec> but you should be on PTO
15:05:03 <nirik> 235063
15:05:40 <nirik> nice. oops
15:05:56 <nirik> when a kitten chews on your yubikey...
15:06:04 <jednorozec> :)
15:06:34 <jednorozec> so
15:06:46 <jednorozec> I was looking into this
15:06:48 <jednorozec> .releng 10338
15:06:49 <zodbot> jednorozec: Issue #10338: Recently retired packages left in inconsistent state (not blocked from koji) - releng - Pagure.io - https://pagure.io/releng/issue/10338
15:08:41 <nirik> yeah, namespace?
15:08:53 <nirik> but why did it just recently break?
15:08:54 <jednorozec> well
15:09:13 <jednorozec> there is a namespace handling function there
15:09:45 <jednorozec> I need to get more access to the toddlers.
15:10:02 <nirik> they are in openshift
15:10:36 <jednorozec> does it support groups? Or I have to make ansible change for each openshift app I want access to?
15:10:37 <nirik> happy to help go over how to get to them...
15:11:10 <nirik> well, you are in main, so you should have access to them all... but perhaps we didn't add you to cluster admins...
15:12:00 <nirik> yeah we didn't. ;(
15:12:27 <nirik> playbooks/groups/os-cluster.yml at the bottom.... but you can bypass that. Just:
15:12:49 <nirik> sudo -i on batcave01, ssh to os-master01.iad2.fedoraproject.org (or stg) and you can use any oc commands from there.
15:13:13 <jednorozec> I will do the PR and get logs by ssh for now
15:13:40 <jednorozec> But my point was i dont know what is wrong
15:14:08 <jednorozec> It looks ok in the sources. It might be network glitch and we catch exception somewhere
15:14:23 <nirik> ok, yeah, I am not sure either. From what I saw, it asked pdc and pdc said it didn't exist, so it must be doing something wrong.
15:14:58 <jednorozec> Yup and I checked the toddler and it is doing it right...
15:16:17 <jednorozec> I will investigate it more.
15:16:21 <jednorozec> moving on
15:16:24 <jednorozec> .releng 10330
15:16:25 <zodbot> jednorozec: Issue #10330: Add epel9 version to Fedora EPEL product in bugzilla.stage - releng - Pagure.io - https://pagure.io/releng/issue/10330
15:17:05 <jednorozec> the stage bugzilla does not allow me to login with releng credentials,
15:17:09 <mboddu> nirik: We talked about bz and adding me as an admin or something to change things, where do we stand on that?
15:17:44 <nirik> mboddu: I thought we decided you were already in the right group? or am I misremebering?
15:18:41 <nirik> is it bugzilla-stage.redhat.com ? or ?
15:18:42 <mboddu> nirik: I dont think I am
15:18:58 <nirik> ah, .stage
15:19:01 <jednorozec> https://bugzilla.stage.redhat.com/
15:19:31 <nirik> mboddu: ok, I can check again...
15:19:36 * nirik makes a note to do that
15:19:38 <mboddu> Thanks nirik
15:20:26 <nirik> bugzilla.stage is really slow here.
15:21:48 <nirik> ok, epel9 added...
15:22:49 <jednorozec> should I close it or will you nirik?
15:23:08 <nirik> I just did.
15:23:15 <jednorozec> thanks
15:23:48 * nirik looks for that group stuff
15:26:14 <nirik> so, if you go to bugzilla.redhat.com and login to your rh account and go to:
15:26:39 <mboddu> I think I need the group name to request or something like that, iirc
15:26:40 <nirik> my links -> workflows -> request group, is one of the ones you can request 'fedora_pm' ?
15:26:46 <nirik> it's fedora_pm
15:27:07 * mboddu checks
15:27:35 <nirik> jednorozec: you might want to request too... ;)
15:27:45 <jednorozec> yeah
15:27:48 * jednorozec logs in
15:28:31 <mboddu> I am not seeing `feodra_pm` in the list
15:28:38 <nirik> ah shoot. ok.
15:28:39 <jednorozec> nope
15:30:32 <nirik> ok, I;ll dig around and find where to request it.
15:30:33 <nirik> :(
15:30:54 <jednorozec> ok, thanks
15:30:57 <jednorozec> so
15:31:04 <jednorozec> I have some progress on
15:31:07 <jednorozec> .releng 10264
15:31:08 <zodbot> jednorozec: Issue #10264: New Repo for "rpms/pypy3.7" incompletely created - releng - Pagure.io - https://pagure.io/releng/issue/10264
15:31:59 <jednorozec> so the main reason packagers use '--no-initial-commit' is when they plan to import different git repo into it.
15:32:29 <jednorozec> https://src.fedoraproject.org/rpms/python3.11
15:32:34 <jednorozec> is a nice example
15:32:58 <jednorozec> creating alias to branch without commits its a tricky thing
15:33:34 <jednorozec> because alias is created for commit and we have 0 commits there
15:33:57 <mboddu> Hmmm, interesting
15:34:27 <jednorozec> so the idea is to allow packagers to provide '--initial-commit-ref <branch/tag/commit_hash>'
15:35:23 <jednorozec> what do you think?
15:35:53 * mboddu lookin
15:36:21 <nirik> sure, seems a bit complex, but if it's not too hard from our end I know a few maintainers will like it
15:36:49 <jednorozec> well, I was unable to find a git magick that would fix this
15:37:52 <mboddu> Can we check with pingou? Since we can create a branch without any commit, how can we not create an alias?
15:37:55 <jednorozec> I did not realize that git alias is actually using commit, but it make sense
15:38:16 <mboddu> ^ well, that answers my previous question
15:38:20 <jednorozec> yup
15:40:03 <jednorozec> we can create orphan branch rawhide and orphan branch main but that is not something we actually want to do
15:40:20 <jednorozec> 'orphan branch' is a git thing
15:42:39 <pingou> the alias points to another ref
15:42:58 <jednorozec> and that have to resolve as commit right?
15:43:42 <pingou> yeah, we need one at one point anyway
15:44:49 <jednorozec> Maybe the solution is to create the main alias after first commit is pushed not when the repo is created.
15:49:07 <mboddu> ^ that would be nice, but that means we need to create a webhook for it
15:50:11 * jednorozec has no clue about pagure and webhooks
15:50:22 <jednorozec> but I would like to learn :)
15:50:41 <pingou> more likely something fedora-messaging based
15:52:19 <jednorozec> well it could be done with post-receive
15:54:13 <jednorozec> I will write up something to the ticket
15:54:18 <mboddu> +1
15:54:29 <jednorozec> #topic Open Floor
15:55:02 <jednorozec> anything for open floor?
15:55:25 <nirik> two quick things...
15:55:49 <nirik> first, I am going to be in a class starting in an hour and running (apparently) for 4h...
15:56:11 <nirik> second: jednorozec can you do the rc if it gets requested today, since I don't think mboddu is really back yet?
15:57:50 <jednorozec> nirik, sure if it comes in reasonable time
15:58:17 <jednorozec> or I will do it first thing in the morning
15:58:37 <mboddu> nirik, jednorozec : I will be around
15:58:48 <mboddu> jednorozec: If you have to leave, just go ahead
15:58:51 <nirik> we still have blockers sadly.
15:59:04 <mboddu> Also, I will be in training which start in an hour as well
15:59:19 * jednorozec is going there as well
15:59:32 <jednorozec> but not for the whole thing, it end 11:00pm local
16:00:37 <jednorozec> and the time is up
16:00:45 <jednorozec> thank you all
16:00:50 <jednorozec> #endmeeting