15:00:27 <jednorozec> #startmeeting RELENG (2022-05-31)
15:00:27 <zodbot> Meeting started Tue May 31 15:00:27 2022 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
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_(2022-05-31)'
15:00:27 <jednorozec> #meetingname releng
15:00:27 <zodbot> The meeting name has been set to 'releng'
15:00:27 <jednorozec> #chair nirik sharkcz pbrobinson phsmoura dustymabe jednorozec
15:00:27 <jednorozec> #topic init process
15:00:27 <zodbot> Current chairs: dustymabe jednorozec nirik pbrobinson phsmoura sharkcz
15:00:42 <phsmoura> morning
15:01:16 <nirik> hello. morning everyone
15:02:21 <jednorozec> hello
15:02:28 <jednorozec> so i just noticed this
15:02:31 <jednorozec> .releng 10816
15:02:32 <zodbot> jednorozec: Issue #10816: Bodhi update FEDORA-2022-a5a2935398 was part of Silverblue compose before pushed to stable - releng - -
15:03:01 * nirik waits for pagure
15:03:59 <nirik> I don't understand how that could happen.
15:04:45 <nirik> oh wait. I know.
15:05:07 <jednorozec> how?
15:05:31 <nirik> The pushes failed on sunday... I thought I had fixed it sunday night, but I didn't. :( I am doing a new push now...
15:05:59 <nirik> but it should only do the ostree part at the end... so hum.
15:07:08 <nirik> It must be somehow related to the failed compose...
15:08:17 <jednorozec> I was going to push it jsut before the meeting
15:08:31 <jednorozec> but there is another compose running
15:08:44 <nirik> yes, I resumed the failed one...
15:09:53 * jednorozec is looking at the github issue
15:10:41 <jednorozec> ok a lot of not relevant stuff
15:13:03 <nirik> ok, it does do the ostree part before I thought.
15:13:19 <nirik> perhaps we should ask pungi to move ostree to the latest they can...
15:14:45 <jednorozec> what difference will it make? that it is running at the end of the compose?
15:14:48 <nirik> or just not worry since it doesn't happen very often that it fails and we don't fix it right away
15:14:56 <nirik> well, what happened was:
15:15:26 <nirik> f36-updates compose started. made repos. ran ostree commit. wrote ostree-commit. failed due to some unsigned packages.
15:15:40 <jednorozec> oh
15:15:40 <nirik> so the updates push failed, and the real repos were not updated.
15:15:44 <nirik> but ostree was...
15:16:31 <nirik> at least I think thats what happened.
15:16:51 <jednorozec> so if it is the last thing running in compose, in this case it will fail on regular rpms so ostree would never do the commit?
15:17:22 <jednorozec> s/so/and
15:17:28 <nirik> yeah, thats my theory...
15:18:51 <nirik> I can update the ticket with my theory...
15:19:05 <jednorozec> please do that
15:21:15 <nirik> hum,
15:21:25 <nirik> perhaps my theory is wrong
15:21:55 <jednorozec> how so?
15:22:20 <nirik> new-updates-sync sync's ostree/silverblue too...
15:22:52 <nirik> so, I am not sure how silverblue got updated then
15:24:07 <jednorozec> huh
15:24:20 <nirik> weird...
15:24:41 <nirik> so I resumed this on sunday... and thought I fixed it... but... I don't see any errors in that push?
15:24:55 <nirik> 2022-05-30 04:45:05 [INFO    ] Compose finished: /mnt/koji/compose/updates/Fedora-36-updates-20220530.1
15:27:16 <nirik> May 30 04:54:11 celery-3[1276]: [2022-05-30 04:54:11,931: INFO/ForkPoolWorker-16]   name:  f36-updates           success:  True
15:28:04 <jednorozec> huh
15:28:38 <nirik> yeah, unsigned... at least thats an error. ;)
15:29:07 <jednorozec> yeah
15:30:09 <nirik> so, I think perhaps:
15:30:34 <nirik> failed compose, I fixed some unsigned things and resumed it. it finished ok, but then the daily one ran right after that and failed?
15:31:04 <nirik> jednorozec: shall I fix and resume? or you want to?
15:31:32 <jednorozec> nirik, go for it
15:32:02 <jednorozec> but I am still not sure that I understand how did the package got into the updates
15:32:40 <jednorozec> I mean ostree ofcourse
15:33:00 <nirik> yeah me either now. ;(
15:34:24 <nirik> I added some questions there.
15:36:26 <jednorozec> ok moving on for now
15:36:30 <jednorozec> .releng 10808
15:36:31 <zodbot> jednorozec: Issue #10808: Cannot request epel8 branch for uptimed using fedpkg - releng - -
15:36:56 <jednorozec> so is this just about updating the exception message to reflect correct URL to pagure not src.fp.o
15:37:15 <nirik> I think so, but would be nice to hear from the user that it fixed their case.
15:39:01 <jednorozec> ok I will take the ticket so when they comment I will get a notification and update fedpkg
15:41:52 <jednorozec> .releng 10553
15:41:53 <zodbot> jednorozec: Issue #10553: Remove nginx 1.[14..20] stream from epel-modular - releng - -
15:42:14 <jednorozec> that slipped under my radar
15:44:27 <nirik> yeah, we have a process for this now right?
15:46:31 <jednorozec> do we?
15:47:02 <nirik> I thought so? perhaps I am thinking of retiring streams in favor of another one?
15:47:39 <jednorozec> well maybe we do have that but it is not part of the releng sops
15:47:44 <jednorozec> let me check the infra
15:49:35 <nirik>
15:52:07 <jednorozec> ok so it is not building new builds for the module
15:53:43 <jednorozec> and maintainer jsut need to set EOL to smoe date to actually make the modules obsolete
15:55:48 <nirik> huh, so whats the difference between that and the above process?
15:56:07 <jednorozec> none i just read that in the doc page
15:56:24 <jednorozec> what they did already is this
15:56:41 <jednorozec> so it doe snot build for el8 but builds for fedora
15:56:49 <jednorozec> if I understand it correctly
15:57:49 <nirik> ah, right, one is just eoling and the other is replacing a stream with another
15:57:53 <nirik> so yeah...
15:58:21 <jednorozec> ok
15:58:33 <jednorozec> we are nearly at the end of time
15:58:41 <jednorozec> #topic Open floor
16:00:55 * nirik has nothing off hand.
16:01:05 <jednorozec> thnak you for your time
16:01:09 <jednorozec> #nedmeeting
16:01:14 <jednorozec> #endmeeting