15:00:37 <jednorozec> #startmeeting RELENG (2022-11-01)
15:00:37 <zodbot> Meeting started Tue Nov  1 15:00:37 2022 UTC.
15:00:37 <zodbot> This meeting is logged and archived in a public location.
15:00:37 <zodbot> The chair is jednorozec. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:00:37 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:37 <zodbot> The meeting name has been set to 'releng_(2022-11-01)'
15:00:38 <jednorozec> #meetingname releng
15:00:38 <zodbot> The meeting name has been set to 'releng'
15:00:38 <jednorozec> #chair nirik sharkcz pbrobinson phsmoura dustymabe jednorozec
15:00:38 <zodbot> Current chairs: dustymabe jednorozec nirik pbrobinson phsmoura sharkcz
15:00:38 <jednorozec> #topic init process
15:00:42 <bittin> joining the meeting in a minute, still in the Fedora GNOME meeting
15:01:04 <phsmoura> morning
15:01:12 * bittin is here now
15:01:13 <bittin> afternoon
15:02:01 <nirik> morning
15:02:12 <jednorozec> good morning/afternoon
15:03:10 <jednorozec> #topic current release cycle
15:04:19 <jednorozec> sadly we postponed for couple of weeks
15:04:31 <nirik> yep.
15:04:34 <jednorozec> currently we have rc-1.5 thanks nirik for making it happen
15:04:46 <bittin> due to this openSSL patch coming today: https://mta.openssl.org/pipermail/openssl-announce/2022-October/000240.html
15:04:51 <bittin> also new RC later today and on Sunday this week
15:05:41 <jednorozec> do we have any new about the openssl CVE?>
15:06:15 <bittin> not yet, later today (tm)
15:06:40 <nirik> anytime now... they said 13-17utc
15:06:51 <jednorozec> yeah we will see how bad it gets
15:07:39 <jednorozec> rc-1.5 is also synced to internal RH mirrors as of this morning
15:07:54 <bittin> yay
15:08:19 <jednorozec> so more people can have a look, I have also changed the way we sync this.
15:08:29 <jednorozec> Probably a good time to write some docs
15:09:05 <jednorozec> we now sync all RC requests so internal QE can have a look
15:09:47 <nirik> docs good
15:10:58 <jednorozec> that should be all for the current release
15:11:03 <jednorozec> #topic tickets
15:11:08 <jednorozec> so I have one
15:11:14 <jednorozec> .releng 111000
15:11:16 <zodbot> jednorozec: An error has occurred and has been logged. Please contact this bot's administrator for more information.
15:11:20 <jednorozec> heh
15:11:43 <jednorozec> heh
15:11:45 <jednorozec> .releng 111000
15:11:47 <zodbot> jednorozec: An error has occurred and has been logged. Please contact this bot's administrator for more information.
15:11:49 <jednorozec> hm
15:11:56 <nirik> 111,000 ?
15:11:58 <jednorozec> https://pagure.io/releng/issue/11100
15:12:18 <nirik> too many 0's on your request. ;)
15:12:27 <jednorozec> https://pagure.io/releng/issue/1110
15:12:27 <nirik> .releng 11100
15:12:28 <zodbot> nirik: Issue #11100: RC 1.1 and 1.2 are missing .composeinfo - releng - Pagure.io - https://pagure.io/releng/issue/11100
15:12:30 <jednorozec> https://pagure.io/releng/issue/11100
15:12:38 <jednorozec> .releng 11100
15:12:39 <zodbot> jednorozec: Issue #11100: RC 1.1 and 1.2 are missing .composeinfo - releng - Pagure.io - https://pagure.io/releng/issue/11100
15:12:42 <jednorozec> aaaa
15:12:55 <nirik> huh.
15:13:01 <nirik> what generates that?
15:13:04 <jednorozec> so the pungi log have a log about writing composeinfo bu its not there
15:13:16 <jednorozec> there is only one pungi change between beta and rc
15:13:23 <jednorozec> not talking about the file names
15:13:57 <jednorozec> but this https://pagure.io/pungi-fedora/blob/f37/f/fedora-final.conf#_66
15:14:40 <nirik> thats just a profiler, shouldn't matter
15:14:42 <jednorozec> I am no sure if related to the compose info but it is different from beta
15:14:52 <jednorozec> hm
15:15:05 <jednorozec> maybe we need finished to create the file?
15:15:11 <nirik> I had to turn it off on rawhide... and we were not really even looking at the profile data anyhow.
15:15:17 <jednorozec> so far all RC composes are FINISHED_INCOMPLETE
15:15:19 <nirik> I wouldn't think so
15:16:04 <jednorozec> I can see the log here
15:16:05 <jednorozec> https://kojipkgs.fedoraproject.org/compose/37/Fedora-37-20221030.0/logs/global/pungi.global.log
15:16:27 <jednorozec> that is the latest RC if you search for composeinfo there is a log about writing it
15:17:08 <nirik> there is a composeinfo.json...
15:17:31 <jednorozec> RH require the file to import stuff to beaker automatically, I created it manually for synced RC on RH mirror
15:17:53 <jednorozec> well yes
15:18:04 <jednorozec> but beaker expects .composeinfo
15:18:26 <nirik> that log seems to be refering to the .json one.
15:18:55 <jednorozec> yeah that is what I thought
15:20:18 <nirik> huh, well, perhaps we could ask lsedlar...
15:20:51 <jednorozec> I will ping him during day tmrw
15:20:55 <jednorozec> moving on
15:21:05 <jednorozec> .releng 11089
15:21:06 <zodbot> jednorozec: Issue #11089: Please, rename Fedora-Server-KVM Checksum file for F37 to be different from ISO - releng - Pagure.io - https://pagure.io/releng/issue/11089
15:21:17 <nirik> we do have a build_composeinfo script...
15:21:29 <jednorozec> nirik, huh
15:21:34 <jednorozec> really? is it used
15:21:54 <nirik> no idea. :)
15:22:02 * jednorozec will look into it
15:22:29 <jednorozec> so changing the filenames for sig files
15:23:02 <jednorozec> this should be fairly simple change but it will probably break a lot of stuff
15:23:11 <jednorozec> because the filename for checksum will change
15:23:23 <nirik> yeah, I'd prefer not.
15:24:18 <jednorozec> but they have a valid argument if you download more than one image in a row you have only the latest checksum
15:25:26 <nirik> true.
15:25:44 <jednorozec> it would be nice to have same filename as the image filename
15:25:52 <nirik> so we have to change it globally? not just server?
15:26:03 <jednorozec> yup
15:26:07 <jednorozec> for everything
15:26:27 <jednorozec> or we can introduce some logick to the pungi-fedora config
15:26:34 <jednorozec> but I would like to avoid that
15:26:40 <nirik> we need to get QA and websites on board.
15:26:47 <jednorozec> yeah
15:27:16 <jednorozec> that is why I am not really considering doing this now, we need to get f37 out and do changes like this later...
15:27:32 <nirik> +1
15:27:38 <nirik> completely agree
15:27:55 <nirik> changes in rawhide first
15:28:31 <jednorozec> for sure but anyway if we break QA stuff now when they have work with 37 they will not be happy
15:29:00 <jednorozec> I just wanted to bring it up because it makes sense
15:29:07 <jednorozec> for rawhide ofcourse
15:29:28 <jednorozec> everything already out will not be changed...
15:29:38 * nirik nods
15:30:17 <jednorozec> .releng 11095
15:30:18 <zodbot> jednorozec: Issue #11095: Please help me unstuck armv7hl Python builds - releng - Pagure.io - https://pagure.io/releng/issue/11095
15:31:29 <nirik> yeah, I'm still working on this.
15:31:29 <jednorozec> nirik, I have one question here
15:31:39 <jednorozec> how did you forced task to stay on one builder?
15:31:41 <nirik> I really would prefer not to have special channel/builders for just python builds
15:32:00 <nirik> koji assign-task --force taskid buildername
15:32:07 <jednorozec> because when I assigned it after oom kill it spawned on different builder...
15:32:10 <nirik> (needs koji admin)
15:32:32 <nirik> yeah, if oom killer kills kojid, it will just go to another builder after that
15:32:43 <jednorozec> ok that is what I did but there is the kernel issue
15:33:20 <nirik> yes, the issue is still there. If we use a LPAE kernel and 40gb memory, it hits OOM in tests.
15:33:32 <nirik> If we use a non LPAE kernel and 3GB memory, it works.
15:33:50 <jednorozec> yeah
15:33:52 <nirik> but that config does NOT work for a bunch of other packages (libreoffice, ceph, etc)
15:34:12 <nirik> I look forward to the day I can retire all those armv7 builders. ;)
15:34:18 <jednorozec> yeah
15:34:53 <nirik> so, we could do a few with 3gb and set python stuff to go to them... but I think I'd like to see if I can think of any other way to get a 'normal' config working
15:36:21 <jednorozec> that would be great
15:37:04 <jednorozec> we will kill the armv7 builds with EOL of f36 so next release cycle right?
15:37:06 <nirik> perhaps I can try some more middle ground, like 16GB mem...
15:37:18 <nirik> yeah, when f36 eols
15:37:24 <jednorozec> good
15:39:40 <jednorozec> So Anton is not here but he is working on a toddler to process scm requests, I hope we can get things working in staging next week
15:40:28 <jednorozec> aww
15:40:33 <jednorozec> unretiremtn requests sorry
15:40:43 <jednorozec> scm requests are alrady working in staging
15:41:08 <nirik> cool.
15:41:50 <jednorozec> I am going to close f37 changes https://pagure.io/releng/issue/10731 https://pagure.io/releng/issue/10768
15:42:18 <nirik> +1
15:45:31 <jednorozec> this one aswell https://pagure.io/releng/issue/10788
15:47:43 <jednorozec> ok that is all from me
15:47:44 <jednorozec> #topic Open Floor
15:48:07 * bittin does not have anything (more RCs incoming to test before the release in 2 weeks)
15:48:46 * nirik doesn't have anything off hand
15:49:06 <jednorozec> I will leave this meeting open until the scheduled end
15:49:23 <jednorozec> so if anything comes up we can discuss
15:49:48 <nirik> ok. I'm gonna go feed some cats... ;) will check back in a min.
15:51:14 <bittin> and i am pretty tired and slow today
15:52:20 <bittin> new openSSL is out now
15:52:22 <bittin> https://www.openssl.org/source/openssl-3.0.7.tar.gz
15:54:08 <bittin> https://mta.openssl.org/pipermail/openssl-announce/2022-November/000241.html
15:54:11 <smooge> release the kraken builders
15:54:49 * bittin is updating my webserver atm
16:00:37 <bittin> *gavel*
16:01:02 <bittin> great meeting, might not be able to attend much in the future as i am starting working again from next week (and will be on my way home when the meeting starts)
16:01:55 <jednorozec> #endmeeting