14:33:05 <dgilmore> #startmeeting RELENG (2016-10-31)
14:33:05 <zodbot> Meeting started Mon Oct 31 14:33:05 2016 UTC.  The chair is dgilmore. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:33:05 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:33:05 <zodbot> The meeting name has been set to 'releng_(2016-10-31)'
14:33:05 <dgilmore> #meetingname releng
14:33:05 <dgilmore> #chair dgilmore nirik tyll sharkcz bochecha masta pbrobinson pingou maxamillion mboddu
14:33:05 <zodbot> The meeting name has been set to 'releng'
14:33:05 <zodbot> Current chairs: bochecha dgilmore masta maxamillion mboddu nirik pbrobinson pingou sharkcz tyll
14:33:08 <dgilmore> #topic init process
14:33:14 <maxamillion> .hello maxamillion
14:33:17 <mboddu> .hello mohanbodd
14:33:19 <sharkcz> .hello sharkcz
14:33:22 <puiterwijk> .hello puiterwijk
14:33:22 <zodbot> maxamillion: maxamillion 'Adam Miller' <maxamillion@gmail.com>
14:33:23 <mboddu> .hello mohanboddu
14:33:26 <zodbot> mboddu: Sorry, but you don't exist
14:33:29 <zodbot> sharkcz: sharkcz 'Dan Horák' <dan@danny.cz>
14:33:31 <zodbot> puiterwijk: puiterwijk 'Patrick "マルタインアンドレアス" Uiterwijk' <puiterwijk@redhat.com>
14:33:34 <zodbot> mboddu: mohanboddu 'Mohan Boddu' <mboddu@bhujji.com>
14:33:39 <nirik> .hello kevin
14:33:40 <zodbot> nirik: kevin 'Kevin Fenzi' <kevin@scrye.com>
14:33:54 * pbrobinson o/
14:34:37 <dgilmore> #topic Alternative Architectures updates
14:34:51 <dgilmore> so I will start with power having moved on the weekend
14:35:00 <dgilmore> things were more broken than I expected
14:35:07 <dgilmore> eclipse stack is a mess
14:35:34 <dgilmore> there was a kernel build friday that led to the buildroot being broken, which was outside our control
14:36:05 <dgilmore> the initial broken deps for the rawhide buildroot was around 1500 lines
14:36:14 <dgilmore> I got it down to about 1000
14:36:37 <dgilmore> #info ppc merged into koji.fp.o for rawhide
14:36:50 <dgilmore> https://pagure.io/pungi-fedora/pull-request/99 needs reviewed and merged
14:37:58 <pbrobinson> dgilmore: I've got that on my list to review, I need to finish up a pair of runroot hosts first
14:38:07 <dgilmore> I have not seen any feedback on anything else yet
14:38:14 <pbrobinson> dgilmore: that will be done this afternoon once I'm off conf calls
14:38:21 <dgilmore> koeschi is disabled du to needing more ppc builders
14:38:35 <pbrobinson> dgilmore: yep, all on my list todo today
14:38:51 <pbrobinson> I meant to do it yesterday but ended sleeping for most of the afternoon
14:39:19 <dgilmore> #info koeschi disabled due to small number of power builders
14:39:42 <dgilmore> #action pbrobinson to reallocate builderd from ppc.koji to koji
14:40:12 <dgilmore> #action dgilmore to file bug on the eclipse breakage
14:40:35 <dgilmore> seems that ppc.koji was ignoring that eclipse-cdt was an ancient build and current builds are FTBFS
14:40:56 <dgilmore> so the entire eclipse stack is broken
14:41:28 <dgilmore> sharkcz: how is s390?
14:41:34 <pbrobinson> dgilmore: I think it needed to be bootstrapped again, I remember something but not exact details of it
14:41:39 <maxamillion> sounds like a busy weekend :/
14:41:53 * dgilmore worked 10 hours or so saturday
14:42:01 <maxamillion> dgilmore: :(
14:42:08 <sharkcz> dgilmore: no problems on s390
14:42:36 <sharkcz> will ask the nodejs team for a re-bootstrap in f26 top include s390x
14:42:45 <dgilmore> #info sharkcz says 2390 is okay
14:43:10 <pbrobinson> sharkcz: once the macro is updated I think most of it should then be fixed once we do that mass rebuild?
14:43:53 <sharkcz> pbrobinson: the mass rebuild will clean it up for sure, dunno if a real bootstrap due cyclic deps is required
14:44:16 <pbrobinson> sharkcz: there's not that many TBH, most are noarch
14:44:27 <sharkcz> for the record s390 also lacks the eclipse-cdt ...
14:44:33 <dgilmore> #undo
14:44:33 <zodbot> Removing item from minutes: INFO by dgilmore at 14:42:45 : sharkcz says 2390 is okay
14:44:36 <pbrobinson> sharkcz: I think they whole lot needed in boostrap is < 20  packages
14:44:38 <dgilmore> #info sharkcz says s390 is okay
14:45:00 <dgilmore> #info sharkcz to ask for a rebootstrap of nodejs to enable s390x support
14:46:04 <dgilmore> #info sharkcz reports eclipse-cdt missing from s390x
14:46:10 <dgilmore> anything else ?
14:46:22 <sharkcz> no
14:46:36 <dgilmore> #topic pagure migration
14:46:51 <dgilmore> #info all issues migrated except for releng
14:47:06 * nirik nods. There was a bug we hit on that one.
14:47:09 <dgilmore> nirik: any progress on the bug you hit for the releng issues?
14:47:27 <dgilmore> #info releng repo hit a big in the migration process
14:47:52 <nirik> not that I have seen...
14:48:01 <linuxmodder> .fas linuxmodder
14:48:01 <zodbot> linuxmodder: linuxmodder 'Corey W Sheldon' <sheldon.corey@openmailbox.org>
14:48:25 <nirik> the bug was due to trac allowing multiple attachments that have the exact same timestamp (like they were attached at exactly the same milisecond)
14:48:33 <dgilmore> nirik: what is the bug for it?
14:48:36 <dgilmore> hi linuxmodder
14:48:40 <linuxmodder> dgilmore,  o/
14:48:42 <nirik> https://pagure.io/pagure-importer/issue/60
14:49:19 <dgilmore> nirik: thanks
14:49:37 * masta looks in
14:49:41 <masta> hey gang
14:49:42 <dgilmore> #info help wanted to fix https://pagure.io/pagure-importer/issue/60 so we can migrate releng to pagure
14:50:15 <pbrobinson> masta: on time as usual I see...
14:50:45 <linuxmodder> dgilmore,  been kinda out of the loop what is the core bug itself?
14:51:01 <puiterwijk> dgilmore: I can get you a patch that'll work in about 5 minute sif you want.
14:51:18 <dgilmore> linuxmodder: as I understand it the issue was that one issue had two attchments the same
14:51:34 <dgilmore> puiterwijk: sure will not say no
14:51:57 <dgilmore> #topic Open Floor
14:52:06 <dgilmore> does anyone have anything to bring up?
14:52:19 <puiterwijk> Small thing
14:52:21 <dgilmore> https://fedorahosted.org/rel-eng/ticket/6488 was just filed
14:52:36 <pbrobinson> on the broken deps?
14:52:49 <puiterwijk> Last week, all builders in staging have been moved to use kerberos. Let me know if you see issues there
14:53:11 <nirik> the */15min atomic compose is busted completely. See releng-cron list for the output...
14:53:12 <dgilmore> #info puiterwijk moved all builders to use kerberos
14:53:22 <puiterwijk> #undo
14:53:35 <puiterwijk> #info puiterwijk moved all staging builders to use kerberos
14:53:46 <puiterwijk> (not prod yet, planning to do that after GA)
14:54:00 <nirik> also we are hitting a bug again in koji where it cannot remove a buildroot and is spamming logs. need to file that upstream and fix our builders today.
14:54:19 <puiterwijk> I'm considering disabling cert auth in staging this week, and see what breaks and who'll complain
14:55:08 <dgilmore> puiterwijk: why did you undo that?
14:55:19 <puiterwijk> dgilmore: because I added the "staging" part :)
14:55:25 <puiterwijk> aka, I added the info that I only moved staging builders
14:55:27 <dgilmore> oh
14:56:14 <dgilmore> nirik: is that because some build process is still running?
14:56:36 <nirik> dgilmore: I think it's because some package made a directory tree thats too deep...
14:57:24 <nirik> todays log file is 14GB already.
14:57:52 <dgilmore> nirik: there is no emails about the atomic-cd cron job in the last day to the releng-cron list
14:58:43 <nirik> odd. I am getting them here.
14:59:12 <nirik> http://paste.fedoraproject.org/466848/92594614/
14:59:40 <puiterwijk> Oh, and the new Bodhi 2.3 release is in production as of last thursday. This one has gated autosigning, and a rewritten push.py.
14:59:53 <dgilmore> that does not necessarilly mean anything is wrong or broken
15:00:26 <dgilmore> but the logging is likely not quite correct
15:00:52 <dgilmore> #info Bodhi 2.3 release is in production as of last thursday.
15:01:08 <dgilmore> puiterwijk: we do not use push.py so you will have to explain what that means
15:01:20 <puiterwijk> dgilmore: sorry. push.py is bodhi-push
15:01:29 <dgilmore> okay
15:03:11 <dgilmore> nirik: please be sure to file a koji issue for the failure to remove a buildroot
15:03:24 * nirik will
15:03:53 <dgilmore> does anyone have anything else?
15:05:02 <dgilmore> #endmeeting