15:00:12 <bowlofeggs> #startmeeting Bodhi stakeholders (2017-02-28)
15:00:12 <zodbot> Meeting started Tue Feb 28 15:00:12 2017 UTC.  The chair is bowlofeggs. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:12 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:12 <zodbot> The meeting name has been set to 'bodhi_stakeholders_(2017-02-28)'
15:00:14 <bowlofeggs> #meetingname bodhi_stakeholders
15:00:14 <zodbot> The meeting name has been set to 'bodhi_stakeholders'
15:00:15 <bowlofeggs> #topic salutations
15:00:17 <bowlofeggs> #chair acarter bowlofeggs dgilmore masta mboddu nirik pbrobinson puiterwijk trishnag
15:00:17 <zodbot> Current chairs: acarter bowlofeggs dgilmore masta mboddu nirik pbrobinson puiterwijk trishnag
15:00:22 <trishnag> .hello trishnag
15:00:25 <zodbot> trishnag: trishnag 'Trishna Guha' <trishnaguha17@gmail.com>
15:00:39 <mboddu> .hello mohanboddu
15:00:40 <zodbot> mboddu: mohanboddu 'Mohan Boddu' <mboddu@bhujji.com>
15:00:45 <marc84> hi everyone
15:00:48 <nirik> hey. Kinda around, but trying to catch up on email flood.
15:00:54 <bowlofeggs> gr33tings!
15:02:11 <bowlofeggs> #topic announcements and information
15:02:13 <bowlofeggs> #info 2.4.0 has been released and deployed to production
15:02:23 <bowlofeggs> that's the only real announcement i have right now
15:02:59 <bowlofeggs> my time on bodhi is still slowed by my involvement with the container registry, so things have still been a little quiter on the bodhi front
15:03:20 <bowlofeggs> let's move on to discuss issues
15:03:28 <bowlofeggs> #topic Looking forward
15:03:30 <bowlofeggs> #info Bodhi's high priority issue list https://github.com/fedora-infra/bodhi/issues?q=is%3Aopen+is%3Aissue+label%3A%22High+priority%22
15:03:31 <bowlofeggs> #info High priority means it's important, but not a show stopper
15:03:33 <bowlofeggs> Any filed issues that aren't on these lists that should be?
15:03:48 <dgilmore> hola
15:04:01 <masta> hi
15:04:10 <bowlofeggs> yesterday i went and cleared the "critical" flag from three issues and switched them to "high priority" - they had been "critical" for a really long time, which says to me they aren't truly critical
15:04:22 <bowlofeggs> so now we just have high priority issues to look at
15:04:54 * trishnag looks
15:05:05 <dgilmore> bowlofeggs: ostree in koji I look forward to
15:05:55 <bowlofeggs> yeah i also look forward to that, though i'm not sure when i will be able to accomplish it
15:06:07 <bowlofeggs> i talked with lsedlar about using pungi to do that
15:06:10 <bowlofeggs> it sounds doable
15:06:41 <dgilmore> it shoudl eb
15:06:42 <dgilmore> be
15:06:52 * nirik looks
15:07:49 <bowlofeggs> it also sounds like signed repos might actually hit koji soon
15:08:04 <dgilmore> bowlofeggs: https://github.com/fedora-infra/bodhi/issues/1173
15:08:09 <dgilmore> I think that is expected
15:08:18 <bowlofeggs> i think getting ostrees into to koji is higher priority, but that could be cool to use too
15:08:43 <nirik> so one thing thats not really a big bug, but people keep bringing up... the one about showing eol releases. Maintainers would be very happy to have that fixed. ;)
15:09:14 <bowlofeggs> dgilmore: oh really? i can't remember who reported that to me, and unfortunately i didn't note it on that report
15:09:54 <dgilmore> bowlofeggs: the signing of packages is entirely disconnected to bodhi
15:09:55 <bowlofeggs> dgilmore: so you think the current behavior is correct? if so, can you comment to that effect on it and we can close it?
15:10:06 <dgilmore> bowlofeggs: I do think it is correct
15:10:07 <bowlofeggs> dgilmore: it is, but bodhi will fail if they aren't signed
15:10:22 <dgilmore> bowlofeggs: the steps if you hit that is to make sure that the packages are signed and resume the push
15:10:22 <bowlofeggs> nirik: yeah i hear about that one often
15:10:26 <nirik> https://github.com/fedora-infra/bodhi/issues/752
15:10:52 <bowlofeggs> dgilmore: cool. if you could add a comment on there that would be helpful,a nd then i'll close it
15:11:02 <bowlofeggs> or i can just quote this text if you dont' want to
15:11:42 <dgilmore> bowlofeggs: commented
15:12:03 <bowlofeggs> nirik: threebean's suggestion sounds pretty doable
15:12:19 <bowlofeggs> nirik: i marked it high priority
15:12:45 <bowlofeggs> thanks dgilmore!
15:13:33 <bowlofeggs> any other issues anyone wants to discuss?
15:15:16 <bowlofeggs> ok, we can move to open floor (i don't have much of an agenda today)
15:15:23 <bowlofeggs> #topic Open floor
15:17:06 <dgilmore> I have nothing
15:17:36 <bowlofeggs> yeah i don't have much either
15:17:48 <bowlofeggs> bodhi 2.4.0 was the first feature release i deployed that wasn't a disaster, so that's nice ☺
15:18:01 * trishnag has nothing on her side - I'm working on https://github.com/fedora-infra/bodhi/issues/1157 .
15:18:17 <trishnag> bowlofeggs: :-)
15:18:17 <bowlofeggs> after 2.2.0 and 2.3.0, i decided to be strict about requiring 100% test coverage on all new code and i think that's why this one went better
15:18:35 <mboddu> bowlofeggs: I am interested in having a fix for https://github.com/fedora-infra/bodhi/issues/1174
15:18:59 <bowlofeggs> mboddu: that one has been a bit controversial
15:19:10 <bowlofeggs> mboddu: if you look at the associated pull request, there's a discussion
15:19:27 <bowlofeggs> i liked dgilmore's suggestion there: https://github.com/fedora-infra/bodhi/pull/1214
15:20:13 <bowlofeggs> the pull request does seem to be stalled though, which makes sense since the proposed solution is completely different
15:20:36 <nirik> I'm not sure we can automatically tell whats causing the failure. would need a human.
15:21:20 <nirik> or there was some plan to do those updates at a different level than bodhi anyhow, wasn't there?
15:21:39 <bowlofeggs> yeah automatically filing a bug could be hard if we don't know for sure what broke
15:22:14 <bowlofeggs> nirik: you mean the ostree updates? i've been wanting to move the execution of that into koji, but i think it would still be driven by bodhi
15:22:48 <nirik> I thought there was talk about moving it to just update 2weekly and decouple it from updates... but I could well be misremembering.
15:23:18 <bowlofeggs> oh yes - i think maxamillion has started doing soemthing like that with a script he runs on bodhi-backend01
15:23:36 <bowlofeggs> so the release is done by that script, but it operates on the same tree that bodhi operates on
15:23:55 <bowlofeggs> so bodhi makes nightly updates to the tree, and every 2 weeks i think that script tags the tree
15:24:08 <bowlofeggs> that's my (very) loose understanding of it
15:24:26 <bowlofeggs> well nightly/daily, depending on where you live i guess ☺
15:24:42 <mboddu> bowlofeggs: then what happens if ostree fails(if it is moved to a script)?
15:25:05 <mboddu> bowlofeggs: is it handled differently and not going to effect our normal mashes?
15:25:30 <bowlofeggs> mboddu: i'm not sure - maxamillion wrote the script and i think he runs it?
15:25:44 <bowlofeggs> i'm not particularly familiar with it - it's not in bodhi's codebase
15:26:39 <mboddu> bowlofeggs: I think it needs more digging.
15:27:28 <bowlofeggs> perhaps - we did identify two issues with it that was preventing the mash from running a week or two ago (when you were on push duty) but those have since been fixed i believe
15:27:50 <bowlofeggs> the main one was that it was running as root, which prevented bodhi from editing it the next time as there were root owned files in the tree
15:27:57 <bowlofeggs> so now it runs as apache
15:28:56 <bowlofeggs> the other issue was that it needed to respect the lock files that bodhi makes when mashing so they don't both edit the tree at the same time
15:29:05 <bowlofeggs> i believe that was fixed as well
15:29:11 <mboddu> bowlofeggs: yes, that is fixed but it failed couple of other before needing other sort of fixes
15:29:28 <bowlofeggs> i'm not sure it's ever been run but once
15:29:35 <bowlofeggs> right?
15:29:37 <bowlofeggs> or has it?
15:30:42 <bowlofeggs> perhaps when we can get bodhi to use koji to handle these trees, we can also use koji to run that script to help with these issues
15:30:55 <mboddu> bowlofeggs: I think so, but I think its okay for now, but I worry about future issues and they are not the same one's and takes lot of time digging into them
15:31:26 <bowlofeggs> yeah that one was a bit tough to debug, as it wasn't immediately obvious that it wasn't a bodhi bug
15:31:37 <bowlofeggs> i wasn't aware that that script had run
15:31:43 <mboddu> bowlofeggs: yep
15:32:07 <mboddu> bowlofeggs: anyway, thats all I got, hoping to have a fix soon
15:32:07 <bowlofeggs> anything else, or shall we end early?
15:33:33 * nirik has nothing else
15:35:50 <bowlofeggs> cool, thanks for coming everyone!
15:35:55 <bowlofeggs> #endmeeting