fedora-qadevel
LOGS
14:00:35 <tflink> #startmeeting fedora-qadevel
14:00:35 <zodbot> Meeting started Mon Apr 25 14:00:35 2016 UTC.  The chair is tflink. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:35 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:00:35 <zodbot> The meeting name has been set to 'fedora-qadevel'
14:00:35 <tflink> #meetingname fedora-qadevel
14:00:35 <zodbot> The meeting name has been set to 'fedora-qadevel'
14:00:35 <tflink> #topic Roll Call
14:01:08 * kparal is here
14:01:08 * mkrizek is here
14:01:09 <tflink> who all's here for the always exciting qadevel meeting?
14:01:21 <tflink> #chair kparal mkrizek
14:01:21 <zodbot> Current chairs: kparal mkrizek tflink
14:01:26 * kparal makes sure he's in the correct room
14:01:58 <tflink> ?
14:02:12 <tflink> kparal: if you're in the wrong room, I think we all are
14:02:19 <tflink> well, the three of us, anyways
14:03:35 <kparal> tflink: you were saying something about an always exciting meeting
14:03:44 <kparal> that confused me
14:03:54 <tflink> ouch, that stings :-P
14:03:58 <kparal> :D
14:04:23 * jskladan is here
14:04:27 <kparal> the fact that you didn't realize until now means it really had to sting, sorry :o)
14:04:34 <tflink> #chair jskladan
14:04:34 <zodbot> Current chairs: jskladan kparal mkrizek tflink
14:04:49 * adamw lurks
14:05:04 <tflink> kparal: denial can be very strong
14:05:05 <jskladan> we got delayed by IRL discussion about names
14:05:15 <jskladan> you know how that goes...
14:05:23 <tflink> #chair jskladan garretraziel lbrabec
14:05:23 <zodbot> Current chairs: garretraziel jskladan kparal lbrabec mkrizek tflink
14:05:27 <tflink> no worries
14:05:48 <tflink> let's get this apparently-not-very-exciting show started
14:06:01 <tflink> #Topic Announcements and Information
14:06:01 <tflink> #info production beaker system is mostly deployed - tflink
14:06:01 <tflink> #info new builds of libtaskotron and resultsdb with namespaces support deployed in dev - mkrizek
14:06:19 <tflink> anything else to add or comments/questions?
14:06:22 <kparal> tflink: it's not your fault, it's me who's lacking any exciting updates each and every week
14:06:49 * kparal was continuing with task-libabigail review, but haven't read the replies yet
14:07:39 <kparal> so what are we missing for beaker?
14:08:06 <tflink> adding hosts, reworking the beaker repo update task to be pungi4 compatible, testing
14:08:35 <kparal> so beaker can't import fedora 24 right now, right?
14:09:10 <tflink> not sure, don't recall if I tried that
14:09:21 <kparal> ok, I just wondered
14:09:32 <tflink> the task that scans koji for nightlies doesn't work with the new layouts
14:09:43 <tflink> IIRC, I imported f24 alpha into stg, though
14:10:04 <kparal> ok
14:10:19 <tflink> yeah, f24 alpha is in stg - the problem is reworking the task to handle new url structure
14:10:20 <dgilmore> kparal: beaker can import rawhide and nightly branched composes
14:10:49 <tflink> I don't think it'll take too long, just something that needs to be odne
14:11:03 <dgilmore> it has to import them from the public mirror though
14:11:30 <adamw> tflink: fedfind fedfind fedfind
14:11:31 <tflink> dgilmore: you mean that beaker can't import nightlies from koji?
14:11:38 <adamw> fedfind knows all about your nightlies
14:11:56 <tflink> adamw: i think the task is older than fedfind :-P
14:11:57 <dgilmore> tflink: internally they import based on the rsync done messages
14:12:13 <dgilmore> tflink: as for importing from koji I never knew it could
14:12:14 <kparal> and adamw knows everything about fedfind. by transition principle... we can use adamw to fix beaker
14:12:19 * adamw goes back to bed
14:12:20 <kparal> *transitivity
14:12:37 <dgilmore> kparal: I suggest you work with internal folk and import just teh same as they do
14:13:13 <tflink> dgilmore: we had it working before the pungi4 change - it involved a script that looked for new nightlies and discarding invalid urls
14:13:29 <dgilmore> tflink: that sounds like you were doing it wrong
14:13:59 <tflink> I'd be interested in learning another way to do it, then
14:14:06 <adamw> tflink: but seriously i do know a few things about working with pungi4 composes, so...if you wanna throw the details at me later please do
14:14:30 <dgilmore> internally they have a script that listens on fedmsg, when the compose rsync is done it fetches http://dl.fedoraproject.org/pub/fedora/linux/development/24/.composeinfo and imports based on it
14:14:46 <tflink> and that's the easy way to do it
14:14:54 <tflink> but not quite what we want
14:15:32 <tflink> the idea was to have the distinct nightly composes in beaker for reproducing issues
14:15:40 <adamw> tflink: that is easy to do too.
14:15:43 <dgilmore> tflink: I am going to be sitting down with the beaker devs in Brisbane in late June early July and try sort out the remaining issues I know about
14:16:01 <dgilmore> tflink: so if there are others I should find out what they are so we can do them also
14:16:18 <tflink> dgilmore: i don't think the problem is with beaker in this case
14:16:26 <dgilmore> tflink: .composeinfo gets a unique daily name
14:16:28 <adamw> tflink: there is a fedmsg when each nightly compose completes, which gives you its compose ID and kojipkgs URL.
14:16:29 <tflink> the problem is the sync script we're using
14:16:42 <dgilmore> tflink: name = Fedora-24-20160425.n.0
14:16:45 <dgilmore> so you can do that
14:16:47 <adamw> or yeah, you can just go from the generic URL and parse composeinfo. it's really not hard either way.
14:16:56 <tflink> that's new, though, right?
14:17:02 <adamw> it's pungi 4
14:17:07 <dgilmore> tflink: not really no
14:17:09 * tflink hasn't touched this code for months and just figured out things were broken late last week
14:17:19 <tflink> dgilmore: the url in compose complete messages?
14:18:10 <tflink> this is rapidly diving down a rabbit hole, though. I'll follow up with dgilmore and adamw outside the meeting
14:18:19 <tflink> any other comments/questions?
14:19:31 * tflink takes the silence as a no
14:19:45 <tflink> #topic Docker Testing
14:19:53 <tflink> this should be pretty quick
14:20:34 <tflink> the bash directive is still in review
14:21:27 <tflink> I'm a little unclear on how the test cases are being handled but I think we're still going to need the semi-generic "runner"
14:21:48 <tflink> lbrabec: do you think that D819 can be done this week?
14:22:12 <lbrabec> yep
14:22:50 <tflink> on a semi-related note, i don't think that the docker-autotest task is working in dev
14:23:14 <tflink> apparently docker is somewhat broken in F24 right now, though
14:23:14 <tflink> at
14:23:21 <tflink> at least that's what it sounded like last week
14:23:28 <adamw> the bug seems to be related to docker package updates, at least on last info
14:23:53 <adamw> oh, or now sgallagh says no. https://bugzilla.redhat.com/show_bug.cgi?id=1322909
14:24:28 <sgallagh> adamw: I am retesting that right this minute.
14:24:34 <sgallagh> Let me get back to you in ten minutes or so
14:25:17 <tflink> anything else on docker testing?
14:25:44 <tflink> actually
14:25:54 <mkrizek> tflink: I don't think docker-autotests failures are an issue of the check itself
14:26:03 <tflink> i was just about to ask about that
14:26:15 * tflink didn't get any farther than noticing it failed
14:26:21 <misc> enfin /win 203
14:26:28 <misc> (oups, sorry)
14:26:36 <garretraziel> there was some error in vim-minimal and vim-enhanced deps
14:26:37 <tflink> IIRC, it crashed, though, right?
14:26:37 <mkrizek> tflink: some conflicts in vim package
14:26:44 <tflink> :-/
14:27:09 <garretraziel> it tried to install docker-vim which tried to install vim-enhanced, but newer version of vim-enhanced had conflicting file with old vim-minimal
14:27:33 <garretraziel> it should get fixed automatically as soon as minion will have updated system (with most recent vim-minimal)
14:27:58 <tflink> shouldn't the old vim-minimal have been upgraded?
14:28:17 <garretraziel> it wasn't
14:28:18 <tflink> come to think of it, the client is f23
14:28:34 <tflink> is there an updated vim-minimal available to match f24?
14:28:50 <garretraziel> not sure
14:29:12 <garretraziel> we've tested it on mkrizek's ntb (which had f23 I presume)
14:29:21 <tflink> ntb?
14:29:27 <garretraziel> laptop
14:29:34 <tflink> ah
14:30:01 <garretraziel> (laptop is called "notebook" in Czech and we are often shortening it to "ntb")
14:30:05 <tflink> ok
14:30:12 <mkrizek> the minion's image is f23
14:31:29 <tflink> garretraziel: can you look into whether the failure was due to bad packages in a repo or attempting to install f24 docker on a f23 system?
14:31:53 <garretraziel> yes, I'll look into it
14:32:06 <tflink> thanks
14:32:21 * tflink assumes nothing else for this topic
14:32:26 <tflink> #topic ABI Checking
14:33:17 <tflink> I don't have anything specific here, just figured it would be wise to see if there were any known issues or topics that should be brought up
14:33:48 <kparal> I asked some questions about our scheduling (how often, which packages) on the mailing list
14:33:57 <kparal> haven't read the replies yet
14:34:20 * tflink also needs to catch up on that thread
14:35:20 * tflink takes that as nothing terribly urgent :)
14:35:22 <tflink> moving on
14:35:27 <tflink> #topic Tasking
14:35:33 <tflink> is anyone in need of tasks?
14:36:11 * kparal has his hands busy with Beta testing
14:36:32 <adamw> ahahahaha.
14:36:49 <jskladan> tflink: nope, /me will have stuff to do
14:36:50 <kparal> what's that sound? :)
14:37:41 <kparal> we're probably going to be quite busy this week
14:37:41 <tflink> kparal: you mean the sound of adamw wanting to be signed up to fix the beaker repo update task?
14:37:50 <kparal> in case we get Beta RC1, that is
14:38:06 <kparal> tflink: almost sounded like that, didn't it?
14:38:08 <tflink> sure but I do ask this question every week :)
14:38:26 <tflink> sounds like not so much
14:38:34 <tflink> which brings us to ...
14:38:55 <tflink> rather, if you do need tasks, let me know if you're not finding anything on the planning board
14:39:01 <tflink> #topic Open Floor
14:39:10 <tflink> anything else that folks wanted to bring up?
14:40:58 * tflink sets fuse
14:42:52 <tflink> thanks for coming, everyone
14:42:52 * tflink will send out minutes shortly
14:42:53 <tflink> #endmeeting