15:01:42 <tflink> #startmeeting fedora-qadevel
15:01:42 <zodbot> Meeting started Mon Nov  7 15:01:42 2016 UTC.  The chair is tflink. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:42 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:01:42 <zodbot> The meeting name has been set to 'fedora-qadevel'
15:01:42 <tflink> #meetingname fedora-qadevel
15:01:42 <zodbot> The meeting name has been set to 'fedora-qadevel'
15:01:42 <tflink> #topic Roll Call
15:01:54 * mkrizek is here
15:01:59 <tflink> #chair jskladan kparal garretraziel mkrizek
15:01:59 <zodbot> Current chairs: garretraziel jskladan kparal mkrizek tflink
15:02:08 * garretraziel is here
15:02:23 * kparal is here
15:02:45 * jskladan is here
15:03:12 <tflink> ok, lets get this party started
15:03:21 <tflink> #topic Announcements and Information
15:03:46 <tflink> #info fixed bug in trigger related to koji tag usage change (T862) - tflink
15:03:46 <tflink> #info deployed new triger release to dev/stg/prod - tflink
15:03:46 <tflink> #info deployed task-python-versions to dev/stg - mkrizek
15:03:46 <tflink> #info poked koji issue claiming 0 artifacts for a fresh new build (T776), fixed in stg, still waiting for production - kparal
15:03:56 <tflink> any questions/comments?
15:04:17 <kparal> I just asked dgilmore about new koji status, no ETA for pushing it into prod yet
15:05:43 <tflink> is it something we can test in stg?
15:06:10 <kparal> I'm not sure
15:06:24 <kparal> I guess we could use koji stg to build textlive, and write a listener for that
15:06:30 <kparal> that would immediately perform the query
15:06:36 <kparal> but it's... work
15:06:44 <kparal> *texlive
15:06:57 * tflink is more worried about the docker stuff, to be honest
15:07:08 <kparal> or cockpit docker, but there I'm not sure we have permissions to do a scratch build
15:07:09 <tflink> but it does sound like a prod change to koji isn't right around the corner
15:07:27 <tflink> I'm not sure that a stg docker build system exists
15:07:34 <kparal> <dgilmore> kparal: as soon as we are unfrozen and think everything is ready we will look at it
15:08:03 <dgilmore> it is not a priority right now
15:08:22 <tflink> ok
15:08:22 <dgilmore> the focus at the moment is fixing rawhide and getting f25 out the door
15:08:39 <dgilmore> but will become a priority after
15:08:54 <jskladan> so, does that means that we should be pushing the work-around, that kparal so detests? :)
15:08:59 <tflink> dgilmore: thanks for the update
15:09:09 <tflink> jskladan: yep, I think so
15:09:22 <kparal> will we actually build and deploy new libtaskotron that the new koji is deployed?
15:09:30 <kparal> *faster than
15:09:45 <tflink> kparal: i think jskladan means the trigger patch for docker images
15:09:58 <jskladan> kparal: how does libtaskotron go with what I'm talking about?
15:09:59 <kparal> right, it's trigger, not libtaskotron
15:10:10 * jskladan damned tflink and his fast typing skills
15:10:11 <kparal> my bad
15:10:49 <tflink> all of a sudden the spiritual energy around me is so dark ... what did jskladan do to curse me?
15:10:53 <tflink> :)
15:11:13 <tflink> anyhow, if there are no other comments on non-docker stuff, this neatly leads into the next topic
15:11:14 <kparal> one concern is that if we push the workaround, we will not know if it's fixed once new koji is deployed
15:11:27 <tflink> kparal: remove the patch and see if it works
15:11:37 <jskladan> but anyway - since we don't really have an actual date on "when new koji is deployed"
15:11:59 <tflink> and I'm getting questions on whether people can start writing tests for docker ...
15:12:18 <tflink> #topic Docker Testing Status
15:12:20 <jskladan> and as tflink says - once it's deployed, we can/will test whether the issue is fixed
15:12:29 <tflink> that was the main thing I wanted to cover here
15:12:44 <jskladan> but doing nothing because "it will be fixed real time soon now" seems like a wrong call
15:13:05 <tflink> #info there are no plans to patch koji immediately for the timing/fedmsg issue
15:13:28 <tflink> #info we will work around the issue for docker triggering for now, will test the fix once it's released into prod
15:13:47 <tflink> and hope that there aren't many texlive builds that people notice
15:13:49 <tflink> :)
15:14:13 <tflink> anything else on this?
15:14:42 <kparal> mkrizek says it affects more packages
15:14:45 <kparal> nothing else from me
15:15:02 <tflink> hope that there aren't many issues with packages until a fix is available
15:15:04 <tflink> :)
15:15:04 <jskladan> well, would be good if you guys could have a look at https://phab.qadevel.cloud.fedoraproject.org/D1031
15:15:44 <tflink> whoops, didn't realize I missed taht
15:16:43 <tflink> anything else on this?
15:16:49 <jskladan> nope
15:17:26 <tflink> ok, moving on
15:17:38 <tflink> #topic Dist-Git Task Storage Proposal (and test case docs)
15:17:57 * tflink wanted to share some updates on that proposal and see if there were thoughts on some possible changes
15:18:13 <tflink> it seems like there is a slight preference for the separate namespaces WRT task storage
15:18:24 <tflink> so I'm planning to update that part of the document
15:18:47 <tflink> has anyone read the discussion on devel@ differently?
15:19:04 <jskladan> nope, that's what I got from that too
15:19:20 * kparal read it a week ago
15:19:25 <tflink> if we're going that route, I was talking to mattdm about https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/message/YWMJBR7SOMFOGJMZYOEAKS5BCRZ2XYP3/
15:19:33 <jskladan> (apart of the doom-saying and Y U no make it how I want comments)
15:19:47 <tflink> and I'm wondering if we might want to start storing testcase docs in dist-git
15:20:08 <tflink> jskladan: silly jskladan, that's going to happen no matter what we change :-P
15:20:46 <tflink> any thoughts on how good/bad of an idea this would be?
15:21:21 <kparal> would we use the docs somehow?
15:21:48 <tflink> some of the details are still TBD
15:22:28 <tflink> but the base concept would be to change from the wiki-based test cases in the wiki which bodhi links to to a set of docs which are rendered from git
15:24:07 <jskladan> tflink: I might be reading a different conversation
15:24:20 <kparal> but that's not related to taskotron in any way, right?
15:24:22 <jskladan> that link is about reading test results
15:24:26 <tflink> not really, no
15:24:47 <tflink> bah, I linked the wrong thing
15:25:16 <kparal> I don't really mind, but I actually quite like wiki here and don't see a problem with it
15:25:45 <kparal> but if people want to store it there and bodhi will implement parsing and rendering, shrug
15:26:08 <tflink> https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/message/OV6562EG3UYZJOA2LU4J6S2VMLJ34BKD/
15:26:34 <tflink> kparal: the issue is more that most people don't know about it or use the wiki pages
15:27:23 <kparal> otoh wiki approach doesn't require commit rights
15:27:43 <kparal> so I'd expect more people to participate this way
15:27:52 <tflink> true but does that matter so much once we have PR capability on dist-git?
15:28:04 <kparal> haha, talking about ease of use
15:28:13 <tflink> AFAIK, very few people are participating as it is
15:28:18 <kparal> so people will have to learn git, clone repo, prepare a patch...
15:28:24 <kparal> tflink: I'm not denying that
15:28:33 <kparal> just... not sold on the idea
15:28:52 <tflink> kparal: other ideas on how to increase usage are welcome - this just made sense to me as a possible way forward
15:29:06 <jskladan> kparal: tflink: neither am I,  honestly, I believe that the real problem is that people hate writing docs...
15:29:27 <jskladan> not that having the possibility to do so (have the docs in dist-git) is bad
15:29:39 <tflink> jskladan: what?!? that can't be true. That's ... UNPOSSIBLE
15:30:23 <tflink> I don't think that many people like writing in the wiki
15:30:28 <tflink> outside of QA, I mean :)
15:30:45 <kparal> I don't really care either way. using wiki seems easier to me personally, but that's just one opinion
15:30:46 <jskladan> it just seems as one of those things where people try to solve the consequence and not the cause
15:31:12 <jskladan> but, whatever - it's not like we will be implementing the bodhi functionality, and I'm not against the idea as such
15:31:34 <jskladan> just seems more like a nice-to-have to me
15:31:38 <tflink> jskladan: the idea would be to make it easier for packagers to write test cases under the assumption that they would actually write test cases
15:31:49 <tflink> but that circles back to your concern
15:32:42 <jskladan> as I said earlier - I'm not against the idea at all, especially since we won't be the ones coding the changes :)
15:32:44 <tflink> anyhow, it sounds like most folks here are not sold on the idea. it can be tabled for now
15:32:57 <jskladan> (as far as I understand the proposal)
15:33:04 <kparal> btw, when pushing a new update in bodhi, there could be a colored bar "there are no test instructions written for this package. see <link> how to do it". in order to raise awareness
15:33:27 <tflink> kparal: that sounds like a good feature request for bodhi :)
15:33:42 <kparal> of course that can be applied both to wiki approach and dist-git approach
15:33:57 <kparal> raising awareness is not really tied to the implementation how to write it
15:34:30 <tflink> I'm not sure the proposal was meant to be as clear as "group X will be implementing a, b c" and more of "if we did this, would it help get more test cases?"
15:35:36 <kparal> raising awareness seems like an easier task than raising awareness + implementing a new system
15:35:45 <kparal> whoever is going to do it
15:36:03 <tflink> are there more comments or concerns about either the testcase stuff or the taskotron part of the dist-git storage stuff?
15:36:20 <kparal> nope
15:37:07 <tflink> ok, moving on
15:37:15 <tflink> #topic Rebuilding Taskotron instances
15:37:24 <tflink> this should be a pretty quick topic
15:37:42 <tflink> our current instances are running F23 and have disk space issues on the taskmaster
15:38:17 <tflink> both of which would be best handled by rebuilding with a newer Fedora (and different VM config)
15:38:32 <tflink> mkrizek: you still up for doing this today?
15:38:37 <mkrizek> yes
15:38:51 <tflink> any objections or concerns?
15:39:19 * jskladan has none
15:40:04 * tflink wonders if jskladan should be roped into this as he's the only RHCSA in here right now ...
15:40:11 <tflink> or am I mis-remembering?
15:40:32 <jskladan> you mean that I should up my qualification? :)
15:40:43 <mkrizek> tflink: or all of us except you? :P
15:40:46 <tflink> well jskladan and pschindl ... that would make far less work for mkrizek and I ....
15:41:06 <jskladan> but mine is for RHEL5
15:41:09 <jskladan> that does not count :D
15:41:10 * garretraziel should do some RH certificates too
15:41:12 <tflink> am I really the only one on our team without a RHCSA?
15:41:24 <tflink> ok, not the only one :)
15:41:37 <garretraziel> :-)
15:41:41 <kparal> I think my certification expired a long time ago
15:41:55 <tflink> in my defense, it's a lot easier for you all to do it than it is for us remotees
15:42:14 <tflink> I'm not even sure it's possible/practical without out-of-pocket costs
15:42:15 * jskladan needs to get that docker stuff done...
15:42:40 <jskladan> tflink: you could schedule exams to around devconf ;)
15:42:44 * jskladan hides
15:42:52 <tflink> anyhow, I seem to have driven us out past the weeds and into the next county
15:43:09 <pschindl> heh, what's going on. I saw my name, did I miss something?
15:43:14 <tflink> jskladan: you can schedule the exams? I thought the schedule was fixed
15:43:27 <jskladan> tflink: there is "The Booth"
15:43:28 <tflink> pschindl: we're volunteering you to take over the sysadmin stuff for taskotron :)
15:43:37 <tflink> since you're more qualified :)
15:43:54 <jskladan> where you can (as far as I understand it) just do the exam any time it's available
15:43:55 <tflink> jskladan: that sounds like the suicide booths from futurama
15:44:02 <jskladan> something like that, indeed
15:44:06 <pschindl> :D ah, ok.
15:45:00 <tflink> which reminds me that the beaker stuff is almost done, there are workarounds for the last issues I was hitting in getting systems set up
15:45:06 <tflink> but that's a whole different topic
15:45:12 <tflink> moving on to ...
15:45:15 <tflink> #topic Open Floor
15:45:43 <tflink> any topics to discuss?
15:46:33 <jskladan> tflink: dunno, how do you feel about your new overlord-to-be The Trump? :)
15:46:45 <jskladan> planning for an escape yet? :)
15:46:57 * tflink tries not to think about the election
15:47:11 <kparal> don't plan on going to CZ, we already have Trump on the throne
15:47:19 <kparal> at least the czech version of it
15:47:29 <tflink> jskladan: no comment
15:47:31 <zdenek> :D
15:47:48 <tflink> kparal: so you all trumped us already?
15:47:49 * tflink ducks
15:47:57 <kparal> :)
15:48:00 <tflink> and throws up a little ...
15:48:02 <jskladan> :D
15:48:30 <tflink> at least the commercials will stop soon
15:48:36 <garretraziel> you could say that he will...
15:48:45 <garretraziel> TRUMP the elections!
15:48:55 <tflink> ow
15:48:58 <tflink> :-P
15:49:29 <tflink> and on that note, I'm setting the fuse for some time between 0 and 5 minutes
15:50:54 * jskladan is a bit overdue for lunch, and as we all know, I'm not a pleasant one to be around, when I'm hungry...
15:51:00 <tflink> thanks for coming, everyone
15:51:15 <tflink> the fuse went off so that jskladan could get some food
15:51:21 <kparal> boom
15:51:23 <jskladan> and this is how the fuse goes BOOM https://www.youtube.com/watch?v=RljexCqQC0w
15:51:25 * tflink will send out minutes shortly
15:51:29 <tflink> #endmeeting