fedoraqa-devel
LOGS
14:01:43 <tflink> #startmeeting
14:01:43 <zodbot> Meeting started Mon Jul 13 14:01:43 2015 UTC.  The chair is tflink. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:43 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:01:43 <tflink> #meetingname fedoraqa-devel
14:01:43 <tflink> #topic roll call
14:01:43 <zodbot> The meeting name has been set to 'fedoraqa-devel'
14:01:50 <tflink> #chair jskladan kparal mkrizek
14:01:50 <zodbot> Current chairs: jskladan kparal mkrizek tflink
14:01:52 * mkrizek is here
14:01:54 * roshi is here
14:01:58 * jskladan is here
14:01:58 <tflink> #chair roshi
14:01:58 <zodbot> Current chairs: jskladan kparal mkrizek roshi tflink
14:02:00 * garretraziel is here
14:02:02 * kparal is here
14:02:05 <tflink> #chair garretraziel lbrabec
14:02:05 <zodbot> Current chairs: garretraziel jskladan kparal lbrabec mkrizek roshi tflink
14:02:10 * kparal pokes lbrabec
14:02:12 <tflink> chairs for everyone!
14:02:42 <garretraziel> yay!
14:02:48 <lbrabec> hi
14:03:13 * lbrabec is here
14:03:22 <tflink> i think that's about everyone that I was expecting
14:03:31 <tflink> who wants to go first with status updates?
14:03:49 * mkrizek stands up, sitting on a chair is too mainstream now
14:03:50 <mkrizek> I can
14:03:58 <mkrizek> #topic mkrizek status report
14:04:14 <mkrizek> #info pushed the logging rework patch
14:04:14 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/D382
14:04:14 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/T495
14:04:14 <mkrizek> #info pushed the distgit directive
14:04:14 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/D383
14:04:16 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/T492
14:04:18 <tflink> mkrizek: that sounds dangerously close to a standup meeting :)
14:04:18 <mkrizek> #info latest stable build in the koji directive
14:04:21 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/T491
14:04:23 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/D417
14:04:26 <mkrizek> #info started artifacts deployment on the dev machine
14:04:28 <mkrizek> #link https://phab.qadevel.cloud.fedoraproject.org/T505
14:04:31 <mkrizek> #info code reviews
14:04:33 <mkrizek> #info need to start preparing for my flock workshop
14:04:36 <mkrizek> #info to get back to the fedmsg thread on @qadevel
14:04:38 <mkrizek> :D
14:05:57 <tflink> sounds like good progress
14:06:09 <tflink> what all are we missing for the abicheck stuff on our end?
14:06:33 <mkrizek> to finish changes in the koji directive, D417
14:07:07 <mkrizek> we still don't know how the check will be run, so that might need some changes as well once that's done
14:07:14 <kparal> and then all deps of T490 will be closed. not sure if we need something else
14:07:40 <tflink> we'll have to follow up with the abicheck folks
14:07:56 <tflink> but from what I remember, that should be everything they needed
14:08:13 <mkrizek> *should*
14:08:24 <tflink> it'd be awesome if we could get that working, even if just in dev, for flock
14:08:28 <kparal> what's the punishment for the s-word?
14:08:35 <tflink> flogging, I think
14:08:54 <kparal> I'll equip Martin and Josef with our whip, when they go to Flock
14:09:01 * tflink adds self-flagellation to his todo list for the day
14:09:07 <tflink> :)
14:09:28 * jskladan heard that freshly boiled spaghetti work fine...
14:09:45 <kparal> chef's approach
14:10:09 * tflink thinks that spaghetti would be better than leather+glass
14:10:16 <lbrabec> only if they are cooked al dente
14:10:17 <tflink> for the receiving end, anyways
14:10:34 * kparal can go next
14:10:38 <tflink> any other questions/comments for mkrizek?
14:10:48 <jskladan> none here
14:10:58 <kparal> thumbs up for mkrizek
14:11:08 <danofsatx> oh, qa-devel
14:11:10 * danofsatx is here
14:11:17 <tflink> danofsatx: welcome
14:11:30 <tflink> kparal: sounds like you're ready to go next?
14:11:31 <jskladan> #topic jskladan updates
14:11:31 <jskladan> #info no real updates here - spent most of the last two weeks either on PTO or SICK
14:11:31 <jskladan> #info spent most of the working time prepairing for Flock & with simple patches
14:11:31 <jskladan> #info in need of tickets
14:11:43 <jskladan> oops, ninja'd kparal there
14:11:52 * kparal prepares revenge
14:12:13 <tflink> jskladan: I'll get some stuff written up today
14:12:39 <tflink> there's some disposable client bits that I've been needing to write up
14:12:44 <jskladan> tflink: cool
14:12:50 * jskladan is looking forward to it
14:13:00 <tflink> unless you want to help with php-packaging-happy-fun-time :)
14:13:58 <tflink> jskladan: how long do you figure it'll take to deal with testdays?
14:14:13 <kparal> php packaging made him speechless
14:14:19 <jskladan> tflink: I guess I should be able to put the ansible playbook up for review tomorrow
14:14:55 <jskladan> hopefully I still remember the whats and whys
14:15:04 <tflink> cool, I think that's a bit higher priority than the disposable stuff - would be nice to kill off the F18 box and another old-cloud instance
14:15:40 <jskladan> ::thumbs_up
14:15:43 <tflink> any other questions/comments before moving on?
14:15:53 <mkrizek> nothing here
14:16:11 <tflink> kparal: want to go next, for real this time? :)
14:16:16 <kparal> #topic kparal's status report
14:16:22 <kparal> #info a small TAP fix in resultsdb_directive
14:16:23 <kparal> #link https://phab.qadevel.cloud.fedoraproject.org/D414
14:16:27 <kparal> #info lots of reviews, particularly with lbrabec's patches to implement disposable clients-related runner changes
14:16:32 <kparal> #info followed up with dist-tag packaging changes, it's not clear yet whether we'll end up having all packages with a dist tag, dgilmore doesn't seem to like that much
14:16:32 <kparal> #link https://bugzilla.redhat.com/show_bug.cgi?id=1238214
14:16:37 <kparal> #info lots of changes to simplify setting a development environment (a few more yet coming)
14:16:38 <kparal> #link https://phab.qadevel.cloud.fedoraproject.org/T523
14:17:34 <tflink> dist-tag packaging?
14:17:48 <kparal> that's related to the koji directive feature
14:18:05 <kparal> and the dist-git directive
14:18:29 <kparal> we want to know which release a build was built against
14:18:39 <kparal> we could rely on .fc22, if all packages had it
14:18:55 <kparal> which they should (because it's mandatory by packaging guidelines), but don't
14:18:56 <tflink> don't they?
14:19:02 <tflink> fun
14:19:04 <kparal> some of them don't
14:19:23 <kparal> there's one exception, though, fedora-release defines those macros, so it can't use them
14:19:49 <kparal> I think that still doesn't preclude having the dist tag in NVR, but dgilmore didn't seem to happy. will follow with him later
14:19:55 <tflink> do we know how many packages don't have a dist tag?
14:20:06 <kparal> about 30, now reduced to about 10
14:20:37 <kparal> the alternative solution is to query Koji of every single package. we wanted to avoid the unnecessary calls, if possible
14:20:52 <kparal> network calls are our #1 reason for crashing
14:20:53 <tflink> we could special-case those 10 packages, too
14:21:10 <tflink> ie - don't ask koji unless name in [bad-packages]
14:21:11 <kparal> yes, that's also possible
14:21:29 <kparal> a bit harder to maintain, though
14:22:08 <tflink> i would think that those 10 packages would be a upper bound
14:22:25 <tflink> could more stuff get though review without having a dist tag?
14:22:34 <kparal> I think the only packages which might get the exception from the packaging comittee are just fedora-release and maybe fedora-repos. I see no reason for the rest
14:22:54 <kparal> dist tag was not mandatory before
14:23:00 <kparal> it got mandatory around this christmas
14:23:36 <tflink> so the number *shouldn't* increase going forward
14:23:42 <kparal> no
14:23:43 * tflink signs himself up for more flogging
14:24:05 <kparal> unless someone makes a typo in the spec file, but we can solve that by simply crashing :)
14:24:16 <dgilmore> kparal: I strongly think you need to talk to koji to get things right
14:24:27 <tflink> yeah, I'm ok with that as long as we're clear on why it's crashing
14:24:43 <kparal> dgilmore: hello. available for a chat in 30 minutes?
14:25:03 <dgilmore> kparal: I am not. I have meetings from 10 on for about 2-3 hours
14:25:19 <dgilmore> 10 Central is in 35 minutes
14:25:35 <kparal> I don't want to hijack this meeting
14:25:41 <dgilmore> kparal: sure
14:26:10 <kparal> let's talk in #fedora-qa now, then
14:26:14 * tflink has other questions as well but will save them for not-in-meeting
14:26:19 <dgilmore> kparal: I am getting my daughter ready to go to summer camp so now is not good for me either
14:26:34 <kparal> ok, some time this week, I'll try to ping you :)
14:27:23 <tflink> garretraziel, lbrabec: do either of you have status updates to report?
14:27:24 <kparal> any other comments, anyone?
14:27:38 <garretraziel> tflink: if you want to hear about OpenQA…
14:28:02 <kparal> I think OpenQA is also relevant in qa-devel
14:28:16 <garretraziel> ok
14:28:20 <garretraziel> #topic jsedlak status report
14:28:24 <kparal> it's mainly jskladan and adamw who are familiar with it, but a few status updates might be interesting for anyone
14:28:38 <tflink> yeah, I don't pretend to know all that much about it :)
14:28:45 <tflink> the details, anyways
14:28:50 <garretraziel> #info been working on running tasks on Docker
14:28:58 <garretraziel> #link https://phab.qadevel.cloud.fedoraproject.org/D420
14:29:17 <garretraziel> #info migrated from subprocess calling to using adam's python library
14:29:19 <garretraziel> #link https://phab.qadevel.cloud.fedoraproject.org/D425
14:29:36 <garretraziel> #info been working on InstallGuide and README for new users :-)
14:29:49 <garretraziel> oh and
14:30:00 <garretraziel> #info finally updated python-flask-restful package
14:30:02 <garretraziel> #link https://bugzilla.redhat.com/show_bug.cgi?id=1061732
14:30:11 <tflink> did that break anything?
14:30:29 <garretraziel> you mean changes in openqa?
14:30:34 <kparal> and, afaik, whole anaconda testing is currently broken due to some gtk changes
14:30:37 <tflink> the flask-restful stuff
14:30:53 <tflink> oh, nvm. i misunderstood
14:31:04 <garretraziel> yup, anaconda is broken
14:31:24 <tflink> i thought you were talking about a version/api change for the packaged python-flask-restful
14:31:28 <garretraziel> if you want to see how our tests are failing, see http://fedora-qa-01.lab.bos.redhat.com/
14:31:28 <kparal> garretraziel: anaconda _testing_ ;)
14:31:44 * tflink was worried that some of our flask-restful using code might break
14:31:50 <garretraziel> tflink: nope
14:32:11 <tflink> note for other folks - that url above is not public, the reasons for that have been discussed elsewhere
14:32:12 <garretraziel> tflink: I just corrected specfile and updated version. it still isn't in repos though
14:32:24 <tflink> yeah, that was the part I misunderstood
14:32:53 <tflink> any other comments/questions?
14:33:56 <tflink> lbrabec: do you have anything to update for status?
14:34:02 <lbrabec> yep
14:34:08 <lbrabec> #topic lbrabec status report
14:34:16 <lbrabec> #info package installation from formulae
14:34:16 <lbrabec> #link https://phab.qadevel.cloud.fedoraproject.org/T300
14:34:17 <lbrabec> #link https://phab.qadevel.cloud.fedoraproject.org/D419
14:34:17 <lbrabec> #info taskotron new cli argument '--local', new unit tests
14:34:17 <lbrabec> #link https://phab.qadevel.cloud.fedoraproject.org/D426
14:34:17 <lbrabec> #info resultsdb frontend search fix and improvement
14:34:19 <lbrabec> #link https://phab.qadevel.cloud.fedoraproject.org/T460
14:34:23 <lbrabec> #link https://phab.qadevel.cloud.fedoraproject.org/D416
14:34:43 <lbrabec> it's mainly work on disposable related stuff, alot was discussed with kparal as he mentioned above
14:36:11 <tflink> cool, looks good
14:36:32 * tflink notes that he needs to find a better way of keeping up with code reviews - his current method is obviously not working all that well :-/
14:36:47 <tflink> any other comments/questiosn?
14:37:15 <kparal> tflink: there's a list of pending reviews on the phab home page, I believe
14:37:28 <tflink> kparal: yeah, but that's not the whole story
14:37:30 <kparal> but once a single person approves it, it disappears for everyone
14:37:37 <tflink> exactly
14:37:38 <kparal> so it's not completely usdeful
14:37:50 <kparal> I just keep tabs open in the browser :)
14:38:03 <kparal> or tag emails in email client
14:38:04 <tflink> there's a python lib for interfacing with phabricator, may poke at it
14:38:35 <tflink> i think that I'm next if there are no more comments/questions
14:38:54 <kparal> go ahead
14:39:19 <tflink> #topic "tflink - status report"
14:39:19 <tflink> #info trying to help with testcloud packaging and testing
14:39:19 <tflink> #info still working on better phabricator packages, want to migrate off of qadevel.cloud before Alpha freeze
14:39:19 <tflink> #info hoping to submit arcanist and libphutil as fedora packages soon, phabricator will take a bit more work
14:39:19 <tflink> #info volunteered to fix the fedorapeople autoindex style issues from the upgrade - https://fedorahosted.org/fedora-infrastructure/ticket/4817
14:39:21 <tflink> #info more work to replace our makefiles - https://phab.qadevel.cloud.fedoraproject.org/D422
14:40:14 <tflink> any comments/questions?
14:40:37 <kparal> doit looks interesting. what's the major advantage over a makefile?
14:41:02 <tflink> written in python
14:41:22 <tflink> but it also has more options for dependencies and importing
14:41:41 <tflink> the biggest driver right now is me wanting to automate more of our release process
14:41:42 <kparal> ok, thanks
14:42:12 * kparal looks forward to the glorious future of daily deployments to dev :)
14:42:12 <tflink> since I've forgotten at least one step for most releases :-/
14:42:42 <tflink> the first beta of buildbot-0.9 was released last week, looking forward to poking at it
14:43:27 <tflink> ok, anyone I forgot to ask for status updates?
14:43:57 * tflink takes that as a no
14:44:06 <tflink> moving on with 15 minutes left
14:44:07 <kparal> no
14:44:10 <tflink> #topic artifacts deployment status
14:44:39 * tflink just wanted to check in quick and make sure that the deployment process was going smoothly
14:44:54 <tflink> or if there was anything needed, problems found etc.
14:45:46 <mkrizek> rpms was updated by a playbook over weekend it seems, so only thing left is running db upgrade for execdb via alembic
14:46:17 * mkrizek has yet to check if anything is broken
14:46:29 <tflink> we should add db update to the playbooks if that's reasonable to get done
14:46:33 * kparal hasn't checked either
14:46:40 <mkrizek> yeah
14:46:53 <mkrizek> I have that on my todo list
14:47:14 <tflink> my  email folder hasn't exploded so I don't think anything is catastrophically broken :)
14:47:54 <mkrizek> might be that email reporting is broken :P
14:47:55 <tflink> sounds like things are progressing smoothly enough
14:48:04 <kparal> mkrizek: +1! :)
14:48:19 <tflink> mkrizek: blasphemy! that's completely un-possible
14:48:24 <tflink> :-D
14:48:56 <tflink> anything else for this topic?
14:49:01 <mkrizek> none here
14:49:03 * tflink assume not
14:49:14 <tflink> #topic tasking
14:49:25 <tflink> who all is needing tasks right now?
14:49:57 * tflink hopes that the testcloud packaging stuff will be resolved later this week and work on libtaskotron can continue
14:50:09 * kparal is finishing dev setup improvements and will be working on T408 implementation, probably with lbrabec
14:50:28 * mkrizek still has stuff to do
14:50:53 * garretraziel will be working on OpenQA
14:50:56 <tflink> ok, sounds like it's just jskladan who's looking for tasks right now?
14:51:02 <jskladan> tflink: seems like it
14:51:26 <tflink> cool, just wanted to make sure that other folks weren't waiting for tasks :)
14:52:12 <tflink> #topic discussions and tool testing
14:52:23 <tflink> not much time for discussion here, mosly a reminder
14:52:58 <tflink> if you have thoughts on the git hosting topic or the upcoming phab instance move, speak soon or it'll be too late
14:53:30 <kparal> too late means we're going to do it or not?
14:54:14 <tflink> #info discussion on git repo hosting is still not resolved, folks should speak up soon if they have opinions
14:54:22 <tflink> kparal: pretty much, yeah
14:54:46 <kparal> I don't have any big concerns there
14:54:48 <tflink> at least the phab move but I think that's less potentially controversial
14:55:01 * tflink wants that old el6 not-ansible-controlled box to go away
14:55:24 * jskladan has no big issues with git hosting either way
14:56:07 * nirik has a item (not sure if it's been discussed, but wanted to bring up in open floor)
14:56:22 <tflink> which brings us to ...
14:56:25 <tflink> #topic open floor
14:56:37 <tflink> nirik: go for it
14:56:49 <nirik> cloud migration. ;) From old to new.
14:57:03 <nirik> in particular I think there's testdays instance to do? or whats the plan for it?
14:57:04 <tflink> is there a hard deadline?
14:57:16 <tflink> nirik: we're planning to kill it
14:57:27 <tflink> jskladan is working on playbooks this week
14:57:33 <nirik> we don't have a deadline, but will probibly come up with one soon.
14:57:35 <tflink> there's a new version that we'd rather deploy
14:57:47 <nirik> ok, depoly in new cloud? or as a real vm?
14:58:07 <tflink> we were hoping to put it in infra but haven't started that process yet
14:58:45 <tflink> I'm hoping to migrate qadevel.cloud in the next couple weeks, taskotron-demo can be destroyed if it hasn't already
14:59:00 <tflink> that's all the cloud instances I can think of off the top of my head
14:59:25 <nirik> ok.
14:59:37 * nirik would like to retire the old cloud as soon as we can.
14:59:43 <tflink> understood
15:00:01 <tflink> understandable, too  :)
15:00:23 <nirik> I'll check for any other instances you folks might have
15:00:50 <nirik> theres:
15:01:31 <nirik> taskotron-demo, qadevel-stg, qadevel, testdays
15:01:38 <tflink> the soft-ish deadline I have in my head is alpha freeze
15:01:49 <tflink> I'd like to have migrated/destroyed all the old cloud instances before then
15:01:55 <nirik> ok
15:02:10 <tflink> the qadevel migration is in progress, demo can be done any time, testdays is in progress
15:02:22 * nirik nods. great.
15:02:51 <tflink> by in progress, I mean that prep work has started - no firm dates on actual migration yet
15:03:03 <tflink> anyhow, we're over time
15:03:20 <tflink> I'll follow up with nirik out of meeting about the cloud instances
15:03:26 <tflink> anything else for open floor?
15:03:30 <nirik> sounds good. thanks
15:03:38 * roshi has nothing
15:03:54 <mkrizek> nothing here
15:04:00 <kparal> neither here
15:04:10 <tflink> ok, I'll close things out, then
15:04:17 * tflink will send out minutes shortly
15:04:21 <tflink> thanks for coming, everyone
15:04:23 <tflink> #endmeeting