rdo
LOGS
15:01:50 <jpena> #startmeeting RDO Packaging Meeting (2015-09-30)
15:01:50 <zodbot> Meeting started Wed Sep 30 15:01:50 2015 UTC.  The chair is jpena. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:02:04 <jpena> #topic Roll Call
15:02:08 <trown> o/
15:02:09 <dmsimard> o/
15:02:12 <elmiko> o/
15:02:30 <jpena> #chair trown dmsimard elmiko apevec
15:02:30 <zodbot> Current chairs: apevec dmsimard elmiko jpena trown
15:02:55 <social> I'm just partially here
15:02:56 <social> o.
15:03:01 <eggmaster> o/
15:03:04 <jpena> #chair social eggmaster
15:03:04 <zodbot> Current chairs: apevec dmsimard eggmaster elmiko jpena social trown
15:03:35 <derekh> apevec: lgtm
15:03:37 <derekh> o/
15:03:41 <mburned> o/ (though trown has more current info than me)
15:03:47 <jpena> #chair derekh mburned
15:03:47 <zodbot> Current chairs: apevec derekh dmsimard eggmaster elmiko jpena mburned social trown
15:04:11 <apevec> jruzicka, number80 - are you at least partially here ?
15:05:33 <number80> apevec: o/
15:05:43 <jpena> #chair number80
15:05:43 <zodbot> Current chairs: apevec derekh dmsimard eggmaster elmiko jpena mburned number80 social trown
15:06:10 <jpena> Agenda is at https://etherpad.openstack.org/p/RDO-Packaging, in case you need to add a last-minute topic
15:06:17 <dmsimard> Completely off topic but I'd really love to have the time the page http://trunk.rdoproject.org/centos7/report.html was generated. I never know which timezone it's in. I think I'll put a review in
15:06:42 <number80> dmsimard: UTC
15:07:05 <number80> I plan to add some js in this page, after pushing jinja2 review
15:07:21 <jpena> so... moving to the first topic?
15:07:23 <apevec> but please keep it "some" :)
15:07:24 <dmsimard> jpena: go
15:07:35 <apevec> agenda is go
15:07:38 <jpena> #topic stable/liberty Delorean
15:07:41 <dmsimard> sorry for sidetracking
15:08:02 <jpena> #link https://trello.com/c/VPTFAP4o/72-delorean-stable-liberty
15:08:12 <apevec> so jpena has instance running
15:08:33 <apevec> I'd like to switch centos7-liberty today
15:08:41 <apevec> currently it's symlink to centos7 i.e. Trunk
15:08:47 <jpena> I've set up a test env in http://209.132.178.247/liberty/status_report.html , using apevec's patch
15:09:14 <jpena> #link https://review.gerrithub.io/248584
15:09:23 <dmsimard> I'll switch khaleesi jobs to use centos7-liberty and open up mitaka jobs
15:09:28 <apevec> dmsimard, as noted in trello card, we'll at first manually create current-passed-ci pointing to trunk
15:09:37 <apevec> dmsimard, yes, that was next question :)
15:09:54 <dmsimard> apevec: wait what
15:10:00 <apevec> dmsimard, I think I saw current jobs use centos7-liberty already
15:10:25 <apevec> dmsimard, lemme express myself in bash :)
15:10:25 <dmsimard> apevec: there'll be current-passed-ci in centos7-liberty and one in trunk right
15:10:41 <apevec> dmsimard, yeah, but we have current users already
15:10:52 <apevec> e.g. openstack puppet upstream CI
15:11:07 <apevec> using centos7-liberty/current-passed-ci (I hope they do!)
15:11:17 <dmsimard> puppet CI uses current-passed-ci
15:11:18 <apevec> (vs /current/)
15:11:20 <apevec> cool
15:11:23 <dmsimard> not sure if it uses centos7-liberty
15:11:24 <trown> apevec: they do
15:11:34 <apevec> EmilienM, ^
15:11:37 <trown> it uses centos7-liberty
15:11:41 <apevec> excellent
15:11:41 <dmsimard> good
15:11:58 <dmsimard> he might want to have tests for mitaka too but let's chat with him later about that
15:12:18 <trown> ya, I wonder about how that will work...
15:12:22 <number80> mitaka could wait final release
15:12:46 <apevec> dmsimard, so my idea is to keep centos7-liberyt/current-passed-ci -> 86/41/864181e522efefd96f367e3174a2a0e74123532e_764a3c59
15:13:05 <apevec> i.e. latest from Sep 28
15:13:14 <apevec> dmsimard, that was manually promoted right?
15:13:23 <dmsimard> trown: gate against centos7-liberty/current-passed-ci in the stable/liberty branches of the puppet modules and then master against mitaka ?
15:13:30 <trown> dmsimard: we have a packstack job automatically updating that now?
15:13:32 <dmsimard> apevec: automatically manually promoted
15:13:34 <apevec> err wrong symlink
15:13:38 <jruzicka> apevec, here, sorry, got lost in doc
15:13:42 <dmsimard> apevec: as in, I manually promoted it with the promote script :p
15:13:58 <trown> dmsimard: ah ya, you could have a different acceptance test on the stable/liberty branch
15:14:10 <dmsimard> I'll chat with EmilienM later
15:14:22 <apevec> centos7-liberyt/current-passed-ci -> ../../../centos-master/86/41/864181e522efefd96f367e3174a2a0e74123532e_764a3c59
15:14:24 <dmsimard> apevec: also yeah, that build from sep28 is good
15:14:46 <dmsimard> apevec: only problem EmilienM reported was the selinux issue and o-s is from the common repo
15:14:48 <apevec> until we have PASS using new liberty CI job
15:14:59 <apevec> yep, new version is in -testing
15:15:48 <EmilienM> dmsimard: what's up
15:16:14 <apevec> EmilienM, we're switching centos7-liberty to actually build from stable/liberty
15:16:31 <apevec> which was brached with RC1s
15:16:35 <apevec> branched even
15:16:51 <apevec> so we're taking care not to break your CI
15:17:54 <apevec> ok, time to write down actions
15:19:01 <dmsimard> #action dmsimard to ensure khaleesi/jenkins runs off centos7-libert
15:19:19 <apevec> #action apevec/jpena switch centos7-liberty keeping Trunk current-passed-ci
15:19:28 <apevec> oh wait, that *aas thing
15:19:56 <dmsimard> EmilienM: do you have a stable/liberty branch yet ?
15:19:58 <apevec> #action derekh to review https://review.gerrithub.io/248726
15:20:00 <trown> I will make sure the rdo-manager job will use that too
15:20:16 <number80> (ping me if needed, I'm getting side-tracked internally)
15:20:19 <apevec> derekh, ^ we need that to fix version dep between neutron and *aas
15:20:24 <EmilienM> dmsimard: no
15:20:36 <EmilienM> dmsimard: we are actually waiting from you and UCA to provide stable packaging.
15:20:58 <apevec> that's next item on agenda :)
15:21:07 <apevec> (rc1 builds)
15:21:09 <derekh> apevec: ack, will do
15:21:45 <dmsimard> EmilienM: ok, when you create the stable/liberty branches, you'll want to keep gating against our centos7-liberty/current-passed-ci repo but the master branch will need to gate againsts centos7/current-passed-ci which will be mitaka
15:22:56 <dmsimard> #action dmsimard to create the mitaka jenkins CI
15:23:29 <apevec> re. rpm-liberty branches: we'll create them as late as possible
15:23:52 <apevec> i.e. only when first incompatbile upstream change happens
15:24:01 <apevec> new binary etc
15:25:01 <number80> apevec: rpm-liberty or rdo-liberty?
15:25:07 <apevec> rpm-liberty
15:25:28 <apevec> that's delorean spec for tracking stable/liberty
15:25:40 <apevec> rdo-liberty is distgit for building packages in Koji
15:25:47 <apevec> for those not in Fedora
15:26:42 <number80> aack
15:26:42 <apevec> maybe I should've picked better name for distgit branch, suggestions welcome!
15:27:08 <number80> no, it's just sickness confusing my mind
15:27:33 <jpena> next topic?
15:27:36 <apevec> but it is confusing, I just could get better name :)
15:27:43 <apevec> could NOT
15:27:48 <apevec> jpena, yes
15:27:58 <jpena> #topic RC1 in Rawhide/CBS cloud7-liberty TODAY Sep 30
15:28:06 <jpena> #link https://trello.com/c/GPqDlVLs/63-liberty-rc1-rpms
15:28:15 <jpena> apevec: your turn
15:28:47 <apevec> right, so optimistic goal is to have all RC1 rebases in Rawhide and CBS Liberty today
15:28:50 <jruzicka> apevec, there's bot r{pm,do}-liberty?
15:28:54 <jruzicka> *both
15:28:55 <apevec> now you tell me what's missing :)
15:29:03 <apevec> jruzicka, yes, see above :)
15:29:09 * jruzicka facepalms
15:29:44 <jruzicka> nvm, I'll get it in time :-p
15:29:57 <number80> apevec: ack
15:30:00 <apevec> re. missing - checklists are in https://trello.com/c/GPqDlVLs/63-liberty-rc1-rpms
15:30:15 <number80> today finishes at 12am PDT ;)
15:30:29 <apevec> number80, can you do RC1 for all storage services (glance, cinder etc) ?
15:30:34 <number80> yes
15:30:51 <apevec> btw what is that aodh thing?
15:31:29 <number80> apevec: ceilometer extracted the alarm service in a separate project
15:31:44 <number80> new ceilometer tradition: a new service per release
15:31:49 <apevec> wow spin-off craze continues
15:32:02 <apevec> well, it all started from Nova
15:32:25 <number80> I'm waiting ceilometer to take over the whole openstack project
15:32:46 <eggmaster> http://annawrites.com/blog/wp-content/uploads/2013/08/trouble-with-tribbles.jpeg
15:32:54 <jruzicka> and then... it's gonna rename itself
15:32:56 <number80> lol
15:32:58 * elmiko chuckles
15:33:13 <number80> back to business gentlemen :)
15:33:23 * jruzicka looking very serious
15:33:43 <jpena> anything else to discuss about RC1 in Rawhide?
15:33:48 <apevec> yeah, let's distribute rc1 builds, I'll be adding @name
15:33:52 <apevec> in the checklists
15:34:12 <apevec> number80, ceilo is yours too or we have new maintainer?
15:34:38 <apevec> trove rc1: vkmc ?
15:34:43 <number80> apevec: we have maintainer but for initial import, it's simpler that I do it
15:34:54 <number80> vkmc: ^
15:35:11 <vkmc> sure
15:35:16 <number80> (aodh is not in fedora dist-git)
15:35:36 <vkmc> rhel osp?
15:35:40 <apevec> ihrachys, can you do neutron RC1  in Rawhide?
15:35:44 <apevec> vkmc, no, Rawhide
15:35:51 <apevec> this is Liberty RC1
15:35:56 <vkmc> apevec, k, thanks
15:36:00 <trown> so services will be in rawhide?
15:36:35 <trown> I thought we were only doing clients and deps in fedora
15:36:43 <apevec> trown, those which are already in Fedora
15:36:48 <ihrachys> apevec: rawhide? really?
15:37:04 <ihrachys> I thought we switch to rpm-liberty?
15:37:05 <ihrachys> rdo-liberty, sorry
15:37:08 <trown> ah ok, so just not new stuff
15:37:26 <apevec> ihrachys, yes, but not yet for those existing in Fedora
15:37:35 <apevec> we'll rebase them then remove :)
15:37:41 <apevec> how silly is that?
15:37:46 <trown> hehe
15:37:47 <ihrachys> apevec: why
15:37:58 <ihrachys> what's going on? who are all those people?
15:38:12 <apevec> ihrachys, well, you can do RC1 rebase only in rdo-liberty
15:38:46 <ihrachys> on my way, I presume. I don't build packages, right? just rebase
15:38:48 <apevec> until we're ready to officially remove openstack-* I'll keep Fedora master in sync
15:38:53 <apevec> ihrachys, yep
15:39:10 <apevec> I'll take care of CBS builds once I see rebase in distgit
15:39:29 <ihrachys> sick
15:39:43 <apevec> yep that's me
15:39:46 * ihrachys on rc1 for rdo-liberty
15:40:12 <apevec> next Heat: zaneb ryansb - can anyone of you do RC1 rebase today?
15:40:18 <apevec> if not let me know!
15:40:34 <apevec> Sahara - elmiko ?
15:40:47 <apevec> ah no egafford ^
15:40:53 <elmiko> do we need a rebase of sahara as well?
15:41:05 <apevec> yes, all services we currently have
15:41:18 <elmiko> k, i'll talk with egafford
15:41:30 <apevec> thanks
15:41:35 <elmiko> i'm not sure if we've cut the upstream rc1, i *think* we have
15:41:51 <apevec> I think I saw it, lemme check
15:42:08 <elmiko> yea, it's up
15:42:17 <elmiko> https://launchpad.net/sahara/liberty/liberty-rc1/+download/sahara-3.0.0.0rc1.tar.gz
15:42:27 <egafford> apevec: Hi there.
15:42:37 * elmiko waves at egafford
15:42:58 * egafford waves back.
15:43:34 <trown> anything else for RC1?
15:44:00 <apevec> Heat
15:44:03 <apevec> Ironic
15:44:17 <trown> who is the maintainer for Ironic?
15:44:21 <apevec> and some clients
15:44:37 <apevec> trown, rdoinfo says: - project: ironic
15:44:38 <apevec> conf: core
15:44:38 <apevec> maintainers:
15:44:38 <apevec> - athomas@redhat.com
15:44:46 <apevec> which I suspect is not accurate?
15:44:59 <trown> ya, that I can take that over
15:45:16 <apevec> trown, please send rdoinfo PR
15:45:20 <trown> will do
15:45:48 <number80> trown: check the clients too
15:45:55 <trown> #action trown send PR to take over maintenance of Ironic
15:46:11 <number80> quite ironic, eh
15:46:48 <apevec> ok, re. Heat I'll ping ryansb / zaneb later
15:47:18 <apevec> #action apevec to ping missing mainters for RC1 rebases
15:47:39 <zaneb> ryansb is so good at rebases
15:48:01 <trown> lol
15:48:02 <number80> zaneb: good occasion to get better ;)
15:48:02 <apevec> zaneb, my thinking as well :)
15:48:21 <ihrachys> apevec: so how do I bump? rdopkg new-version apparently does not behave from rdo-liberty
15:48:34 <apevec> jruzicka, ^ fix it :)
15:48:54 <apevec> ihrachys, that's why I did rebase on master ;)
15:49:10 <apevec> ihrachys, just rename it temp locally to "master"
15:49:12 <ihrachys> apevec: that's bad excuse, especially for *aas that are not in fedora
15:49:13 <apevec> to make rdopkg happy
15:49:28 <apevec> rdopkg is easy to cheat
15:50:35 <apevec> ihrachys, let jruzicka know how it breaks
15:50:50 <number80> jruzicka: reminds me we need a switch for projects that have no patches
15:51:12 <number80> (--no-patches-dont-ask-for-a-branch)
15:51:20 <apevec> and add issue in https://github.com/redhat-openstack/rdopkg/issues/
15:51:34 <number80> I already did ^^
15:51:42 <number80> one week ago
15:52:00 <apevec> number80, next are ping/bribe/send patch to jruzicka  :)
15:52:08 <apevec> whatever you prefer
15:52:36 <apevec> we could make tooling topic for the next meeting
15:52:51 <apevec> to collect RFE ideas and set priorities
15:52:55 <apevec> jruzicka, ^ ?
15:52:56 <number80> yup
15:53:11 <apevec> ok, getting off topic, let's move on
15:53:15 <apevec> jpena, next please!
15:53:24 <jpena> let's go
15:53:35 <jpena> #topic Package needs version bump
15:53:43 <jruzicka> oh yes
15:53:54 <jruzicka> no patches branch, that's been on the table for quite some time
15:53:57 <jpena> I think this topic was raised by chandankumar
15:54:33 * apevec looks
15:55:00 <apevec> chandankumar is not around, but yeah that topic is clear, we need those bumps
15:55:19 <apevec> os-brick is reviewed?
15:55:37 <jpena> no, I'm working on it, number80 is reviewing
15:55:40 <apevec> what is falcon?
15:55:55 <number80> wtf?
15:56:45 <number80> it's been built a month ago ...
15:56:45 <jpena> I assume it is python-falcon
15:56:47 <apevec> it's updated in Rawhide, I'll check CBS
15:57:21 <apevec> yeah, needs update in cloud7-common
15:58:30 <apevec> designateclient update - ihrachys ?
15:58:48 <apevec> NB clients do stay in Fedora
15:58:56 <number80> apevec: this one is vkmc's
15:59:04 <vkmc> yeah ^
15:59:10 <apevec> ok
15:59:31 <apevec> please add yourself as maintainers: in rdoinfo
15:59:31 <ihrachys> apevec: meh. I think I will start to search for a replacement for myself in packaging world. I am depressed by it.
16:00:22 <apevec> ihrachys, sorry to hear that, but anyone you recommend will be welcome
16:00:55 <ihrachys> apevec: I just fail to keep up with all packages I got to maintain.
16:01:14 <jruzicka> ihrachys, that does sound like too many packages, then ;)
16:01:16 <ihrachys> those clients, I completely forgot about them. and I also have python-networking-*
16:01:52 <ihrachys> jruzicka: or maybe there's too much process around RDO
16:02:30 <jruzicka> ihrachys, lolno
16:02:44 <jruzicka> there's too much change but that's unevitable
16:02:45 <apevec> ihrachys, jruzicka is hard at working automating all that
16:03:07 <jruzicka> few more years to stabilise, automate and mature, then you won't even know there's a process :-p
16:03:21 <number80> well, I understand the concern and I'd like to adress it but not the right time :(
16:03:21 <trown> :)
16:03:38 <number80> (but yes, automation is the answer)
16:03:53 <number80> let's finish the meeting first
16:03:55 <jruzicka> yes, yes, you'll all get fired once I finish.
16:03:57 <jruzicka> </joke>
16:04:10 <jpena> let's move on then
16:04:19 <apevec> jruzicka, no, we'll get to debug your toolz
16:04:25 <jpena> #topic New Package python-reno
16:04:33 <jruzicka> apevec, that should be oh so pleasant :)
16:04:45 <number80> jpena: this is no requirements
16:04:46 <jpena> chandankumar has started the review, https://bugzilla.redhat.com/show_bug.cgi?id=1267496
16:04:55 <apevec> ah automated RElNotes
16:05:07 <apevec> this will be used for upstream stable relnotes
16:05:17 <jpena> so it's needed, isn't it?
16:05:19 <apevec> and we should use it in our automation for stable updates
16:05:21 <apevec> yes!
16:05:26 <trown> cool
16:05:32 <number80> btw, can we make a rule that every single review does not require a topic?
16:05:48 <number80> (or we'll end up with 3h meetings)
16:05:48 <apevec> number80, ack - unless it's problematic
16:05:57 <number80> *mods*
16:06:14 <apevec> but this was good to point out, as part of automation topic
16:06:18 <number80> yes
16:06:19 <apevec> (which was offtopic :)
16:06:23 <apevec> anyway
16:06:56 <apevec> same for the next topic about SCM requests
16:07:14 <apevec> that's part of normal procedure, nothing to discuss JFDI
16:07:22 <number80> +1
16:07:31 <jpena> ok, nothing else to discuss about python-reno or SCM requests... We're almost there
16:07:37 <number80> yes
16:07:46 <jpena> #topic chair rotation for next meeting
16:07:53 <jpena> any volunteer?
16:07:55 <trown> I will take it
16:07:59 <apevec> THANKS!
16:08:01 <number80> amen
16:08:10 <jpena> #action trown to chair next meeting
16:08:19 <jpena> thanks trown ;)
16:08:27 <jpena> #topic open floor
16:08:28 <trown> thanks jpena for chairing today
16:08:34 <number80> +2
16:08:39 <apevec> +V
16:08:48 <apevec> Merge!
16:08:59 <number80> it's good to see some rotation
16:09:04 <elmiko> i'm curious about the state of barbican
16:09:37 <apevec> is it still under review?
16:09:45 <trown> I have rdo-manager manually installing via delorean liberty packages, working on getting the khaleesi job working
16:09:51 <elmiko> i *thought* we had a good packages build
16:10:14 <apevec> elmiko, it is in delorean trunk, yes
16:10:23 <apevec> and builds there
16:10:28 <apevec> trown, good news!
16:10:34 <jpena> btw, barbican is failing on centos7-kilo
16:10:44 <elmiko> apevec: does that mean it's slated for release in rdo?
16:10:51 <elmiko> jpena: well that stinks =(
16:11:01 <jpena> elmiko: http://trunk.rdoproject.org/kilo/centos7/64/67/6467361a5d5bd78203c70b3ac174a04870f17bbb_cfc0615c/rpmbuild.log
16:11:10 <apevec> trown, dmsimard - would be good to include rdo-mgr to vote in delorean CI promotion job
16:11:14 <elmiko> jpena: thanks, i'll take a look
16:11:34 <trown> apevec: ya that would be the goal after getting a job up and passing
16:11:40 <dmsimard> apevec: we're not quite there yet
16:11:47 <dmsimard> apevec: it's in the todo, though
16:12:09 <apevec> elmiko, if review has been done, then yes
16:12:27 <elmiko> apevec: ok, thanks
16:13:39 <apevec> elmiko so looks like number80 finished pkg review https://bugzilla.redhat.com/show_bug.cgi?id=1190269
16:13:56 <elmiko> yea, i thought we had finished that portion
16:14:07 <number80> apevec, elmiko it's been imported
16:14:07 <apevec> elmiko, next step is for Greg to do SCM request
16:14:18 <apevec> number80, oh how?
16:14:32 <apevec> I don't se SCM request
16:14:53 <elmiko> xaeth_afk: ^^
16:14:57 <apevec> review done Sep1
16:15:01 <number80> apevec: in github only
16:15:04 <apevec> ping Sep 14
16:15:09 <apevec> today Sep 30
16:15:20 <apevec> bad email filtering?
16:15:28 <apevec> BZ => spam
16:15:33 <elmiko> i'll follow up with xaeth (Greg Swift) about the scm req.
16:15:53 <apevec> thanks
16:16:06 <trown> are we still in open floor, or should we end meeting?
16:16:06 <apevec> #action elmiko  follow up with xaeth (Greg Swift) about the scm req for openstack-barbican
16:16:17 <apevec> uhm yes, we're overtime
16:16:20 <jpena> so, if there is nothing left, we will end the meeting in 3...
16:16:21 <apevec> jpena, let's close
16:16:31 <elmiko> sorry, but thanks for answering my questions =)
16:16:39 <jpena> #endmeeting