15:00:20 <ykarel> #startmeeting RDO meeting - 2017-07-12
15:00:23 <openstack> Meeting started Wed Jul 12 15:00:20 2017 UTC and is due to finish in 60 minutes.  The chair is ykarel. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:27 <openstack> The meeting name has been set to 'rdo_meeting___2017_07_12'
15:00:34 <ykarel> Please add your topic to https://etherpad.openstack.org/p/RDO-Meeting if not already added
15:00:42 <apevec> dmsimard, can you try first with docker-lastest in Extras ?
15:00:51 <apevec> docker-latest
15:00:57 <ykarel> #topic Roll call
15:00:59 <apevec> it's parallel-installable
15:01:06 <apevec> o/
15:01:06 <rbowen> o/ Here
15:01:08 <chandankumar> \o/
15:01:18 <dmsimard> apevec: ah, docker-latest is 1.13.1, I'll try that.
15:01:20 <amoralej> o/
15:01:30 <apevec> dmsimard, docker is also from Extras, not base
15:01:34 <PagliaccisCloud> yo
15:01:41 <ykarel> #chair rbowen apevec chandankumar amoralej
15:01:42 <openstack> Current chairs: amoralej apevec chandankumar rbowen ykarel
15:01:50 <ykarel> #chair PagliaccisCloud
15:01:52 <openstack> Current chairs: PagliaccisCloud amoralej apevec chandankumar rbowen ykarel
15:02:18 <rdogerrit> David Moreau Simard created rdo-infra/ansible-role-rdo-kolla-build master: Use docker-latest (1.13.1) instead of docker (1.12.6)  https://review.rdoproject.org/r/7615
15:02:40 <ykarel> so first topic
15:02:43 <ykarel> #topic mass scratch rebuilds script [hguemar]
15:03:04 <ykarel> number80, yt?
15:03:27 <number80> o/
15:03:35 <ykarel> #chair number80
15:03:36 <openstack> Current chairs: PagliaccisCloud amoralej apevec chandankumar number80 rbowen ykarel
15:04:49 <dmsimard> apevec: hmm. I also have another idea, let's discuss after meeting :)
15:05:07 <ykarel> number80, you added this topic: mass scratch rebuilds script [hguemar]
15:06:09 <number80> Yes, we had reports that some dependencies don't build currently, so I'll work on a script that will launch a scratch build for every packages in master to detect FTBFS
15:06:38 <number80> so we'll be able to spot them
15:06:42 <rbowen> Sweet
15:06:54 <jschlueter> amoralej, apevec: in RDO CI do we have any jobs running tempest
15:07:03 <apevec> jschlueter, meeting
15:07:19 * jschlueter waits
15:08:01 <number80> EOF from me
15:08:04 <rdogerrit> Alan Pevec proposed rdoinfo master: Update SQLAlchemy to 1.1.11 for Pike  https://review.rdoproject.org/r/7591
15:08:04 <rdogerrit> Alan Pevec created rdoinfo master: Update openvswitch to 2.7.1 for Pike  https://review.rdoproject.org/r/7616
15:08:11 <ykarel> next topic
15:08:14 <ykarel> #topic     Reminder: https://review.rdoproject.org/etherpad/p/rdo_low_hanging_fruits and https://github.com/redhat-openstack/easyfix/issues for "nice to have" or "easy fix" issues that are not getting addressed
15:08:21 <ykarel> chandankumar, ^^
15:08:56 <chandankumar> ykarel: i have just started adding it, i have 4-5 issues in queue for the same
15:09:05 <chandankumar> would be good fo getting started
15:09:13 <chandankumar> *for
15:09:25 <chandankumar> that's it from myside.
15:09:26 <amoralej> number80, we can run dlrn bootstrap from time to time
15:09:31 <rbowen> I'll be promoting it some more in the coming weeks, and in the newsletter (which is very late this month, but still coming)
15:10:00 <number80> amoralej: it's more about dependencies that are the issues but yes, we can do that too +1
15:10:04 <ykarel> chandankumar, looks good
15:10:26 <rbowen> There's probably also some tickets in https://github.com/redhat-openstack/website/issues that are suitable for moving over.
15:10:27 <number80> good job chandankumar
15:11:36 <ykarel> So next topic
15:11:46 <ykarel> #topic Status of pike preparation
15:11:51 <ykarel> amoralej, ^^
15:11:52 <amoralej> that's mine
15:12:02 <amoralej> pike dates are becoming closer :)
15:12:14 <apevec> closer than ever!
15:12:19 <amoralej> i'm running first bootstrap in new centos-pike already
15:12:30 <amoralej> https://trunk.rdoproject.org/centos7-pike-bootstrap/report.html
15:12:50 <amoralej> then, i'll run head-only maybe after latest version of libraries
15:13:10 <dmsimard> ah, yes, DLRN renames and carry-over have always been tricky
15:13:12 <dmsimard> good job
15:13:25 <amoralej> it should be a bit easier this time, but yes
15:13:33 <amoralej> transition is always tricky
15:13:48 <rdogerrit> Merged openstack/magnum-distgit rpm-master: switch from oslosphinx to openstackdocstheme  https://review.rdoproject.org/r/7602
15:14:01 <number80> Yep
15:14:05 <amoralej> so, branches will be cut at RC1, iirr
15:14:28 <amoralej> which is in second week of aug
15:14:37 <apevec> yes, first RC is branching point
15:14:38 <amoralej> and at that time we should move promotion jobs
15:14:38 <apevec> upstream
15:14:46 <amoralej> right?
15:15:05 <apevec> amoralej, hopefully based on dlrn api
15:15:08 <apevec> amoralej, hopefully based on dlrn api
15:15:21 <apevec> i.e. no more single pipeline
15:15:25 <amoralej> mmmm, is anyone actively working on it?
15:15:33 <apevec> good question
15:15:36 <amoralej> we need upstream promotion also
15:15:45 <apevec> weirdo is reporting to dlrn api
15:15:54 <apevec> need to check oooq jobs
15:16:02 <apevec> amoralej, periodic?
15:16:05 <dmsimard> The fact that jobs are reporting does not mean we can promote with it
15:16:07 <amoralej> will new upstream promotion running in review.r.o be ready for then
15:16:08 <amoralej> yes
15:16:16 <apevec> that's being moved to review.rdo
15:16:23 <amoralej> dmsimard is right
15:16:24 <dmsimard> We need to write *something* to promote repositories based on criterias
15:16:32 <dmsimard> I discussed this already with weshay and adarazs
15:16:32 <amoralej> reporting doesn't mean we have it ready
15:16:43 <amoralej> and some logic to detect the hashes to test
15:16:43 <dmsimard> It should not be too hard to write, but someone needs to do it
15:16:45 <apevec> yes but it's first step
15:17:09 <apevec> amoralej, it all starts from "consistent" as it does now
15:17:14 <dmsimard> tbh I'm super busy with containers and migrations so I won't have time short term to work on promotions
15:17:22 <amoralej> consistent is ok
15:17:30 <amoralej> next step is current-tripleo
15:17:47 <amoralej> when is expected to have periodic in review.r.o?
15:17:48 <rdogerrit> Christopher Brown proposed openstack/senlin-distgit rpm-master: Inital import of senlin spec file  https://review.rdoproject.org/r/7470
15:18:11 <apevec> need to check w/ panda and weshay
15:18:15 <amoralej> ok
15:18:38 <amoralej> then if we want to have new current-tripleo-rdo based on dlrn api we have some work to do
15:18:45 <dmsimard> rlandy got multi node jobs working yesterday on review.rdo
15:18:54 <dmsimard> I'm not sure where we are on OVB
15:18:58 <dmsimard> but progress is good
15:19:33 <amoralej> i can spend some time on implementing jobs with dlrn-api but i'd like to progress with oooq generic jobs, (dmsimard i'll need your help with it, too)
15:20:04 <ykarel> dmsimard is without chair for too long
15:20:08 <ykarel> #chair dmsimard
15:20:09 <openstack> Current chairs: PagliaccisCloud amoralej apevec chandankumar dmsimard number80 rbowen ykarel
15:20:19 <dmsimard> amoralej: weshay said he would be able to work (or get someone to work) on implementing DLRN API reporting from oooq jobs
15:20:42 <amoralej> #action amoralej to check how hard/easy would be to implement promotion pipeline with dlrn-api
15:21:11 <dmsimard> amoralej: can you share your findings with adarazs and I ?
15:21:18 <amoralej> dmsimard, sure
15:21:29 <Duck> quack o/
15:21:34 <dmsimard> amoralej: the only challenge to think about is pre-required steps for job running
15:21:35 <ykarel> #chair Duck
15:21:36 <openstack> Current chairs: Duck PagliaccisCloud amoralej apevec chandankumar dmsimard number80 rbowen ykarel
15:21:41 <dmsimard> amoralej: for example containers and images must be built prior
15:21:43 <amoralej> and i'll add the topic of pike promotion to next tripleo-ci meeting
15:21:46 <dmsimard> amoralej: regardless if we have dlrn api or not
15:22:11 <amoralej> dmsimard, yes, we need to think in how each job will discover what hash should it use
15:22:27 <amoralej> that will have implicit dpendencies
15:22:29 <amoralej> between jobs
15:23:45 <amoralej> i think that's it from my side about pike preparation so far
15:24:04 <ykarel> so can we move to next topic
15:24:32 <amoralej> i think so
15:24:34 <ykarel> #topic Update of CBS automation
15:24:42 <ykarel> amoralej, that's also yours
15:24:53 <amoralej> that's mainly for awareness
15:25:13 <amoralej> so, cbs tagging automation is in place
15:25:23 <rdogerrit> David Moreau Simard proposed rdo-infra/ansible-role-rdo-kolla-build master: Use docker-latest (1.13.1) instead of docker (1.12.6)  https://review.rdoproject.org/r/7615
15:25:47 <ykarel> amoralej, any link?
15:26:00 <amoralej> #info anyone can propose promotion of CBS tags by sending reviews to rdoinfo, example https://review.rdoproject.org/r/#/c/7534/
15:26:51 <amoralej> so, now we have automated CBS builds for new releases in stable versions and automated promotion
15:27:17 <amoralej> which is nice, but we still need to improve some things
15:27:30 <amoralej> and note that dependencies builds are still manual
15:27:33 <dmsimard> amoralej++
15:27:38 <amoralej> anyway i think it's a step ahead
15:27:55 <apevec> yes, it is big step forward!
15:28:04 <dmsimard> amoralej: this will help us keeping stable releases up to date so much
15:28:11 <apevec> ...towards getting us all out of the job!
15:28:13 <dmsimard> it's been a huge pain point, good job
15:28:23 <amoralej> yeah, i really hope so
15:28:27 <rdogerrit> Merged rdo-infra/ci-config master: Revert "Expose the kolla_force_build parameter in the generic build job"  https://review.rdoproject.org/r/7600
15:29:03 <amoralej> and that's what i had
15:29:09 <ykarel> Ok
15:29:16 <ykarel> #topic chair for next meeting
15:29:44 <amoralej> i can chair
15:29:54 <ykarel> #action amoralej to chair next meeting
15:30:03 <apevec> that was quick sell :)
15:30:03 <ykarel> #topic open floor
15:30:07 <Duck> quack
15:30:11 <rdogerrit> Merged rdo-infra/ci-config master: Also collect kolla logs when destroying virtual machines  https://review.rdoproject.org/r/7601
15:30:26 <Duck> dmsimard: so is it ok for the web builder RDO Cloud allocation?
15:30:35 <Duck> Google crashed when opening the doc again
15:30:51 <dmsimard> Duck: yes, we requested additional quotas but let's include it in the spreadsheet
15:31:00 <Duck> fine :-)
15:31:13 <Duck> will move it soon then
15:32:05 <Duck> dmsimard: time next week to play Ansible?
15:32:35 <Duck> I'm officially requestion a slice of your time :-)
15:32:37 <dmsimard> Duck: I've started here and there slowly to fix some things
15:32:55 <dmsimard> Duck: I also have a draft about how we want to use vault
15:33:10 <Duck> please involve me in
15:33:15 <dmsimard> Duck: will do
15:34:38 <rbowen> Duck: Any progress on moving mailing lists?
15:34:50 <rbowen> Duck: Are you waiting on me for anything about that?
15:34:59 <Duck> I need to ask you some questions
15:35:07 <rbowen> ok. I'm available any time you need.
15:35:24 <rbowen> We also have the infra list, which I forgot about, but that's broadcast only, so should be easy.
15:35:25 <Duck> Also I'm also waiting for my hyperkitty ticket to be replied
15:35:33 <rbowen> ok, good.
15:35:36 <Duck> because there's migration problems remaining
15:35:41 <rbowen> But if Im holding it up, please let me know.
15:36:09 <Duck> If you can push abompard forward, it would help :-)
15:36:21 <Duck> he took over the ticket but nothing's happening
15:36:29 <rbowen> ok, I'll see what I can do. :-)
15:36:44 <rbowen> Meanwhile, Apache Pony is always an alternative option. :-)
15:36:52 <rbowen> http://ponymail.apache.org/
15:37:16 <Duck> that's a fonrtend alternative yet, but that does not solve migration
15:37:23 <rbowen> Oh, right. Gotcha
15:37:40 <Duck> except if you just want the archives without ML setups and users
15:38:12 <Duck> and we have all the rules for hyperkitty working (on fresh sites and import that worked)
15:38:21 <rbowen> Well, what we talked about in this meeting was just leaving the archives as they are, and starting fresh.
15:38:23 <Duck> let me find the link
15:38:35 <Duck> really?
15:38:36 <rbowen> If we *can* migrate the archives, obviously we want to, but if it's a blocker, we can just do a fresh start.
15:38:51 <Duck> daring :-)
15:39:03 <rbowen> Did I interpret that right, folks? That we're ok with just doing a fresh start and referencing the old archives as they are?
15:39:21 <Duck> rbowen: https://gitlab.com/mailman/hyperkitty/issues/144
15:39:37 <dmsimard> rbowen: well you don't have consensus or quorum from just chatting with Duck :p
15:39:58 <rbowen> I mean from when we discussed it in this meeting a month ago.
15:40:03 <Duck> anyway having this software improved would be nice
15:40:15 <rbowen> indeed
15:40:23 <number80> 963456
15:40:27 <dmsimard> rbowen: Oh, I might have missed that. I don't know -- I don't personally mind so long as they're easy to find
15:40:28 <number80> SORRY
15:40:29 <rbowen> I'll bug Aurelien and see what we can make happen.
15:40:30 <apevec> rbowen, that's what I remember
15:40:43 <apevec> keep old archives where they are
15:41:09 <apevec> we can lookup in meeting minutes
15:41:42 <rbowen> I can try to get broader consensus on the mailing list, but, yeah, that's what I remember from the meeting.
15:41:56 * number80 nods
15:42:04 <Duck> let's ask the community
15:42:28 <Duck> ho
15:42:30 <rbowen> I'll start a thread on the mailing list --- or wake up the old one.
15:42:41 <Duck> is anyone proficient in Copo building?
15:42:59 <Duck> I need a fixed package for Mailman3
15:43:13 <Duck> one of the deps breaks the external auth
15:43:36 <Duck> and I've got a strange failure so I would be happy to get help
15:43:45 <number80> Duck: mail me
15:43:51 <Duck> "Copr" sorry
15:43:58 <number80> Got it :0
15:44:09 <number80> I'll fix it for you
15:44:10 <Duck> hey, difficult to type with wings :-)
15:44:16 <Duck> thanks number80
15:44:22 <apevec> Cop:reloaded ? what does it mean anyway...
15:44:27 <number80> #action number80 help Duck with mailman3 package
15:44:33 <Duck> (it built fine on a VM so it should…)
15:45:05 <Duck> apevec: Copr, the build service around Fedora
15:45:34 <apevec> yeah, I know
15:45:41 <apevec> I'm just badly trying to be funny
15:45:51 <number80> :)
15:45:52 <apevec> Duck, it's where RDO started :)
15:46:10 <apevec> actually, no, it started in EPEL
15:46:22 <apevec> then removed after breaking its rules
15:46:24 <Duck> number80: http://copr.fedorainfracloud.org/coprs/duck/osas-infra-team-rpm-repo/package/python-urllib3/
15:46:46 <Duck> the other option is to take your package in your repo but I'd like to learn and understand
15:46:56 <number80> Boys are born amongst cabbages, girls amongst roses and RDO in the middle of apples :)
15:47:17 <snecklifter> blimey, thats deep
15:47:17 * Duck :-))
15:47:41 <number80> Duck: grant me access to your copr (FAS: hguemar)
15:48:05 <Duck> ok
15:48:28 <Duck> number80: I think you need to request it on the UI first no?
15:48:54 <number80> maybe it changed then
15:49:11 <Duck> number80: yep, that's what it says on the "permissions' tab
15:49:11 <number80> oh yeah
15:49:16 <number80> done
15:49:43 <Duck> number80: done
15:49:58 <ykarel> so anything else to discuss
15:50:18 <Duck> no Sir
15:50:30 <rdobot> [sensu] RESOLVED: master.monitoring.rdoproject.org - check-delorean-master-current @ http://tinyurl.com/ybrtapm3 |#| No build failures detected: http://trunk.rdoproject.org/centos7-master/report.html
15:51:02 <ykarel> So we can close early, still 10 minutes left
15:52:25 <ykarel> Thanks all
15:52:27 <ykarel> #endmeeting