16:00:18 <sdake> #startmeeting kolla
16:00:22 <openstack> Meeting started Wed Jul 27 16:00:18 2016 UTC and is due to finish in 60 minutes.  The chair is sdake. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:26 <openstack> The meeting name has been set to 'kolla'
16:00:30 <sdake> #topic rollcall
16:00:34 <coolsvap> o/
16:00:35 <akwasnie> hi all :)
16:00:35 <duonghq> o/
16:00:37 <sdake> o/
16:00:43 <inc0> o/
16:00:48 <rhallisey> hi
16:01:01 <hrito> o/
16:01:34 <srwilkers_> o/
16:02:10 <sdake> notice some fresh faces - welcome duonghq hrito srwilkers_
16:02:11 <lrensing_> o/
16:02:16 <wirehead_> o/
16:02:22 <srwilkers_> thanks :D
16:02:25 <duonghq> nice to see you
16:02:50 <sdake> #topic announcements
16:03:07 <sdake> Heidi was kind enough to make a video of the mascot selection process
16:03:37 <sdake> https://zoom.us/recording/play/Adnpi67_uwd1qw3GZsD_CTvXj0hjSPQd48dHRxvCR9IWKzyT-bjFnIW7BtZYZUKo
16:04:11 <sdake> when someone has finished the video let me know pls :)
16:04:44 <rhallisey> we won :)
16:05:07 <rhallisey> spoiler
16:05:12 <Daviey> Sorry, i'm late chaps.
16:05:21 <inc0> yay
16:05:31 <inc0> so no more honey badger?
16:05:37 <sdake> inc0 the bloodshot koalla bear with glue involved is off the table :)
16:05:51 <inc0> so I was thinking
16:06:07 <inc0> koala bear gluing openstack squarey logos witha  glue
16:06:09 <inc0> no sniffing
16:06:19 <sdake> thats the only announcement i have - anyone from community have anything?
16:06:37 <Daviey> kolla bear is what i always think of... but to bring inc0's idea in, it should be sniffing glue
16:06:57 <inc0> :D
16:07:05 <srwilkers_> +1
16:07:14 <inc0> it is a junky animal anyway
16:07:39 <sdake> fwiw we dont have input into the mascot creation process
16:07:40 <inc0> anyway, let's call it our secret mascot
16:07:47 <inc0> duh
16:07:49 <wirehead_> Drop bears.
16:07:55 <sdake> ;)
16:08:11 <rhallisey> honey badger
16:08:16 <inc0> narwhal
16:08:19 <sdake> #topic newton-3
16:08:32 <wirehead_> A drop bear sniffing glue is all things to all people.  At least all of the people who count.
16:08:51 <sdake> #link https://launchpad.net/kolla/+milestone/newton-3
16:09:14 <sdake> we have a whole lot of stuff planned for milestone 3
16:09:39 <sdake> it seems unreasonable to think we will finish all of that work
16:10:04 <sdake> milestone 3 will be a feature freeze
16:10:17 <sdake> versus past releases where we put features in the rcs
16:10:25 <sdake> we wont be doing that at all this cycle
16:10:27 <coolsvap> yes we have many bugs which are either not updated or not valid anymore
16:10:33 <sdake> because we applied for stable:follows-policy
16:11:15 <sdake> coolsvap at this point i'm more concerned about the features landing in m3 then fixing bugs
16:11:18 <coolsvap> and i think 10+ bugs targeted as a result of docimpact
16:11:36 <coolsvap> sdake: agree
16:11:49 <sdake> the problem with that view is people look at it, go "too much" and feel overwhelemd and don't know where to start
16:12:10 <sdake> in past cycles 2 weeks before milestone 3, we cut all the features that were not in good progress or better
16:12:17 <sdake> (liberty/newton)
16:12:25 <sdake> liberty/mitaka I mean
16:12:32 <sdake> i htink we should stick to that
16:12:49 <sdake> any objections?
16:13:42 <sdake> ok well i dont see any alternative anyway so objections probably8 don't matter :)
16:14:19 <sdake> anyone willing to tackle some of the bug triage?
16:14:40 <coolsvap> I am
16:15:02 <sdake> cool
16:15:05 <sdake> yo udont have to do it all
16:15:12 <sdake> but any way you can help is appreciated
16:15:31 <coolsvap> :)
16:15:44 <sdake> #topic customization of  dockerfiles
16:15:57 <sdake> inc0 did you add this to the agenda?
16:16:04 <inc0> that's old agenda
16:16:11 <sdake> no its definately new
16:16:12 <inc0> but, patchsets are there
16:16:19 <sdake> ya lets discuss that
16:16:31 <inc0> I'll review them all today
16:16:40 <sdake> about half the customization patches are in the queue
16:16:54 <inc0> also I'll work on customization of base, it's by far hardest
16:17:03 <sdake> i'd like an ack from inc0 on each of em if he can commit to that
16:17:28 <sdake> inc0 is that possible or too much?
16:17:37 <inc0> possible, will do
16:17:40 <sdake> cool
16:17:47 <sdake> I just want to make sure they are precisely correct
16:17:57 <sdake> since if we break this now, it not super ideal :)
16:18:25 <sdake> #topic repo split
16:18:42 <sdake> ryan's vote expired today without consensus
16:18:51 <sdake> that siad, I think we should try voting again on this point
16:18:56 <sdake> maybe the core team was AFK
16:19:03 <sdake> its summer time
16:19:08 <sdake> lots of PTO and whatnot
16:19:29 <sdake> rhallisey you mind shooting out a new voting email in 1-2 weeks?
16:19:34 <rhallisey> ya will do
16:19:55 <sdake> probably right now is not the right time
16:19:56 <rhallisey> just keep an eye out for that everyone
16:19:58 <sdake> because we are still testing the backportss review policy
16:20:03 <rhallisey> roger
16:20:14 <sdake> i think the backports review policy seems to be working
16:20:15 <coolsvap> I am not still not sure about my views about it
16:20:37 <sdake> coolsvap I htink that will factor into everyone's voting
16:20:39 <coolsvap> but lets revisit it in couple of weeks
16:21:27 <sdake> #topic kolla-kubernetes
16:21:29 <rhallisey> ya we can debate further on the ML
16:21:44 <sdake> rhallisey have an update for us?
16:22:02 <rhallisey> so still blocked on the api_interface patches
16:22:15 <sdake> that was my fault right
16:22:26 <rhallisey> I mean you raised a good point
16:22:39 <rhallisey> but I don't think we should fully address the config issue now
16:22:50 <rhallisey> because of the repo split/config split
16:22:55 <sdake> can you commit to  the following
16:22:59 <wirehead_> I think we need to accept that we're looking for the "right now, so that Kolla-kubernetes doesn't have a giant pile of carried patches" solution.
16:23:05 <wirehead_> Instead of the "end" solution.
16:23:06 <rhallisey> ^ yes
16:23:12 <rhallisey> that's my point
16:23:18 <wirehead_> The reviews have been fixating on the end solution.
16:23:25 <sdake> "I wont add any new variables to globals.yml after this variable"?
16:23:38 <sdake> rhallisey see above question can yo ucommit ^^
16:23:47 <rhallisey> before the config split, it's unlikely
16:24:01 <sdake> no i mean before milestone3
16:24:10 <rhallisey> before m3 no
16:24:20 <wirehead_> So, if you look at my Nova patch, you'll see all of the variables we're altering and adding to get things going.
16:24:25 <sdake> ok well i'll change my vote to unblock things
16:24:42 <sdake> wirehead_ the issue is putting things in globals.yml is permanentn
16:24:56 <sdake> wirehead_ because of how openstack deprecation policies work
16:25:13 <sdake> if we do rework on globls.yml its a huge pita
16:25:25 <sdake> that was the basis of my complaint on ryan's patch
16:25:25 <sdake> i haven't seen your nova patch
16:26:09 <sdake> here we are making a semi-permanent compromise to unblock your work
16:26:16 <sdake> that we can possibly undo later
16:26:33 <sdake> but if 5 variables needed in globals.yml - thats a whole nother world of pain :)
16:27:07 <wirehead_> Well, the way that we're progressing forwards right now is that Kolla-kubernetes requires you to add stuff to globals.yml.
16:27:07 <sdake> wirehead_ mind linking your review for after meeting to see what the issue is
16:27:41 <wirehead_> We're OK with kolla-kubernetes users needing to paste some lines into globals.yml and Kolla users having an unmodified globals.yml file?
16:27:42 <sdake> wirehead_ ok i get that part - but ryan's addition was a nonessential variable
16:28:03 <sdake> wirehead_ ya actually thats a fantastic solution
16:28:31 <wirehead_> https://review.openstack.org/#/c/347572/
16:28:36 <sdake> kolla-kubernetes is on a release:cycle_independent model
16:28:56 <rhallisey> sdake, if that's the case we can remove the global var
16:29:06 <rhallisey> but the j2 login underneath will need to stay
16:29:14 <sdake> which means if we dont polute the globals.yml namespace, no harm is done to kolla-ansible
16:29:25 <wirehead_> Yeah.
16:29:28 <wirehead_> We can do that for https://review.openstack.org/#/c/327925/
16:29:30 <sdake> j2 login
16:29:38 <wirehead_> j2 logic
16:29:49 <sdake> i'll have a look after meeting
16:29:50 <rhallisey> logic
16:30:08 <sdake> thanks wirehead for solving that problem :)
16:30:47 <wirehead_> Oh, and Nova's control plane is starting to work as a Kube service. :)
16:30:55 <sdake> wirehead_ anything that I can do to help communicate that idea?
16:31:06 <sdake> nice :)
16:31:17 <rhallisey> wirehead_, nice
16:31:28 <wirehead_> sdake: Might be worth it to just post an official summary of what needs to change with that patch for it to get +2'd
16:31:55 <sdake> wirehead_ that would be atypical, but I think an email explaining how we are relaxing reviews aroudn kolla-kubernetes
16:32:10 <sdake> because of this change to documenting globals.yml
16:32:29 <sdake> i'd find that helpful for the core team
16:32:57 <sdake> your complaint was that everyone is looking at the end solution
16:33:09 <sdake> which is valid :)
16:33:52 <sdake> ok well then anything else on kolla-kubernetes?
16:33:59 <rhallisey> nothing atm
16:34:05 <rhallisey> thanks
16:34:28 <sdake> rhallisey would you mind sending out a email explaining the review approach because of globals.yml  cut and paste?
16:34:43 <rhallisey> ok
16:35:19 <sdake> might help to point people at the irc log of the meeting index on the technical details
16:35:34 <sdake> #topic open discussion
16:35:58 <inc0> so, we don't have any input to mascot creation?
16:36:01 <inc0> really?
16:36:17 <sdake> inc0 I am not really sure
16:36:23 <coolsvap> pip installs in binary containers
16:36:35 <sdake> inc0 up first then coolsvap has floor
16:37:03 <inc0> sdake, can you ask?
16:37:05 <sdake> inc0 I can find out if you like what hte process will be
16:37:07 <inc0> we have grand plans;)
16:37:49 <sdake> inc0 it may be hiding on the mailing lsit somewhre
16:37:57 <sdake> inc0 give me a couple hours
16:38:06 <sdake> coolsvap ure up
16:38:12 <coolsvap> #link http://lists.openstack.org/pipermail/openstack-dev/2016-July/100024.html
16:38:42 <coolsvap> i wanted to bring it in meeting
16:39:35 <coolsvap> inputs are most appreciated
16:39:36 <sdake> coolsvap mind bringing people up to speed on the status of the thread
16:40:05 <coolsvap> so there is a request to do pip install in networking-sfc container review
16:40:30 <coolsvap> which I think we have not allowed till now
16:40:57 <sdake> i am pretty sure we have allowed source only containers
16:41:14 <sdake> murano as an example
16:41:31 <sdake> how is this one different from that?
16:41:48 <coolsvap> i am fine with source only containers
16:41:53 <coolsvap> please have a look at https://review.openstack.org/#/c/344930/
16:42:17 <coolsvap> https://review.openstack.org/#/c/344930/6/docker/neutron/neutron-networking-sfc-agent/Dockerfile.j2
16:42:20 <coolsvap> oh its updated
16:42:35 <coolsvap> https://review.openstack.org/#/c/344930/1/docker/neutron/neutron-networking-sfc-agent/Dockerfile.j2
16:42:37 <coolsvap> patch set 1
16:43:30 <coolsvap> do we allow pip install networking-sfc in binary containers
16:43:38 <sdake> coolsvap its hard to review in meeting time - but I htink what woudl help is what are peoples positions
16:43:43 <sdake> where is the problem taht needs fixing
16:44:41 <coolsvap> thats my question
16:44:56 <sdake> i see :)
16:45:01 <coolsvap> i am fine if its discussed outside meeting or mailing list
16:45:28 <sdake> coolsvap i'll ask the core team to have a look at that thread
16:45:55 <sdake> so we can make a determination
16:46:08 <sdake> my general position is pip and yum should not be mixed unless absolutely necessary
16:46:35 <sdake> anyone else have anythign to add?
16:47:25 <sdake> any other open items?
16:47:31 <coolsvap> I think we have had situations where packages were not present in distos where we had source only containers
16:47:46 <sdake> coolsvap yup that is correct
16:48:35 <sdake> murano was not in rdo at one point (maybe still) and the binary container didn't build but source did from upstream
16:49:12 <sdake> ok well thanks for coming  folks
16:49:25 <sdake> #endmeeting