16:00:28 <inc0> #startmeeting kolla
16:00:29 <openstack> Meeting started Wed Feb  8 16:00:28 2017 UTC and is due to finish in 60 minutes.  The chair is inc0. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:33 <zhubingbing> o/
16:00:33 <openstack> The meeting name has been set to 'kolla'
16:00:46 <inc0> #topic rollcall, w00t!
16:00:50 <sdake> zhubingbing you jumped the gun :)
16:00:50 <inc0> you know what to do
16:00:53 <sdake> w00t
16:00:55 <berendt> o/ woot
16:01:00 <egonzalez> woot o/
16:01:03 <zhubingbing> woot
16:01:05 <krtaylor> o/
16:01:10 <zhubingbing> w00t!!
16:01:12 <zhubingbing> -)
16:01:22 <akwasnie> Hi:)
16:01:30 <inc0> welcome back akwasnie!
16:01:37 <akwasnie> Finally!
16:01:37 <sdake> long time no see akwasnie - welcome back :)
16:01:38 <Jeffrey4l> o/
16:01:46 <zhubingbing> welcome back!!
16:01:49 <akwasnie> Thanks! I missed you guys
16:01:58 <akwasnie> I am in Usa now:)
16:02:01 <sp_> 0/
16:02:03 <vhosakot> 0/
16:02:16 <egonzalez> welcome o/
16:02:37 <mandre> o/
16:02:43 <inc0> right... let's move on, agenda seems busy
16:02:49 <inc0> #topic announcements
16:03:05 <sdake> i've got one
16:03:14 <inc0> soo, we are in the middle of pre-release
16:03:27 <inc0> so I'd encourage everyone to focus on bugs in kolla and kolla-ansible
16:03:36 <sdake> i am doing a seminar at 2:00 PM MST for 2 hours on all 3 kolla deliverables - see ml for more details :)
16:03:40 <inc0> so we can deliver another kickass release early March
16:03:48 <sdake> on feb 15th that is :)
16:03:58 <Jeffrey4l> we may have to create rc1 tag and ocata branch this week.
16:04:01 <sdake> and inc0 ^^^ :)
16:04:04 <caowei> o/
16:04:07 <klindgren> o/
16:04:08 <duonghq> o/ sorry for late, my fiber connection has gone by someway
16:04:20 <gema> o/
16:04:28 <inc0> ok, let's move on to meeting itself
16:04:29 <jascott1> o/
16:04:32 <krtaylor> quick introduction, I'm getting more involved with the intention of making things work on ppc64le architecture
16:04:39 <inc0> #topic dumb-init  vs tini
16:04:39 <Jeffrey4l> then fix bug require a real bug number, and fix master first then back port into ocata branch.
16:04:40 <srwilkers> o/
16:04:45 <inc0> Jeffrey4l, you're up
16:04:48 <pbourke> o/
16:04:49 <Jeffrey4l> yes.
16:04:51 <Jeffrey4l> thanks.
16:05:05 <Jeffrey4l> it is related to this patch https://review.openstack.org/430108
16:05:12 <zhubingbing> agree jeffrey41
16:05:22 <britthouser> 0/
16:05:22 <Jeffrey4l> some background:  docker 1.13 provide init process now.
16:05:31 <Jeffrey4l> which is using tini
16:05:40 <Jeffrey4l> #link https://github.com/krallin/tini
16:06:16 <Jeffrey4l> so i think we should sync with docker use the same init solution?
16:06:19 <sdake> my two concerns when Jeffrey4l raised this last night were the license may have been funky - that is not the case - and why fix what isn't broken - which is  open for debate
16:06:21 <inc0> Jeffrey4l, I'd say not in Ocata
16:06:47 <berendt> dumb-init has still active upstream?
16:06:53 <inc0> and in Pike we can explore even pinning to docker 1.13
16:06:58 <Jeffrey4l> another reason is: dumb-init 1.2.0 works wired.
16:07:05 <sdake> inc0 or introducing entirely new runtimes even ;-)
16:07:12 <Jeffrey4l> that why we revert the dumb-init 1.2.0 patch.
16:07:33 <Jeffrey4l> inc0, this is not related the docker 1.13. this patch do not pin kolla to docker 1.13.
16:07:37 <inc0> we don't know if tini works any better
16:07:43 <Jeffrey4l> check the change, the gate works.
16:07:53 <inc0> I know, but you're saying docker 1.13 will use it out of the box right
16:07:54 <inc0> ?
16:07:56 <berendt> i would prefer to not change as long as dumb-init is working for us
16:08:07 <pbourke> dumb-init is not working for us
16:08:12 <pbourke> its busted in latest relase
16:08:17 <pbourke> tini is the upstream docker solution
16:08:25 <berendt> but not yet, right?
16:08:40 <Jeffrey4l> we hit this issue https://github.com/Yelp/dumb-init/issues/136
16:08:45 <inc0> so I'm really reluctant to make this kind of change after ocata-3:/
16:09:12 <Jeffrey4l> inc0, if we wanna to pin docker 1.13 use the its init solution.
16:09:12 <pbourke> we could stick with dumb-init 1.2 till pike
16:09:19 <berendt> is it not possible to fix dumb-init?
16:09:29 <inc0> Jeffrey4l, that's my point
16:09:36 <Jeffrey4l> we have to run container like "docker run --init xx" a new parameter.
16:09:40 <sdake> we are not pinning to 1.13 in ocata release :(
16:09:43 <inc0> we can drop our own init alltogether
16:09:47 <inc0> not in ocata sda
16:09:50 <inc0> sdake,
16:09:53 <sdake> right
16:09:56 <sdake> ok all good then :)
16:09:57 <Jeffrey4l> ok. i am fine with that ;)
16:10:05 <inc0> what we have now, works, issue is with dump-init 1.2
16:10:09 <inc0> which we don't have to use
16:10:33 <inc0> but that puts dump-init on notice (I recently learned new phrase in English!)
16:11:06 <inc0> and in Pike we can explore either switching to tini or looksing init completely and pin to docker 1.13
16:11:16 <inc0> but that will require some discussion in ops community
16:11:21 <Jeffrey4l> ok. i will give a -2 to my patch until Pike.
16:11:48 <inc0> right, anything else in that topic?
16:11:58 <Jeffrey4l> i am done .
16:12:12 <berendt> should we open a blue print for it?
16:12:22 <inc0> berendt, +1 to that
16:12:24 <klindgren> would tini allow someone to provide a default overide-able start command for each container?
16:12:53 <Jeffrey4l> klindgren, tini not. but docker 1.13 yes.
16:13:08 <berendt> Jeffrey4l: do you open a blueprint for it?
16:13:11 <Jeffrey4l> there is a parameter "--init-path" in docker 1.13
16:13:13 <inc0> klindgren, we could do this with or without tini
16:13:13 <Jeffrey4l> berendt, np. i will
16:13:23 <inc0> with templates
16:13:49 <inc0> but that's different discussion:) feel free to ping me after meeting and we'll talk about it
16:13:52 <Jeffrey4l> another topic is pin docker 1.13. not it is not properly for current. ;)
16:14:00 <inc0> ok moving on
16:14:02 <inc0> #topic PTG Schedule finalization
16:14:31 <inc0> #link https://etherpad.openstack.org/p/kolla-pike-ptg-schedule
16:14:53 <inc0> still 2 kolla-k8s sessions not filled
16:15:38 <inc0> 1 session not filled:)
16:18:29 <sdake> would be nice to know who pink was to provide clrity on that potentially duplicate session
16:18:41 <sdake> i think operator feedback is not really a design session topci which #5 looks like
16:19:47 <vhosakot> team, sorry to say I'll miss the PTG as I'm attending my company's conference in Berlin the same exact days as PTG.  good thing is, I'll be presenting kolla to the customers and will try to get more contributors for us :)
16:20:34 <inc0> vhosakot, shame, but alcohol we would drink will only get older this way = better
16:20:41 <sbezverk> sdake: hmm I added it and it sounds clear to me
16:20:46 <vhosakot> inc0: haha :)
16:21:05 <sdake> sbezverk cool - coud yoou ahve a look and tell me if it is still clear?
16:21:07 <krtaylor> I guess it is safe to assume that each main number in the etherpad = one 40-50 min session?
16:21:14 <inc0> sbezverk, that's the issue, everything sounds clear to someone who made it;)
16:21:23 <inc0> krtaylor, correct
16:21:26 <sdake> krtaylor that is what we have done at every midcycle in tehe past ;)
16:21:27 <vhosakot> are the PTG seesions available on webex or recorded by any chance?
16:21:39 <sdake> vhosakot inc0 is sorting out remote participation
16:21:46 <inc0> you'll be able to call into it
16:22:00 <inc0> but  I don't think we'll record as it's legally iffy
16:22:07 <vhosakot> sdake: great, yes, I'd love to catch up on videos/recordings please.
16:22:25 <krtaylor> thanks, just wanted to be clear as I will be in a cab for the first couple :(
16:22:26 <vhosakot> inc0: cool, calling in is great
16:23:11 <inc0> krtaylor, we can bring you up to speed afterwards
16:23:18 <krtaylor> thx!
16:23:30 <sdake> question what is day 1 #2 about?
16:23:59 <inc0> sdake, we have some strange wiring between projects now
16:24:18 <sdake> so this is cross-repo gating or somethign ele?
16:24:25 <inc0> kolla using ansible gates, kolla-ansible configs in kolla-k8s
16:24:27 <inc0> all that stuff
16:24:27 <sdake> would like to have some context to prep properly :)
16:24:37 <sdake> i see - could you add more detials on that one ?
16:24:42 <sdake> i think the rest ono day 1 are pretty clear
16:24:44 <inc0> sure
16:24:50 <sdake> now for day 2:)
16:24:58 <egonzalez> yep, looks very confussion by the name
16:27:20 <sdake> day 2 has some topics
16:27:32 <sdake> rather adjustments
16:29:46 <gema> multiarch? where would it go?
16:29:49 <gema> waiting list?
16:30:06 <gema> coffee break?
16:30:39 <krtaylor> lol
16:31:21 <inc0> right, I think schedule looks good
16:31:34 <inc0> we'll take one more look at it next week
16:31:41 <inc0> and I'll put hours there:)
16:32:27 <inc0> #topic release notes (timebox 15min)
16:32:47 <sdake> please for the love of god do release notes for kolla-kubernetes
16:32:57 <inc0> I'd like everyone to try to recall changes that has been made in this release and confirm that reno notes has been made
16:32:59 <sdake> a lot of peopel read them and track our progress
16:33:21 <inc0> now I'm talking about kolla-ansible/kolla;)
16:33:28 <sdake> inc0 ahh sorry :)
16:33:28 <inc0> as we actually release these
16:33:37 <Jeffrey4l> prepare an etherpad link?
16:33:40 <sdake> inc0 right - my statement still stands and I hoep you supoprt it:)
16:33:50 <inc0> Jeffrey4l, good point, let me do this
16:34:14 <zhubingbing> +1
16:34:27 <inc0> #link https://etherpad.openstack.org/p/kolla-ocata-release-notes
16:34:49 <inc0> sdake, there are many ways to track progress, release notes aren't normally meant for this;)
16:35:18 <inc0> but back on track, let's note what is not in reno and collectively fix these
16:35:31 <inc0> I'm especially interested in upgrade notes, this has to be up to date
16:41:24 <sdake_> can someone link the release notes etherpad, my network dc'ed
16:42:02 <berendt> sdake_: https://etherpad.openstack.org/p/kolla-ocata-release-notes
16:42:11 <sdake_> tx
16:49:30 <inc0> ok we're running out of out timebox:)
16:49:51 <inc0> few more minutes and let's leave some time for open discussion for a change
16:51:46 <inc0> allright, so everyone, don't stop after meeting, take a longer look at release notes
16:51:53 <inc0> these are really really important
16:52:05 <inc0> especially if some step needs to be made between upgrades
16:52:30 <inc0> also, we need to fill them so please add missing notes to repo
16:52:36 <inc0> #topic Open discussion
16:52:55 <inc0> 8 minutes left, anyone has anything they'd like to share
16:52:55 <inc0> ?
16:53:24 <gema> I'd like to ask for reviews of this patch: https://review.openstack.org/#/c/430940/
16:54:18 <jascott1> can day 1 #7 be about helm features we might need also?
16:54:49 <sdake_> jascott1 that would be a better use of time I think
16:55:10 <Jeffrey4l> gema, how can i get ppc docker base image?
16:55:13 <jascott1> we might be able to cover it all in the 'helming hour'
16:55:16 <Jeffrey4l> does docker provide this?
16:55:29 <gema> Jeffrey4l: krtaylor may be able to answer that
16:56:23 <krtaylor> not off the top of my head, let's discuss after meeting and I'll find out
16:56:37 <gema> Jeffrey4l: may be worth adding the question to the patch
16:56:47 <Jeffrey4l> yep.
16:56:49 <gema> and bring the conv there
16:57:25 <krtaylor> btw, also reviewing that now
16:57:50 <gema> thanks!
16:58:22 <Jeffrey4l> OT: dumb-init do not provide ppc edition
16:59:03 <inc0> meh
16:59:15 <inc0> well we're removing it anyway;)
16:59:18 <inc0> ok guys
16:59:22 <inc0> thank you for coming!
16:59:29 <inc0> #endmeeting kolla