16:01:17 <inc0> #startmeeting kolla
16:01:18 <openstack> Meeting started Wed Sep 14 16:01:17 2016 UTC and is due to finish in 60 minutes.  The chair is inc0. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:21 <openstack> The meeting name has been set to 'kolla'
16:01:26 <inc0> #topic rollcall
16:01:29 <pbourke> o/
16:01:30 <inc0> woot for kolla please
16:01:34 <vhosakot> o/
16:01:38 <duonghq> o/
16:01:38 <vhosakot> wooot!!
16:01:39 <dave-mccowan> o/
16:01:44 <egonzalez90> woot o/
16:01:50 <hrito> o/
16:02:10 <wirehead_> o/
16:02:13 <mliima__> \o/
16:02:21 <sdake> hey folks
16:02:22 <sdake> inc0 can you makeme chair please
16:02:30 <inc0> #chair sdake
16:02:31 <openstack> Current chairs: inc0 sdake
16:02:33 <sdake> thanks
16:02:38 <inc0> wanna take over?
16:02:41 <sdake> inc0 - i'll handlethis meeting please
16:02:45 <inc0> sure
16:02:51 <sdake> sorry for tardinss -was otp
16:03:01 <inc0> no sweat, it's all yours
16:03:01 <sdake> hey folks ;)
16:03:08 <sdake> #topic announcements
16:03:26 <coolsvap> o/
16:03:28 <mliima__> hey sdake :)
16:03:51 <sdake> As you have seen I am passing on responsbilities for kolla one of our two new fearless leaders
16:04:06 <sdake> I'd ask you to give them your support when one is selected
16:04:13 <sdake> both are fantastic candidates
16:04:29 <sdake> a bit about how the transition willbehanlded
16:04:55 <sdake> ptl transitions can sometimes be shakey because otgoing ptls and incoming ptls don't get along or have different agendas
16:05:04 <sdake> that is not the case for our community
16:05:30 <sdake> which is great - less problems to worry about ;)
16:05:52 <sdake> after the ptl is elected, I will act as release steward for newton until 3.0.0
16:06:02 <rhallisey> yo
16:06:21 <sdake> the new ptl will take on responsibility for running meetings and whatnot
16:06:30 <sdake> and communication with the fuoundation
16:06:51 <sdake> once elected
16:07:00 <zhubingbing> hi
16:07:11 <sdake> the implication of all this is yo uwill hve two leaders asking for posibly two different things
16:07:40 <sdake> i'd like to ask folks to remember 3.0.0 (newton) has 269 bugs atm and we have a whole lot of work to do to execute the release
16:07:54 <sdake> its our community rsponsibility to release 3.0.0 on time
16:08:13 <sdake> it is my responsibility as release steward to drive that forward
16:08:28 <sdake> change in eladership is a foudning princple ofopenstack
16:08:31 <sdake> and a fantastic thing
16:09:04 <sdake> i'm glad i feel kolla is in great shape enough that i can pass responsibilities to a new leader without them having to struggle
16:09:12 <sdake> any quetions on that?
16:09:29 <egonzalez90> sdake, are you leaving kolla or just leaving PTL?
16:09:36 <sdake> i am remaining core
16:09:42 <sdake> why on earthw ould i leave kolla
16:10:21 <sdake> these guys running for ptl have earned the right for the responsibility
16:10:34 <sdake> I could continue on as ptl forever likely - but that isn't how i roll
16:11:00 <sdake> any other questions?
16:12:03 <sdake> ok - thats my monologue any other communitiy updates?
16:12:12 <inc0> rc1
16:12:16 <inc0> is lurking at us
16:12:43 <inc0> so rc1 happends tomorrow
16:13:03 <inc0> which means no more feature merges and we branch
16:13:12 <inc0> it will create stable/newton branch
16:13:13 <sdake> oh thanks for mentioning that
16:13:33 <inc0> well, not we, but it will be there
16:14:00 <sdake> because of the brarnch, every bug will need to be backported targeted for 3.0.0 (from master)
16:14:26 <sdake> and require a bug id unforunately
16:15:25 <sdake> the agenda hasn't been updated, so lets roll on the fly :)
16:15:29 <inc0> we'll need to bind source builds to stable versions instead of master
16:15:43 <inc0> sdake, I'd like to talk about kolla-ansible split
16:16:01 <sdake> ok - added to my brains' agenda :)
16:16:09 <sdake> #topic rc1
16:16:42 <sdake> #ink https://launchpad.net/kolla/+milestone/newton-rc1
16:16:51 <sdake> pleasee opeen that up
16:17:04 <sdake> we have made tremendous progress in the last 12 days
16:17:13 <sdake> seems like our businest time
16:17:22 <sdake> everyone is doing a fantasitc job reviewing and crnaking out code
16:17:37 <sdake> really nice work:)
16:17:52 <sdake> i'd like to guide folks to review the blueprints that are outstanding
16:18:01 <sdake> ahead of fall else
16:18:09 <sdake> and drive to getting them merged
16:18:13 <sdake> senlin is done ffrom what i can tell
16:18:21 <sdake> nto sure if it merged over ngiht ornot
16:18:34 <egonzalez90> yup, just marked now as implemented
16:18:42 <egonzalez90> merged an hour ago
16:18:58 <sdake> nice job egonzalez90 and crs for beating the review into shape ;)
16:19:21 <sdake> any othe rblueprints ready to merge?
16:19:36 <pbourke> i see two that need review
16:19:51 <pbourke> https://review.openstack.org/#/c/358305/
16:19:52 <sdake> networking-sfc is problematic at this point - but I asked jeffrey4l to take over the review, but Im not sure if he has it passing the gate yet
16:20:13 <pbourke> https://review.openstack.org/#/c/346842/
16:20:23 <pbourke> sfc is ready to rock
16:20:27 <sdake> vmtp seems as good as its going to get
16:20:32 <sdake> pbourke nice - gate green then
16:21:05 <sdake> and the answer is yes (second review)
16:22:30 <sdake> lets walk down lsist real quick
16:22:45 <egonzalez90> https://blueprints.launchpad.net/kolla/+spec/performance-monitoring
16:22:45 <zhubingbing> i agree
16:23:19 <inc0> we're super close
16:23:21 <sdake> i think performance monitoring is not going to make it
16:23:28 <sdake> but its so close
16:23:38 <zhubingbing> why /?
16:23:41 <inc0> it still can
16:23:42 <sdake> zhubingbing has done a great job getting that rolling
16:23:53 <inc0> just one +2 missing on both reviews
16:24:06 <sdake> inc0 cool - just woke up so havnet processed launchpad yet
16:24:15 <sdake> ok lets go ahead and review those
16:24:22 <inc0> let's make it a focus guys, we're super close
16:24:30 <inc0> and people wanted this feature badly
16:24:34 <sdake> ya - focus on the priorities et in reviews
16:24:40 <sdake> rather in launchpd
16:25:02 <sdake> inc0 is anyting needed here: https://blueprints.launchpad.net/kolla/+spec/upgrade-ceph-to-jewel
16:25:11 <inc0> I already closed it
16:25:14 <inc0> it's doen
16:25:21 <sdake> nice
16:25:22 <sdake> let me refresh
16:25:44 <sdake> #ink https://blueprints.launchpad.net/kolla/+spec/kuryr-docker-plugin
16:25:59 <sdake> the kuryr keystone v3 support merged yesterday
16:26:05 <sdake> this was blocking kuryr from merging in kolla
16:26:21 <sdake> as a result of that keystone v3 support, thee review needs some rework
16:26:31 <sdake> huikiang said it was cutting it close
16:26:36 <sdake> so that may not make newton
16:26:55 <sdake> #link https://blueprints.launchpad.net/kolla/+spec/neutron-vpnaas-agent
16:27:21 <sdake> anyone know status of this
16:27:38 <zhubingbing> me
16:27:44 <zhubingbing> it is finshd
16:27:52 <pbourke> https://review.openstack.org/#/c/335502/
16:27:54 <sdake> merged o rready for review?
16:28:02 <zhubingbing> This I have been tested
16:28:07 <pbourke> needs +a
16:28:08 <zhubingbing> yes
16:28:21 <sdake> pbourke's +2 is good enough fforme
16:28:37 <sdake> lets still do a thourough job of reviewing these
16:29:29 <zhubingbing> This I have been in the local test by
16:29:40 <zhubingbing> i think it can merged
16:30:00 <sdake> blue, green, and purple are all positive colors
16:30:13 <sdake> rc1 looks like its in good shape
16:30:39 <sdake> looks like we may only be missing kuryr out of our original plan
16:30:50 <zhubingbing> yes
16:31:05 <zhubingbing> miss kuryr
16:31:21 <sdake> i haven't given up on kuryr making the releae
16:31:22 <sdake> it has 24 hours
16:31:25 <sdake> on to bugs
16:31:34 <sdake> if fy o uhae a bug in progress and can get it merged
16:31:35 <sdake> please do so
16:31:50 <zhubingbing> i can do it
16:31:55 <zhubingbing> link give me
16:31:55 <sdake> crs please prioritize bugs in "in progress state"
16:32:04 <sdake> do which  zhubingbing
16:32:10 <zhubingbing> kuryr
16:32:22 <sdake> zhubingbing sync up wth huikang, that is his baby
16:32:47 <zhubingbing> np
16:32:56 <sdake> anything triaged or confirmed bugwise will go into rc2
16:33:05 <sdake> we hve a cuple weeks to sort out rc2
16:33:12 <sdake> NO FEATURES IN RC2
16:33:21 <sdake> now if a feature we implemented is broken, thats a bug
16:33:32 <sdake> so those kinds of fixes will be permitted
16:33:32 <zhubingbing> https://review.openstack.org/#/c/298894/ i fixed it now
16:34:04 <zhubingbing> i am going to contact huikang
16:34:11 <sdake> any quetions on our rc1 planning?
16:34:47 <zhubingbing> https://blueprints.launchpad.net/kolla/+spec/performance-monitoring
16:34:53 <zhubingbing> why close it
16:35:05 <sdake> close?
16:35:09 <zhubingbing> progress status be ok
16:35:21 <sdake> its in needs revie
16:35:29 <sdake> telegrpah and grafana haven't merged yet
16:35:42 <zhubingbing> ok
16:35:47 <zhubingbing> i understand
16:35:48 <inc0> zhubingbing, did you add release note for it?
16:35:50 <inc0> in reno?
16:35:51 <sdake> (lr atleast not as of midnight last night my time)
16:36:26 <zhubingbing> which
16:36:37 <sdake> zhubingbing all of em
16:36:51 <sdake> every bluerpint implementtoin should have a release note
16:36:51 <sdake> and some bugs too
16:36:52 <inc0> well I'd really keep it "we added monitoring stack"
16:36:59 <zhubingbing> ok
16:37:21 <sdake> zhubingbing we can hash that out inreviews if your available today
16:37:42 <sdake> but ya general monitoring stack idea inc0 proposed is good- along with individual components that were added
16:37:45 <zhubingbing> I don't think there's any problem.
16:37:52 <sdake> cool
16:38:00 <sdake> if there are bugs, we can sort em out in rc2
16:38:03 <inc0> we don't have release notes yet, but that's minor, we'll add it
16:38:26 <sdake> ok 8 minutes timebox for k8s update
16:38:32 <sdake> #topic kubernetes update
16:38:47 <rhallisey> so
16:38:51 <sdake> rhallisey al yours :)
16:39:02 <rhallisey> quickstart guide has been getting ironed out
16:39:08 <inc0> rhallisey, multinode?
16:39:12 <rhallisey> the services are falling into place nicely
16:39:24 <rhallisey> inc0, it will work for both
16:39:28 <rhallisey> aio & multinode
16:39:29 <inc0> rhallisey, cool
16:39:54 <rhallisey> the next steps come as how the cli commands will be run
16:40:18 <rhallisey> the community talked about using ansible/workflow tools
16:40:45 <rhallisey> I prototyped this in the past using the ansible-kube module
16:41:04 <rhallisey> that patch still exists, but it will call the granular cli now
16:41:20 <rhallisey> ok that's all the updates
16:41:26 <rhallisey> anyone else?
16:41:37 <sdake> i have something to add
16:41:51 <sdake> kolla-kubernetes has fantastic potential
16:42:00 <sdake> in order to execute, we need core reviewers to review teh work
16:42:06 <rhallisey> indeed
16:42:08 <sdake> the review queue is jammed up
16:42:14 <inc0> I think we're more constrained on people writing code now;)
16:42:21 <sdake> i kno wwe hae alot of jbectives
16:42:31 <sdake> inc0 yup - thats comming
16:42:43 <sdake> but please add kolla-kuberenetes to your watch list
16:42:45 <sdake> and review it
16:42:52 <sdake> review aptchess related to it
16:42:58 <rhallisey> oh one more thing
16:43:08 <rhallisey> when will the repo split be re visisted?
16:43:17 <sdake> thats next on agenda
16:43:19 <inc0> in about a minute I hope?
16:43:22 <rhallisey> ok
16:43:32 <duonghq> I think the k8s's doc should be kept in sync with kolla one
16:43:38 <sdake> any questions for kolla-kubernetes?
16:43:54 <rhallisey> duonghq, in terms of the quickstart?
16:44:01 <duonghq> yup
16:44:13 <rhallisey> well for deps and some of the pieces yes
16:44:28 <rhallisey> I think it needs to re use a bunch of it
16:44:46 <duonghq> just a note, I want to discuss further on kolla after this meeting
16:44:52 <duonghq> something about PBR issue
16:44:55 <duonghq> so, go on
16:45:00 <rhallisey> cool thanks duonghq
16:45:09 <sdake> ok timebox is up
16:45:21 <sdake> 12 minutes or kolla-ansible repo spit
16:45:24 <sdake> #topic kolla-ansible repo split
16:45:33 <sdake> inc0 all yours
16:45:46 <inc0> okj
16:45:58 <inc0> so in midcycle we discussed splitting, this needs to happen
16:46:09 <sdake> agreed
16:46:09 <inc0> I kept hearing about it from number of different sources
16:46:24 <inc0> question was timing and we all agree that newton is not a good time
16:46:28 <inc0> but newton is coming to an end
16:46:54 <inc0> so when we branch out stable/newton , repo split becomes an option
16:47:12 <sdake> the downside of that is our 260+ bug backports become more difficult
16:47:14 <inc0> easiest way to do it would be to pretty much split dir ansible
16:47:18 <inc0> and tool kolla-ansible
16:47:30 <inc0> not that much tho
16:47:49 <inc0> since we'll keep similar repo structure in reality
16:47:58 <inc0> just one fix needs to go to two repos instead of one
16:48:09 <sdake> yup- and i think our test of seprating containers from ansible has worked out well
16:48:09 <inc0> I think we could prep small shell script to help with that
16:48:14 <inc0> should be entirely doable
16:48:34 <sdake> i'd prfer keeping fixes separate as we have one in the past
16:48:38 <apuimedo> sdake: thanks for not giving up on us
16:48:42 <inc0> in Ocata timeframe we could look at moving configs to shared kolla as well
16:48:58 <apuimedo> we have a patch for a Dockerfile, maybe hui kang can reuse bits of it
16:49:04 <sdake> apuimedo i dont give up on people - but sometimes I give up on companies :)
16:49:07 <apuimedo> sent it yesterday and it already has +2
16:49:18 <apuimedo> sdake: that's a good policy
16:49:44 <inc0> soo, my question to our glorious community is
16:49:51 <sdake> i think the timing is right for a vote on the repo spit
16:49:54 <inc0> does anyone specifically don't want it to happen
16:50:09 <inc0> we will repeat this question on ML
16:50:10 <sdake> inc0 my onhy concern is backports
16:50:22 <inc0> sdake, I'll take a look at automating it
16:50:24 <sdake> other then that, wfm
16:50:31 <inc0> should be matter of more clever shell script
16:50:42 <inc0> with a dash of git magic
16:50:54 <inc0> or even not
16:51:13 <inc0> because we'll backport 2 repos into one
16:51:25 <inc0> so we can just apply exactly same diffs
16:51:37 <sdake> i'd prefer backport process o be gerrit cherrypick
16:51:43 <sdake> rather then working with a git repo directly
16:51:58 <inc0> not sure if we can do that really
16:52:08 <inc0> but creating a patch on mitaka will be simple
16:52:16 <inc0> or newton
16:52:20 <rhallisey> this may be a little annoying, but it's a must for the community imo
16:52:35 <inc0> I agree,
16:52:41 <sdake> i totally agree rhallisey
16:52:51 <sdake> its not a matter of annoyance to me
16:52:57 <sdake> its a matter of delivering 3.0.0 on time
16:53:23 <sdake> lets have a vote on ml and answer two questions 1) what the repo split is 2) when it should happen
16:53:26 <inc0> sdake, I don't think this is endangered in any way
16:53:29 <sdake> inc0 i can do that or you can
16:53:52 <inc0> sdake, since you have most questions, please write an email and I'll clarify
16:54:02 <inc0> cuz if you have questions others will too;)
16:54:21 <sdake> my question is around timing
16:54:28 <sdake> others questions may be about wht repo split is
16:54:39 <inc0> once ocata starts, we can do it, earlier the better
16:54:54 <sdake> if you prefer I draft the email, i'll present a couple options for when
16:55:12 <sdake> inc0 if it makes you feel better as soon as 3.0.0 is in the can, I'd do t he repo work myself personally
16:55:33 <inc0> sdake, let's both work on it
16:55:41 <inc0> but agree, 3.0.0 is priority
16:55:51 <sdake> what i mean is, I am supportive of the idea
16:55:56 <sdake> i dont care who does the work :)
16:56:04 <sdake> but i think its important
16:56:08 <sdake> and would do it myself if need be
16:56:33 <inc0> so I don't think we have much opposition to it, but let's start ML thread
16:56:45 <sdake> by lets you mean me :)
16:56:52 <inc0> you're PTL man
16:56:56 <inc0> :P
16:56:56 <sdake> ok will do shortly after this meeting
16:57:00 <inc0> thanks
16:57:01 <sdake> inc0 anyone can propose votes
16:57:07 <sdake> coolsvap is the only one that has
16:57:14 <rhallisey> shoudl I re raise :D
16:57:15 <sdake> no rhallisey did as well forgive me
16:57:18 <sdake> memory foggy
16:57:33 <inc0> and rhallisey did it on this particular subject too!;)
16:57:35 <sdake> ok i'll take care of it this mroning
16:57:37 <sdake> you havemy word
16:57:39 <inc0> thanks
16:57:45 <sdake> #topic open discussion
16:57:51 <sdake> 3 minutes
16:57:52 <sdake> dense meeting
16:57:53 <inc0> make sure to include all questions you'd like to clarify
16:58:02 <sdake> apologies for not alot o extra time for open discussion
16:58:15 <sdake> #topic thanks for making kolla happen
16:58:22 <inc0> sooo...one thing from me
16:58:25 <inc0> once RC releases
16:58:33 <sdake> see topic
16:58:35 <inc0> we need to test the crap out of M->N upgrades
16:58:52 <sdake> inc0 agree - testing needs to be a priority for al new services and n->m ugprades and reconfigure
16:59:01 <zhubingbing> i agree
16:59:05 <inc0> all services, priority is old actually
16:59:10 <inc0> as we need to keep them up and running
16:59:10 <sdake> anyway
16:59:18 <inc0> especially that we changed few moving parts
16:59:21 <sdake> This team is fantastic
16:59:22 <inc0> underneath
16:59:33 <inc0> sdake, woot for kolla?;)
16:59:33 <duonghq> thank you everybody
16:59:34 <sdake> one of the finest I've had the pleasure of serving as a leader for
16:59:51 <sdake> Your all rockstar engineers with a bright future ahead
17:00:15 <sdake> inc0 - out o  time :)
17:00:22 <sdake> we can try the woot next time :)
17:00:26 <pbourke> inc0: not sure woot for kolla is catching on :p
17:00:34 <inc0> pbourke, just wait.
17:00:35 <sdake> #endmeeting