14:02:38 <enikanorov_> #startmeeting neutron lbaas
14:02:39 <openstack> Meeting started Thu Jun 26 14:02:38 2014 UTC and is due to finish in 60 minutes.  The chair is enikanorov_. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:02:40 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:02:43 <openstack> The meeting name has been set to 'neutron_lbaas'
14:02:52 <sballe> Sorry for not having been active lately. In face to face meeting which limit my online interaction
14:02:54 <blogan> thanks enikanorov
14:03:22 <sbalukoff> sballe: It happens to the best of us. ;)
14:03:34 <enikanorov_> i think we need to start with choosing a volunteer for the lbaas meetings chair :)
14:03:37 <sballe> sbalukoff, :-)
14:03:48 <jorgem> I can spearhead that
14:04:09 <enikanorov_> jorgem: that would be great
14:04:18 <german__> jorgem +1
14:04:21 <sballe> lbaas meeting chair: does that mean runnign the lbass meeting?
14:04:26 <jorgem> Unless someone else REALLY wants to lol
14:04:28 <enikanorov_> sballe: yep
14:04:31 <sbalukoff> +1 to jorgem running these. :)
14:04:38 <enikanorov_> running the meeting, updating meeting wiki page
14:04:38 <sballe> +1 jorgem
14:04:49 <blogan> +1 jorgem
14:05:05 <jorgem> Sounds good. I had some thoughts though that I'd be interested to hear people opinions on
14:05:16 <ctracey> ehlo
14:05:43 <s3wong> hello
14:05:50 <jorgem> Does eveyone like the way the weekly standup etherpad doc is working for them thus far?
14:05:57 <enikanorov_> ok, jorgem, welcome to the Iron Throne!
14:05:58 <sbalukoff> jorgem: yep!
14:06:03 <jorgem> woot!
14:06:05 <german__> yep
14:06:19 <sballe> ye
14:06:25 <blogan> enikanorov: are you saying he's joffrey?
14:06:43 <enikanorov_> blogan: lol no
14:06:50 <blogan> well i would have agreed lol
14:06:55 <s3wong> blogan: ready for the purple wedding?
14:07:06 <rm_work> blogan: no, spoilers, he's tommen
14:07:11 <blogan> s3wong: ahh yeah
14:07:45 <jorgem> So does everyone have items they want to address today?
14:07:56 <jorgem> There was no agenda so let's try and set one up.
14:08:06 <enikanorov_> jorgem: how about updates from those who are assigned?
14:08:16 <sbalukoff> Do we know if markmcclain is going to join?
14:08:24 <blogan> sbalukoff: he said hello
14:08:25 <sbalukoff> I'd love to hear an update on his flavor framework write-up idea.
14:08:27 <markmcclain> sbalukoff: I'm here:)
14:08:29 <rm_work> he said hi earlier sbalukoff
14:08:31 <rm_work> :P
14:08:34 <sbalukoff> Oh!
14:08:36 <sbalukoff> Yay!
14:08:39 <markmcclain> #link https://review.openstack.org/#/c/102723/2/specs/juno/neutron-flavor-framework.rst
14:08:40 <sbalukoff> Sorry, it's early.
14:08:46 <jorgem> Okay so updates are on the agenda. More than what is on the weekly standup I'm guessing
14:08:59 <sbalukoff> Sweet!
14:09:03 <german__> markmcclain woot!
14:09:04 <sbalukoff> That's quick. XD
14:09:12 <sballe> cool
14:09:23 <jorgem> Nice thx mark
14:09:41 <ctracey> One question that came in earlier in the week was around resource naming.  Sounded like the direction was to (re)name 'pools' to 'nodepools'. Can we get clarification on that?
14:09:56 <ctracey> blogan: ^^
14:09:57 <german__> nodepools +1
14:10:23 <blogan> ctracey: it wasn't renaming to nodepools it was just a stopgrap until that one issue is figured out
14:10:33 <ctracey> correct.
14:10:51 <blogan> ctracey: which I will need to talk to markmcclain or enikanorov about or both
14:10:56 <ctracey> my question is whether we think this is necessary and/or what the future impact will be
14:10:58 <german__> I still like nodepools more than pools :-)
14:11:19 <jorgem> Is this something that can be addressed here or the ML?
14:11:23 <sballe> I agree'
14:11:50 <ctracey> im fine with either venue
14:11:58 <evgenyf> I have update for TLS: new patch was commited, please reiew it. Also, implementation for "old" model was commited for first review https://review.openstack.org/#/c/102837/1. it will be rebased to new model when it lands
14:12:02 <ctracey> (we were looking for agenda items :) )
14:12:36 <blogan> ok lets keep it one topic at a time, right now we have flavor framework, nodepools, and TLS.  jorgem which should be discussed first?
14:12:39 <jorgem> So far I have the following agenda items:
14:12:40 <jorgem> 1) Updates
14:12:40 <jorgem> 2) Node pools vs pools naming.
14:12:40 <jorgem> 3) Mark McClain's flavor blueprint
14:12:40 <jorgem> 4) evgenvy has update for TLS blueprint
14:13:06 <blogan> lets start with the top
14:13:13 <jorgem> sounds good to me.
14:13:14 <german__> since we are alreday 12 mins in - let's close the agenda
14:13:21 <jorgem> +1
14:13:34 <avishayb__> and L7 switching doc as well https://review.openstack.org/#/c/99709/8
14:13:42 <jorgem> Okay, so any important updates that aren't in the weekly standup doc?
14:14:14 <jorgem> #link https://etherpad.openstack.org/p/neutron-lbaas-weekly-standup
14:14:24 <jorgem> In case anyone is wondering where the standup is
14:14:34 <blogan> we've got the extensin, plugin, and db piece of the refactor nearly complete
14:15:07 <blogan> next would unit tests and drivers, which includes the mammoth agent piece
14:15:55 <jorgem> Any other updates from other teams?
14:16:09 <german__> I updated the etherpad
14:16:13 <ctracey> I have much of the CLI work done and will be committing more of it today
14:16:28 <ctracey> I have also been hacking on devstack to get that up and working
14:16:49 <jorgem> As a side note, I was going to have the wiki be the central place for links and such to etherpads etc.
14:17:09 <jorgem> Because, I think we all can agree managing bookmarks suck :)
14:17:20 <sbalukoff> jorgem: +1
14:17:25 <markmcclain> good idea
14:17:26 <ctracey> yesterday we decided to co-locate our work under https://github.com/oslbaas....just while we are bootstrapping
14:17:26 <german__> + 1
14:17:52 <jorgem> ctracey: Could you link the full url please
14:18:09 <jorgem> #link https://github.com/oslbaas
14:18:11 <jorgem> nm I got it
14:18:33 <jorgem> Okay, let's move to item 2: node pools vs pools naming
14:18:51 <jorgem> Who wants to kick off that discussion?
14:18:59 <blogan> i wil
14:19:01 <blogan> l
14:20:01 <blogan> the reason I went with the different name is because since we are going ot have v1 of the lbaas api running alongside v2 at first, if two extensions use the same resource name then the neutron API will set it up so that the attributes of v1 pool and v2 pool will be combined
14:20:59 <blogan> so v1 pool has subnet_id and lb_method, v2 doesn't have subnet_id and has lb_algorithm, when a v2 pool is created the API will reject the call because it does not have subnet_id and lb_method
14:21:32 <ctracey> yes. I need to take a deeper dive into this code, but if the way I understand it is true, this seems like a limitation of the routing framework.
14:22:16 <blogan> if everyone just wants to change it to nodepools then fine, we should also change members to nodes.
14:22:24 <blogan> however I still prefer pool of nodepool
14:22:36 <ctracey> pool over nodepool?
14:22:50 <ctracey> i prefer pool
14:22:57 <blogan> yeah of = over
14:23:12 <blogan> any suggestions, ideas, opinions?
14:23:13 <jorgem> Which blueprint is this relevant to? I'd like to capture that.
14:23:19 <ctracey> the point I am trying to touch is that once we name it we should not rename
14:23:21 <blogan> obj model refactor
14:24:22 <ctracey> but i am surprised by this as it kind of makes our namespace change from /lb to /lbaas meaningless
14:24:24 <sbalukoff> ctracey: +1
14:24:30 <blogan> https://review.openstack.org/#/c/89903/
14:24:41 <jorgem> #link https://review.openstack.org/#/c/89903/
14:24:56 <blogan> ctracey: not totally meaningless, but definitely less meaningful
14:25:06 <ctracey> +1
14:25:36 <blogan> if anyone has any future suggestions, ideas, opinions take it to the ML
14:25:48 <ctracey> sounds good....I will formulate some today
14:25:53 <jorgem> Sounds good. I need to understand the problem in more depth myself.
14:25:54 <ctracey> after digging more
14:26:04 <blogan> ill shwo you today jorgem
14:26:13 <jorgem> Okay, let's move on to item 3 then.
14:26:33 <jorgem> Item 3 is markmcclain's flavor bp
14:26:53 <jorgem> Has everyone had a chance to look at it? I see some comments on it.
14:27:00 <enikanorov_> yes, I did
14:27:08 <german__> I just saw the link - so no
14:27:09 <sbalukoff> I have not yet had a chance to review it thoroughly.
14:27:20 <sballe> ditto
14:27:21 <enikanorov_> i've left comments on gerrit
14:27:32 <jorgem> For those that did, anything you want to weigh in on?
14:27:39 <german__> maybe we should table it until we all read it?
14:27:58 <rm_work> german__: +1
14:28:00 <markmcclain> enikanorov: adding responses to questions
14:28:02 <s3wong> well, FYI, there will be an IRC meeting tomorrow at 17:30 UTC on #openstack-meeting-3 in hope to decide on the flavor framework
14:28:03 <jorgem> german__:  That's fine with me. Just wanted to see if we should be focusing on something.
14:28:36 <german__> agreed
14:28:36 <sbalukoff> s3wong: Oh thank goodness it isn't any earlier than this meeting. XD
14:28:56 <sbalukoff> jorgem: I'm fine with tabling it until tomorrow's meeting.
14:29:15 <sballe> sbalukoff, +1
14:29:18 <jorgem> markmcclain: Anything we should pay special attention to when reviewing the doc?
14:30:16 <markmcclain> jorgem: not really.. it follow last week's discussion
14:30:25 <jorgem> Okay, well let's move on to item 4 then...
14:30:46 <jorgem> Item 4 is TLS blueprint update. evgenyf: You want to kick that off?
14:30:55 <evgenyf> Yes
14:31:13 <evgenyf> New patch was pushed today earlier
14:31:19 <evgenyf> Please review it
14:31:30 <vjay2> * Added marks blueprint to https://etherpad.openstack.org/p/juno_lbaas_mid_cycle_meetup_reviews
14:31:41 <ptoohill> please also review related BP https://review.openstack.org/#/c/100931/
14:31:51 <ptoohill> which is related to this topic
14:31:57 <evgenyf> I also commited implementation for first review
14:32:00 <jorgem> #link https://review.openstack.org/#/c/100931/
14:32:04 <evgenyf> it's related to current model
14:32:09 <sbalukoff> evgenyf: Looking at the commit comments, does your proposal bring TLS to the old model?
14:32:44 <evgenyf> sbalukoff:Yes, it will be rebased to new one when it land
14:33:02 <sbalukoff> So, I have a problem with bringing TLS to the old model: I don't think we should do it.
14:33:07 <sbalukoff> I think we should only have TLS on the new model.
14:33:11 <german__> sbalukoff +1
14:33:12 <ptoohill> i sorta agree here
14:33:17 <ptoohill> +1
14:33:22 <vjay2> +1
14:33:33 <ctracey> sbalukoff: +1
14:33:35 <sballe> +1
14:33:48 <evgenyf> sbalukoff: It will be for new model only, this review is just for getting first feedback on implementation
14:33:58 <sbalukoff> It seems like it will only be more work to port your proposal over, potentially requiring additional work for people writing shims.
14:34:05 <sbalukoff> (For no real benefit)
14:34:43 <sbalukoff> evgenyf: Ok, I'm confused then. What's the point of proposing TLS that touches the old model at all?
14:35:13 <ctracey> this should probably be extended to all of lbaas...that we draw a line in the sand that all new features go to v2 only.
14:35:24 <german__> ctracey +1
14:35:25 <sbalukoff> ctracey: +1
14:35:41 <jorgem> ctracey: +1
14:35:41 <TrevorV> ctracey, +1
14:35:52 <johnsom> ctracey: +1
14:36:05 <rm_work> +1
14:36:08 <sballe> +1
14:36:18 <ctracey> this will be hard enough to get done as it is
14:36:27 <sbalukoff> I hear that!
14:36:42 <sbalukoff> Can I get an 'Amen brother'?
14:36:50 <ctracey> and well we have monkey patching :)
14:36:54 <german__> isn't +1 the short form
14:37:21 <sbalukoff> german__: Yea, sorry. Again, I blame it being early. That and I'm a bit of a clown.
14:37:35 <german__> coffee helps
14:37:37 <german__> :-)
14:37:53 <vivek-ebay> ctracey +1
14:37:57 <ptoohill> forgot the 'eh'
14:37:58 <jorgem> So I'm hearing to implement TLS work on refactored code crrect?
14:38:15 <sbalukoff> jorgem: *only* on refactored code.
14:38:15 <german__> yes, all new features only for v2 object model
14:38:50 <ctracey> TLS in this sense would also serve as a carrot to hopefully entice users to upgrade
14:39:02 <german__> we will have a stick, too
14:39:04 <evgenyf> Do we know when to expect the new model code to be commited?
14:39:17 <german__> blogan?
14:39:20 <sbalukoff> Definitely before Juno. ;)
14:39:24 <jorgem> evgenyf: From what blogan tells me it will be pretty soon.
14:39:40 <ptoohill> evgenyf, theres a fork that can be viewed with the changes if youd like to begin some sort of work against the new object model
14:39:46 <jorgem> He went offline as his internet is being flakey right now btw
14:39:56 <evgenyf> Good:) so I will rebase my code on it and re-commit
14:40:03 <ctracey> i will be refocusing some of my efforts to go through the object model today
14:40:07 <sbalukoff> evgenyf: Awesome! Thanks!
14:40:14 <jorgem> evgenyf: Thanks!
14:40:14 <ctracey> giving any help they need there
14:40:32 <jorgem> okay, are we ready to move on to next topic?
14:40:33 <german__> let's not put too much development outside of the official neutron channels
14:40:35 <TrevorV> correct me if I'm wrong, but VijayB was also helping with that
14:40:39 <ptoohill> evgenyf, could you link where your code lives
14:40:43 <sballe> german__, +1
14:40:52 <ctracey> yep
14:41:17 <ctracey> german__: can you elaborate?
14:41:43 <evgenyf> ptoohill: https://review.openstack.org/#/c/102837/1
14:41:44 <german__> if we all work on the branch we will have a huge commit coming
14:41:49 <ptoohill> excellent, thank you
14:41:58 <german__> which might be difficult to bring back in
14:42:06 <german__> so we should stagger it somehow
14:42:29 <VijayB> Hi TrevorV : I just joined - am out of context - are we talking about the neutron interfaces or about the new extension itself?
14:42:33 <ctracey> i have broken my work into multiple reviews
14:42:37 <ctracey> and will be adding more
14:42:48 <german__> ok, just making sure...
14:42:49 <evgenyf> ptoohill: there is that fork you mentioned?
14:42:57 <ptoohill> one sec
14:43:19 <ptoohill> #link https://github.com/brandonlogan/neutron
14:43:21 <german__> I don't want us to create a prallel universe
14:43:25 <ctracey> i have been operating with multiple feature branches that I occasionally roll up into a single feature branch
14:43:27 <ctracey> for testing
14:43:35 <evgenyf> ptoohill: Thank you
14:43:37 <jorgem> german__: you mean multiverse :)
14:43:54 <german__> yep, and then we have trouble unifying it
14:43:56 <ptoohill> there is also the 'base' repo that was linked above
14:43:58 <TrevorV> VijayB, We were talking about new Object Model, and blogan couldn't answer for himself, so since you were working with him I thought you might have an idea, but I think that topic has passed now
14:44:09 <ptoohill> blogan branch forks from here: https://github.com/oslbaas/neutron
14:44:59 <jorgem> Okay, last item is the L7 switching blueprint. avishayb__ could you provide link and start discussion on this?
14:45:16 <VijayB> TrevorV: ah ok, thanks :) I guess we can take up additional questions on #openstack-lbaas later if people have any...
14:45:24 <vjay2> could we have all these links captured (or is it already there?)
14:45:34 <german__> the meeting bot will do it
14:45:35 <avishayb__> here:https://review.openstack.org/#/c/99709/8
14:45:56 <vjay2> i checked this for reviews to be done https://etherpad.openstack.org/p/juno_lbaas_mid_cycle_meetup_reviews
14:46:02 <german__> but jorgem is also adding them to the wiki
14:46:09 <avishayb__> I think we are almost there.. waiting for another review.
14:46:22 <jorgem> german__: correct, I will search and compile a list of relevant links
14:46:41 <german__> thanks
14:46:53 <jorgem> #action Jorge to unify the many links floating around
14:47:19 <vjay2> could you paste the wiki link which is in development here?
14:47:22 <jorgem> I will reach out on the channel today to get links from everyone if they aren't in the weekly standup doc
14:47:44 <jorgem> #link https://wiki.openstack.org/wiki/Neutron/LBaaS
14:47:51 <jorgem> I will use this as the central wiki page
14:47:54 <vjay2> thanks!
14:48:15 <jorgem> Okay we have about 10 mins left.
14:48:31 <evgenyf> for TLS topic: guys please give your feedback for today's commit on https://review.openstack.org/#/c/98640/12
14:48:53 <jorgem> Okay, anything else related to TLS?
14:48:56 <TrevorV> #link https://review.openstack.org/#/c/98640/12
14:48:59 <evgenyf> It makes signs to be completed soon:)
14:49:26 <jorgem> Going once...
14:49:39 <jorgem> Okay, let's switch to last topic then.
14:50:03 <jorgem> avishayb__: L7 switching bp. Go ahead and kick that off please.
14:50:22 <avishayb__> I just did, you missed it..
14:50:39 <jorgem> oops! sorry
14:50:45 <avishayb__> design doc here https://review.openstack.org/#/c/99709/8
14:50:48 <TrevorV> #link https://review.openstack.org/#/c/99709/8
14:50:56 <avishayb__> no pb :)
14:51:00 <jorgem> avishayb__: Anything specific you wanted to talk about?
14:51:01 <sbalukoff> avishayb__: My only comment is that I don't like delaying until after Juno for a few of those compare types (as I've said a few times now)
14:51:29 <avishayb__> sbalukoff I took care of that
14:51:35 <sbalukoff> Oh good!
14:51:41 <sbalukoff> I'll have another look at it today, then.
14:51:51 <avishayb__> thanks
14:52:12 <sbalukoff> Anyway, the L7 stuff is dependent on the new object model and API landing.
14:52:20 <sbalukoff> But it shouldn't be delayed much after that.
14:52:35 <avishayb__> yep - waiting for the fiest commit of the new model
14:52:44 <avishayb__> *first
14:53:17 <sbalukoff> It would be great to get the spec approved as soon as possible, though.
14:53:37 <ptoohill> same with tls specs
14:53:45 <sbalukoff> Yes, indeed!
14:54:00 <sbalukoff> Seems like there's less contention about the L7 spec.
14:54:16 <ptoohill> agreed ;)
14:54:23 <sbalukoff> (Though that may simply be fewer people reviewing it. ;) )
14:54:31 <vjay2> Actually, Haven't had a chance to look at the L7 switching blueprint (i dont know who else other than sbalkoff did)
14:54:35 <german__> that;s what I am thiniking
14:54:47 * vivek-ebay brb
14:54:59 <sbalukoff> So, since this group thrives on contention, let's get a few more eyes on it. ;)
14:55:05 <avishayb__> its time for you guys to have a look there ..
14:55:08 <german__> #action everyone review TLS spec
14:55:22 <sbalukoff> #action everyone review L7 spec
14:55:24 <jorgem> #action Eveyone look at L7 switching blueprint
14:55:36 <sbalukoff> D'oh!
14:55:36 <jorgem> lol
14:55:52 <jorgem> well now it is important lol
14:55:58 <german__> this is settled
14:55:58 <sbalukoff> Doubly so.
14:56:09 <ptoohill> more coffee
14:56:20 <jorgem> Okay so we have about 5 mins left. Here is a recap thus far:
14:56:21 <german__> for the lat 5 minutes, rmwork brabican update or shouldw e take that to the channel
14:56:32 <jorgem> 1) Updates
14:56:33 <jorgem> Located @ https://etherpad.openstack.org/p/neutron-lbaas-weekly-standup
14:56:33 <jorgem> 2) Node pools vs pools naming.
14:56:33 <jorgem> Moving discussion to ML
14:56:33 <jorgem> 3) Mark McClain's flavor blueprint
14:56:33 <jorgem> Everyone please review. Meeting tomorrow on this item.
14:56:33 <jorgem> 4) TLS blueprint update (evgenyf)
14:56:33 <jorgem> We agreed to move dev effort to v2 only.
14:56:34 <jorgem> 5) L7 switching blueprint (avishavb)
14:56:34 <jorgem> Everyone please review bp.
14:56:35 <ptoohill> not sure rm_work is available at the moment
14:56:44 <german__> ok, channel it is then
14:56:46 <rm_work> i am
14:56:51 <ptoohill> so, he is :)
14:57:03 <german__> I saw a +2 -- so we are good now?
14:57:14 <rm_work> Proposing an update to that BP today
14:57:39 <sbalukoff> rm_work great!
14:57:40 <rm_work> splitting the work and into multiple CRs so we can get the first one in and address Arvind and Igor's comments on later CRs
14:57:47 <sbalukoff> Want to add another action item to re-review it?
14:57:52 <rm_work> That's pretty much it
14:58:03 <rm_work> I can poke peope when I post it
14:58:03 <german__> k, sounds good
14:58:07 <rm_work> It'll be later today
14:58:13 <sballe> rm_work, please do
14:58:20 <rm_work> kk
14:58:35 <TrevorV> Just don't poke too hard
14:58:43 <jorgem> Looks like everyone is quite busy so the important thing is to cross-review. At least, that seems like the biggest takeaway from today's meeting.
14:58:49 <jorgem> Good stuff!
14:58:55 <sbalukoff> Yay!
14:59:20 <jorgem> BTW did we decide any deadlines on stuff?
14:59:31 <german__> no,
14:59:36 <jorgem> 30 secs left…ready…set…go!
14:59:43 <sballe> bye
14:59:46 <german__> bye
14:59:46 <jorgem> lol
14:59:53 <avishayb__> bye
14:59:53 <sbalukoff> Haha
14:59:55 <ptoohill> Thank you all, good meeting! buhbye
14:59:56 <jorgem> okay, I guess I'll ask that later
14:59:59 <sbalukoff> Thanks, y'all!
14:59:59 <rm_work> o/
15:00:06 <vjay2> Bye
15:00:07 <TrevorV> \o/
15:00:15 <rm_work> #endmeeting
15:00:15 <jorgem> let me try to end this thing...
15:00:20 <rm_work> nope
15:00:25 <jorgem> #end-meeting
15:00:36 <jorgem> hmm
15:00:40 <jorgem> we don't have access :(
15:00:43 <rm_work> does enikanorov have to do it?
15:00:48 <rm_work> since he started it?
15:00:52 <jorgem> yes until the torch is passed to me hehe
15:00:55 <eglynn> #endmeeting whatever
15:00:55 <rm_work> enikanorov_:
15:01:00 <rm_work> oh
15:01:07 <rm_work> #endmeeting neutron_lbaas
15:01:08 <DinaBelova> :D:D:D
15:01:09 <jd__> now we'll have to wait 2 minutes
15:01:15 <jorgem> #kill −9 meeting
15:01:22 <sbalukoff> Haha
15:01:22 <eglynn> enikanorov: there?
15:01:23 <jd__> it auto end after one hour AFAIK
15:01:23 <llu-laptop> :)
15:01:26 <jorgem> dang
15:01:26 <TrevorV> #sudo kill -9 meeting
15:01:28 <TrevorV> Silly kid
15:01:38 <enikanorov_> #endmeeting