14:20:00 <dougwig> #startmeeting neutron lbaas
14:20:01 <openstack> Meeting started Thu Oct 23 14:20:00 2014 UTC and is due to finish in 60 minutes.  The chair is dougwig. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:20:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:20:06 <openstack> The meeting name has been set to 'neutron_lbaas'
14:20:07 <dougwig> standby, folks
14:20:10 <sbalukoff> Haha!
14:20:12 <dougwig> agenda
14:20:14 <dougwig> #link #link https://wiki.openstack.org/wiki/Network/LBaaS#Meeting_23.10.2014
14:20:18 <dougwig> #topic Announcements
14:20:21 <dougwig> #link https://etherpad.openstack.org/p/lbaas_reviews
14:20:24 <dougwig> #link https://review.openstack.org/#/c/123468/
14:20:28 <dougwig> #link https://etherpad.openstack.org/p/kilo-neutron-summit-topics-distilled
14:20:42 <dougwig> #topic lbaas ref driver - add agent, or octavia?
14:20:43 <ptoohill> Fair enough if thats the case, I wont abandon it yet but will focus on Octavia till we have clear path forward regarding the ref impl.
14:20:44 <dougwig> ok, go
14:21:30 <dougwig> i don't think this decision changes our focus; most of us should still be working on octavia.
14:22:35 <TrevorV_> agreed dougshelley66
14:22:35 <blogan> if we can get v2 into K by updating the agent and current ref impl, is it worth it to do it then?
14:22:39 <TrevorV_> dougwig: *
14:22:42 <ptoohill> Well yea, but there is some changes that would need to be done to get the 'current' ref impl updated and could knock those out early as possible if we were going that route.
14:22:53 <ptoohill> Just wanted to clarify things is all
14:23:21 <dougwig> my sense is that it's not a huge amount of work to get it back with an agent.  am i wrong on that?
14:23:33 <blogan> not huge, but still a decent amount
14:23:59 <ptoohill> also waiting on the barbican interface/workflow to be 'finalized'
14:24:38 <blogan> another question
14:24:44 <blogan> will the octavia driver use the agent?
14:25:11 <dougwig> will its rest api be async?
14:25:16 <blogan> yes
14:25:38 <blogan> but that doesn't mean we have to use the agent
14:25:49 <dougwig> then it depends on whether we put the final status update in a neutron agent or inside octavia?
14:26:46 <blogan> well neutron lbaas is definitely going to need to know about the final status update
14:27:14 <blogan> i just think the agent is overkill, and an extra queue we don't need to go over
14:27:30 <blogan> so the driver can spin up threads
14:27:40 <xgerman> as I said earlier we should focus on Octavia and just let it land when we are ready
14:27:53 <dougwig> right.  which we can either do with an eventlet "agent" inside the neutron-server where the driver was called, as a separate agent, or inside octavia.  a separate neutron agent seems the least clean of all of those solutions.
14:28:05 <ptoohill> Which may not be till L  ?
14:28:13 <xgerman> I am ok with that
14:28:19 <ptoohill> fair enough
14:28:23 <xgerman> (we got new directions)
14:28:35 <blogan> xgerman: oh really?
14:28:52 <dougwig> xgerman: agree, but i'm happy with focusing on neutron lbaas during the neutron lbaas meeting.  :)
14:29:02 <xgerman> :-)
14:29:03 <ptoohill> If thats the case then we wouldnt have to worry about the agent updates which would be the 'bigger' chunk of work for the current ref impl
14:29:11 <blogan> xgerman: new directions that will take you off of octavia?
14:29:44 <xgerman> no, we are just more comfortable with stuff in stackforge as opposed to opesntack
14:29:46 <dougwig> or new directions that you aren't so strict "must be openstack" ?
14:29:50 <dougwig> ahh, nice.
14:29:53 <xgerman> yep, that one
14:29:56 <blogan> well thats a total 180
14:29:58 <sbalukoff> xgerman: Oh, that is new.
14:30:18 <xgerman> yep
14:30:21 <blogan> watch Rackspace become strict about it being openstack now
14:30:26 <xgerman> lol
14:30:40 <sbalukoff> Haha
14:30:48 <dougwig> blogan: split off all of your unprofitable consumer baggage and maybe some of your priorities will change too.
14:30:59 <dougwig> :)
14:31:04 <blogan> lol
14:31:22 <dougwig> maybe they can call the new PC division "Compaq".
14:31:33 <sbalukoff> Haha!
14:31:41 <xgerman> and us DEC?
14:32:07 <dougwig> only if you're running hp cloud on titaniums.
14:32:10 <dougwig> itaniums
14:32:20 <xgerman> ;-)
14:32:31 <dougwig> auto-correct didn't want to acknowledge that processors existence
14:32:45 <dougwig> ok, i've got us in the weeds...
14:32:50 <johnsom> Hey, Itanium had some good ideas....
14:32:50 <dougwig> #topic Open discussion
14:33:03 <dougwig> anything else for today/
14:33:04 <dougwig> ?
14:33:13 <ptoohill> Im on way in, dropping the line. Later guys!
14:33:23 <dougwig> ptoohill: later
14:33:38 <blogan> methinks we're done
14:33:44 <dougwig> alright, if nothing else, let's get out of here early
14:33:46 <dougwig> going once...
14:34:02 <dougwig> going twice...
14:34:15 <sbalukoff> Yay! Back to bed!
14:34:22 <TrevorV_> lulz
14:34:24 <blogan> time change is soon
14:34:26 <TrevorV_> \o
14:34:31 <blogan> this will be even earlier
14:34:32 <xgerman> bye
14:34:38 <dougwig> #endmeeting