16:01:40 <Sukhdev> #startmeeting networking_ml2
16:01:40 <openstack> Meeting started Wed Oct 12 16:01:40 2016 UTC and is due to finish in 60 minutes.  The chair is Sukhdev. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:44 <openstack> The meeting name has been set to 'networking_ml2'
16:01:58 <Sukhdev> andreas_s : are you here?
16:02:08 <andreas_s> Sukhdev, yup
16:02:09 <andreas_s> hi
16:02:14 <Sukhdev> yamahata : are you here for ML2 meeting
16:02:21 <Sukhdev> andreas_s : hi
16:02:42 <Sukhdev> #topic: Agenda
16:02:50 <Sukhdev> #link: https://wiki.openstack.org/wiki/Meetings/ML2#Meeting_October_12.2C_2016
16:03:06 <Sukhdev> #topic: Announcements
16:03:43 <Sukhdev> rkukura is celebrating holiday festivities - so, he will not be joining us today
16:04:37 <Sukhdev> I was in the middle of job change - hence, some interruptions to these meetings for past few weeks
16:04:51 <Sukhdev> hopefully, things will go smooth going forward
16:05:15 <Sukhdev> Anybody has any announcements?
16:05:22 <andreas_s> oh ok - but you're still working on Neutron?
16:05:28 <andreas_s> seems so..
16:05:32 <Sukhdev> yes
16:06:08 <andreas_s> cool!
16:06:15 <andreas_s> no announcements
16:07:11 <Sukhdev> #topic: Ocata planning
16:07:28 <Sukhdev> rkukura and I discussed this briefly yesterday
16:08:12 <Sukhdev> Since he is not here today - I think perhaps we postpone this discussion to next week's meeting
16:08:38 <andreas_s> will you guys be at the summit?
16:08:49 <Sukhdev> I asked him to think about the issues/agenda we want to cover in Ocata release
16:09:12 <Sukhdev> andreas_s : rkukura is planning on skipping - I will be there most likely
16:09:38 <andreas_s> Sukhdev, cool! I'll be there as well
16:09:39 <Sukhdev> andreas_s : I have not made any travel arrangements yet - but, will know details by next meeting
16:09:54 <andreas_s> ok
16:10:05 <Sukhdev> I am thinking we will schedule a session for ML2 - same way we did in Austin
16:11:01 <Sukhdev> andreas_s : plan on attending next week's meeting - we will go over some of the ideas as to what we should cover in this cycle
16:11:14 <Sukhdev> we can cover details in the summit
16:11:31 <andreas_s> ok, I'll be joining next week!
16:11:52 <Sukhdev> #topic: Bugs/RFEs
16:12:28 <Sukhdev> andreas_s : floor is yours - can you update the team?
16:12:38 <andreas_s> ok
16:12:58 <andreas_s> Regarding "portbinding information for nova live migration"
16:13:15 <andreas_s> Seems like this is getting more and more attention
16:13:25 <andreas_s> #link https://review.openstack.org/#/c/309416/
16:13:35 <andreas_s> john Garbutt started writing a nova spec for this
16:13:44 <andreas_s> #link https://review.openstack.org/375580
16:14:20 <andreas_s> and this topic will be discussed in the nova/neutron crossproject session on Wednesday
16:14:34 <andreas_s> #link https://etherpad.openstack.org/p/ocata-nova-neutron-session
16:15:11 <andreas_s> but I'm still not sure if the Neutron design (like I proposed) will work out...
16:15:42 <Sukhdev> I saw the earlier version of your spec - have not had a chance to review the latest version -
16:15:57 <andreas_s> also John seems to have a different understanding of what he gets.... he was expecting a general portbinding API - that's not what we have planned
16:16:11 <andreas_s> Sukhdev, no, no updates, maybe a sequence diagram or so, but nothing content wise
16:16:41 <andreas_s> so if you guys could join that Wednesday session, that would be helpful!
16:17:00 <Sukhdev> so, cross project discussion is excellent idea to close that gap
16:17:15 <Sukhdev> Yes, I will plan on participating in this session
16:18:04 <andreas_s> Sukhdev, so you're the opinion that splitting up the portbinding into an activation stage and a creation stage is technically doable?
16:18:48 <andreas_s> create would only call bind_port - activate would call update_port_*commit
16:19:29 <Sukhdev> this will impact lots of mech drivers
16:20:10 <Sukhdev> andreas_s : we have to keep in mind that there are baremetal (BM) use cases which rely on bind_port calls
16:20:18 <andreas_s> yes - that's why I'm not 100% convinced that this is the right approach...
16:20:19 <andreas_s> ok
16:20:24 <Sukhdev> also the HPB use cases as well
16:20:49 <Sukhdev> we have to make sure that we maintain the backward compatibility
16:21:12 <Sukhdev> otherwise we will have very unhappy folks :-)
16:21:35 <Sukhdev> Hence, a cross project discussion is a great idea
16:21:51 <andreas_s> Sukhdev, right - I guess my point is - I can't asess the impact of this change... that's were I need the expertise of the ml2 folks
16:22:03 <andreas_s> The alternative would be to have 2 active ports
16:23:11 <andreas_s> but probably some mech drivers won't support that - so some legacy mode would be required as well...
16:25:06 <Sukhdev> andreas_s : One suggestion -
16:25:22 <Sukhdev> put these issues and ideas in the etherpad
16:25:35 <Sukhdev> this will give folks something to think about ahead of time
16:25:43 <andreas_s> Sukhdev, good point!
16:26:26 <Sukhdev> wisdom states that nothing will be accomplished in 40 minute session - unless, people are well prepared with various ideas
16:27:12 <andreas_s> yeah. I hope that we can at least make a decision if this is something nova/neutron wants or not
16:27:34 <Sukhdev> once people come into the session prepared to discuss the details, then we can make decisions
16:27:34 <andreas_s> I'm mean not the prposal in detail, but the idea of fixing the communicaton
16:28:09 <Sukhdev> right - that is a good point - but, you want to be ready with some concrete details
16:28:18 <andreas_s> ok, so don't have anything else on that other than I will update the etherpad...
16:28:22 <andreas_s> I
16:28:36 <Sukhdev> sounds good - thanks for the update
16:28:45 <andreas_s> welcome
16:29:11 <Sukhdev> anybody else has anything they would like to discuss in the agenda?
16:29:24 <Sukhdev> #topic: Open Discussion
16:29:58 <andreas_s> nothing from my side
16:30:17 <Sukhdev> any body wants to cover anything related to ML2 or barcelona summit?
16:30:28 * Sukhdev waiting
16:30:58 <Sukhdev> I guess we are done
16:31:03 <Sukhdev> Thanks for attending
16:31:11 <Sukhdev> bye
16:31:16 <Sukhdev> #endmeeting