16:02:41 <Sukhdev> #startmeeting networking_ml2
16:02:42 <openstack> Meeting started Wed Mar 18 16:02:41 2015 UTC and is due to finish in 60 minutes.  The chair is Sukhdev. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:43 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:02:45 <openstack> The meeting name has been set to 'networking_ml2'
16:03:04 <Sukhdev> Folks, welcome to ML2 weekly meeting
16:03:14 <Sukhdev> #topic: Agenda
16:03:16 <yamahata> hello
16:03:20 <Sukhdev> #link: https://wiki.openstack.org/wiki/Meetings/ML2
16:03:58 <Sukhdev> #topic: Announcements
16:04:09 <Sukhdev> I have couple of announcements -
16:04:33 <Sukhdev> As you may know I chair L2-Gateway subgroup meetings
16:05:05 <Sukhdev> and, we announced this week that the initial version of L2 GW API is implemented and ready for testing
16:05:16 <shivharis-1> congratulations!
16:05:31 <Sukhdev> I thought I pass that information here - this may be of use to you folks
16:05:38 <yamahata> awesome
16:05:43 <yalie> con!
16:05:50 <Sukhdev> The wiki is here - https://wiki.openstack.org/wiki/Neutron/L2-GW
16:06:02 <Sukhdev> There is lots of good information
16:06:15 <Sukhdev> yamahata, yalie shivharis-1: Thanks
16:06:42 <Sukhdev> At the end of the wiki page, you will se devstack instructions to launch the L2 Gateway
16:07:18 <rkukura> Sukhdev: Wiki looks great! Does this involve any changes to ML2 or its L2 agents?
16:07:22 <Sukhdev> We tested the full connectivity between VMs running on neutron networks with the bare metal servers running in legacy networks
16:07:39 <Sukhdev> rkukura: No
16:07:57 <rkukura> even better ;)
16:08:30 <Sukhdev> It works in conjunction with ML2 - it is implemented as a service plugin - like LBaas, VPNaas, etc
16:08:57 <Sukhdev> Initial implementation uses OVSDB
16:09:33 <Sukhdev> hardware VTEP schema - so VXLAN to VLAN interconnectivity via HW switches
16:09:45 <shivharis-1> curious, what else can be used other than ovsdb?
16:10:17 <Sukhdev> shivharis-1: back-end can be anything - initial implementation uses OVSDB
16:10:32 <Sukhdev> This is just like any other service API
16:11:14 <Sukhdev> Another announcement is K3 feature freeze is this week -
16:11:15 <shivharis-1> Sukhdev: will talk to u offline, this is still announcements
16:11:28 <Sukhdev> shivharis-1: sure
16:11:42 <Sukhdev> K3 content is here - https://launchpad.net/neutron/+milestone/kilo-3
16:12:05 <Sukhdev> If you are working on anything for K3, be sure your patch/bug is included in this list
16:12:21 <Sukhdev> Anybody has any other announcement to make?
16:12:30 * Sukhdev waiting
16:12:57 <Sukhdev> #topic: ML2 Drivers Decomposition
16:13:24 <Sukhdev> For past several weeks, this has been most discussed topic in this meeting
16:13:45 <Sukhdev> I noticed that a lots of progress has been made in this regard
16:13:56 <Sukhdev> I see lots of drivers have moved over already
16:14:06 <sadasu> I have a question for armax regarding : https://review.openstack.org/#/c/155436/14
16:14:43 <sadasu> One review comment regarding VIF_DRIVER_PROFILEID being moved over to vendor side
16:15:05 <sadasu> is armax here?
16:15:30 <Sukhdev> sadasu: those comments seem very trivial
16:15:53 <sadasu> yeah, before I make the change, need to  make sure
16:16:17 <sadasu> this constant defines the key for the VIF_DETAILS dictionary
16:16:22 <Sukhdev> sadasu: what is the concern?
16:16:49 <sadasu> just not sure if it is a good idea to move it to vendor side, it is consumed by Nova VIF driver and from their perspective
16:17:15 <sadasu> won't it be difficult to find where this is being set?
16:17:59 <Sukhdev> sadasu: His is suggesting put it where it is consumed - perhaps he is not aware who the consumers are
16:18:29 <sadasu> also, VIF_DETAILS is meant to hold feature specific values, do we want those keys defined in one place rather than spread over?
16:18:32 <Sukhdev> sadasu: So, if you believe, it belongs here, just simply answer in the review comments, and that should be good
16:19:07 <sadasu> Sukhdev: did that already...
16:19:51 <Sukhdev> sadasu: then I do not think there should be an issue -
16:20:14 <sadasu> I understand the overall goal is not to contaminate Neutron code with vendor stuff, but I believe this doesn't fall under that category
16:20:18 <sadasu> lets move on
16:20:34 <Sukhdev> Anything else?
16:20:42 <Sukhdev> https://github.com/openstack/neutron/blob/master/doc/source/devref/contribute.rst#how-to
16:20:57 <Sukhdev> See the growing list of drivers/plugin
16:21:35 <Sukhdev> Anything else on this topic?
16:22:04 * Sukhdev waiting
16:22:36 <Sukhdev> #topic: ML2 Sync/Task Flow
16:23:14 <Sukhdev> Looks like manish is not here
16:23:14 <Sukhdev> https://review.openstack.org/#/c/154333/
16:23:24 <Sukhdev> Did folks have opportunity to review the patch?
16:24:03 <Sukhdev> I hope I was able to articulate the issue which I see which can be solved by this
16:24:19 <rkukura> I have not yet had a chance to followup to the followups on my comment
16:25:20 <Sukhdev> last week manish mentioned that he will take the discussion off-line. I was hoping to see an invite for discussion - did not see it
16:25:49 <shivharis-1> i think most folks were busy with k3 deliverables
16:26:07 <Sukhdev> shivharis-1: I think so
16:26:08 <rkukura> shivharis-1: right - I’m hoping we can focus more on this between kilo and the summit
16:26:35 <shivharis-1> rkukura: me too
16:26:38 <Sukhdev> I actually like what he has in the patch - so, this is a good start
16:27:14 <Sukhdev> Anyway, since manish is not here, I suggest we table this
16:27:45 <Sukhdev> Unless anybody want to express openion on this?
16:27:59 * Sukhdev waiting
16:28:24 <Sukhdev> #topic: Bugs
16:28:42 <Sukhdev> shivharis-1: ola
16:28:43 <shivharis-1> We still have 4 high-prio bugs
16:29:22 <shivharis-1> i am not sure if these will make it at this stage,these may have to be moved for the final K release
16:30:07 <rkukura> shivharis-1: I’m making progress on https://bugs.launchpad.net/neutron/+bug/1367391, but there is substantial work remaining
16:30:08 <openstack> Launchpad bug 1367391 in neutron "ML2 DVR port binding implementation unnecessarily duplicates schema and logic" [High,In progress] - Assigned to Robert Kukura (rkukura)
16:30:42 <shivharis-1> rkukura: this can then be K release
16:31:05 <shivharis-1> rkukura: heavens wont fall if it is moved out, right
16:31:38 <rkukura> I don’t see a problem with missing k-3, other than less chance of it making kilo at all
16:31:50 <shivharis-1> i think all seem to fall in that category, i am not holding my breath for these to get into k3
16:32:31 <shivharis-1> i am hoping we make these in K, atleast
16:32:50 <Sukhdev> we still have lots of time for K3
16:32:51 <shivharis-1> Anyother questions folk have on bugs?
16:33:06 <shivharis-1> anyone needs review help for medium bugs?
16:33:29 <shivharis-1> Sukhdev: K3 in tomorrow
16:34:11 <Sukhdev> shivharis-1: I do not believe that is true for the bugs - they can continue to go in all the way to the end of the month, I beieve
16:34:12 <shivharis-1> Sukhdev: k3 moved out?
16:34:50 <Sukhdev> shivharis-1: depends upon the criticality of the bug -
16:34:58 <rkukura> launchpad says kilo-3 is expected in 7 hours
16:34:59 <banix__> Sukhdev: that’s correct
16:35:00 <shivharis-1> Sukhdev: this is not my undestanding, once it is tagged it is over - we move to the next milestone
16:35:52 <Sukhdev> I may be confused between k3 and RC
16:35:55 <banix__> kilo-3 is tomorrow but bug fixes can get in afterward
16:36:11 <Sukhdev> we can get this bug into Kilo release itself
16:36:12 <shivharis-1> I will verify with infra
16:36:25 <banix__> #link https://wiki.openstack.org/wiki/Kilo_Release_Schedule
16:36:53 <Sukhdev> banix__: thanks
16:37:14 <shivharis-1> looking...
16:37:18 <Sukhdev> yes, the release is not until end of April - so, I believe we have time to get the bugs in
16:38:18 <shivharis-1> k3 is Mar 19, you can continue on bugs for K release after that - but not the k3 milestone
16:38:37 <shivharis-1> k3 is a tagged in the code base
16:38:43 <Sukhdev> shivharis-1: correct
16:39:12 <Sukhdev> shivharis-1: for the features
16:39:35 <Sukhdev> Anything else on the bugs?
16:39:36 <rkukura> Keep in mnd that only critical bug fixes are likely to get in after kilo-rc1, which is scheduled for early April, I think
16:39:52 <Sukhdev> rkukura: correct
16:40:22 <shivharis-1> anything else on bugs/any help needed for reviews?
16:40:23 * Sukhdev waiting
16:40:49 <Sukhdev> #topic: Open Discussion
16:41:09 <Sukhdev> Any body wants to discuss anything?
16:41:37 * Sukhdev waiting
16:41:59 <Sukhdev> Looks like we are done
16:42:03 <banix__> bye
16:42:08 <Sukhdev> Thanks folks for joining
16:42:13 <Sukhdev> bye
16:42:17 <shivharis-1> bye
16:42:20 <rkukura> bye
16:42:21 <Sukhdev> #endmeeting