16:01:38 <Sukhdev> #startmeeting ironic_neutron
16:01:39 <openstack> Meeting started Mon Mar  7 16:01:38 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 'ironic_neutron'
16:01:54 <lazy_prince> o/
16:01:55 <vdrok> o/
16:01:59 <vsaienko> o/
16:02:04 <sambetts> o/
16:02:07 <Sukhdev> #topic: Agenda
16:02:13 <Sukhdev> #link: https://wiki.openstack.org/wiki/Meetings/Ironic-neutron#Meeting_March_7.2C_2016
16:02:17 <devananda> o/
16:02:36 <Sukhdev> Feel free to add to the agenda, as we go along
16:02:44 <Sukhdev> #topic: Announcements
16:03:05 <Sukhdev> M3 was last week, right?
16:03:27 <sambetts> Sukhdev: Yes
16:03:38 <sambetts> http://releases.openstack.org/mitaka/schedule.html
16:04:01 <Sukhdev> Anybody has any announcement to make before we dive into the agenda?
16:04:26 <Sukhdev> lazy_prince : welcome back
16:04:46 <Sukhdev> #topic: Update on patches
16:05:13 <Sukhdev> #link: https://etherpad.openstack.org/p/ironic-neutron-mid-cycle
16:05:48 <Sukhdev> thanks to jroll for the refactor effort
16:06:07 <Sukhdev> I notice two patches are already merged
16:07:31 <Sukhdev> devananda : what is the update on the API patches?
16:07:37 <devananda> Sukhdev: jroll is on vacation for the next few days and has asked me to try to keep his refactor patches going
16:08:44 <devananda> I'm juggling a few things, but will try to make some progress on that while he's gone
16:09:47 <Sukhdev> devananda : thanks - what a troop :-)
16:10:16 <sambetts> Is this patch https://review.openstack.org/#/c/206163 likely to make it into Mitaka?
16:10:34 <devananda> sambetts: I do not think so
16:11:08 <sambetts> So Mitaka == No LAG Support?
16:11:31 <Sukhdev> sambetts : yes
16:12:05 <Sukhdev> sambetts : mitaka+portgroup patch = LAG support
16:12:33 <devananda> it also has a lot of -1's and no updates in a month
16:12:52 <Sukhdev> devananda : on the API related patches, other than cleaning up of comments - any other major issue?
16:12:53 <sambetts> Sukhdev: sure, does anyone know if Nova does inter releases or is it going to be Newton we see that release?
16:13:38 <Sukhdev> sambetts : it was held up because of API patches - kind of chicken-n-egg issue
16:13:43 <devananda> sambetts: Nova does coordinated releases with milestones
16:13:49 <devananda> sambetts: so yes, that means Newton
16:14:11 <sambetts> Ah I see, ok thanks
16:14:16 <devananda> Sukhdev: https://review.openstack.org/#/c/285852/ needs reviews
16:14:44 <Sukhdev> devananda : yes it does - I did review it though
16:15:05 <Sukhdev> devananda : it is mostly ready to go
16:15:17 <Sukhdev> lazy_prince : you should look at it as well
16:15:25 <devananda> Sukhdev: thanks. it also needs reviews from ironic cores.
16:15:28 <sambetts> its on the top of my list too
16:15:32 * lazy_prince looking..
16:15:47 <devananda> I will rebuild my local test env to play with it today
16:16:17 <sambetts> I've -1d the driver update patch with regards to it needing rebasing ontop of this one
16:16:19 <Sukhdev> I have not tested after refactor - anybody else had a chance to test?
16:16:40 <vdrok> I will look into that too
16:16:41 <Sukhdev> sambetts: I saw it :-)
16:16:49 <sambetts> Unfortunatly not, been focused on CI work
16:17:37 <Sukhdev> how about ironicclient patch - https://review.openstack.org/#/c/206144
16:17:44 <Sukhdev> this one is ready to go as well
16:18:55 <Sukhdev> this one looks ready as well - https://review.openstack.org/#/c/206244/
16:19:06 <vdrok> yep, but we have to merge portgroups api and network providers before client
16:20:10 <vdrok> and anyway, IIUC ironicclient was released so this won't go in mitaka client even if merged
16:20:53 <Sukhdev> vdrok : are you serious?
16:21:24 <devananda> we can cut a client release right after that
16:21:24 <vdrok> Sukhdev, I think so, client for mitaka is 1.2.0, it was released last week
16:22:10 <Sukhdev> when is mitaka RC1?
16:22:38 <vdrok> Mar 14-18
16:23:24 <sambetts> Final Release for client libs was last week though according to the schedule
16:24:22 <Sukhdev> I did not realize the clients cut-off deadline is earlier!!! yikes
16:24:37 <devananda> Sukhdev: it has been posted on http://releases.openstack.org/mitaka/schedule.html since the cycle started
16:25:12 <devananda> we can cut a 1.3 release of the client and publish it to pypi once the API lands in the server
16:25:30 <devananda> though openstack's coordinated release will continue to use 1.2 as the basis for Mitaka stable testing
16:26:24 <Sukhdev> I see - that works.
16:27:45 <Sukhdev> anything else on the patches?
16:27:59 <Sukhdev> Lets try to review and push them this week, if possible
16:28:41 <Sukhdev> #topic: Vlan aware BMs
16:28:55 <Sukhdev> sambetts : I kept this topic on the agenda for this week as well
16:29:23 <Sukhdev> Do you want to discuss this any further?
16:29:33 <baoli> Sukhdev: I added a few links to that topic, and I'd love to see people commenting on them.
16:30:05 <sambetts> Yup, I want to push for a bit of discussion around the comments left on the RFE too
16:31:40 <sambetts> baoli has left some interesting ideas about how we should handle Ironic port to Neutron port mapping instead of storing them in the extra field, and I'd like to see some feedback there
16:32:32 <sambetts> devananda: do you know when the summit session suggestions open so I can write up something for this like you suggested at the midcycle?
16:35:37 <devananda> sambetts: the newtn design summit is april 25-29. I imagine we'll start discussing the sessions around the first of april
16:35:59 <devananda> sambetts: the list won't be finalized until shortly before the summit, around april 18th
16:36:24 <devananda> sambetts: could you post a link to that discussion from baoli?
16:36:31 <Sukhdev> this will be a good topic for design summit -
16:37:18 <sambetts> devananda: cool, I know jroll was organising the number of sessions last meeting, so I'd guessed it would be once we'd heard back about that
16:37:25 <devananda> yep
16:37:34 <sambetts> devananda: https://review.openstack.org/#/c/277853/2/specs/approved/VLAN-aware-baremetal-instances.rst I can't link to it directly, but L41 there
16:38:29 <devananda> oh, that's going to take more time to read
16:39:07 <devananda> in general, I agree with the sentiment that the ports.extra field should not be used to store anything that affects ironic or neutron behavior in a released feature
16:39:35 <devananda> it's intended to be an opaque field that Ironic code never introspects or depends on
16:40:28 <Sukhdev> devananda: +1
16:41:17 <Sukhdev> anything else on this?
16:41:26 * Sukhdev waiting
16:41:36 <devananda> not from me
16:41:50 <Sukhdev> #topic: Open Discussion
16:41:53 <sambetts> I think further dicussion on the spec would be great :)
16:42:31 <sambetts> So another patch has grabbed my attention recently
16:42:33 <Sukhdev> sambetts: sounds good - I will review baoli 's comment/suggestion as well
16:42:44 <sambetts> its somewhat related
16:42:46 <sambetts> https://review.openstack.org/#/c/284025/
16:43:27 <sambetts> its an interesting bug in the current virt driver code
16:43:32 <Sukhdev> #link: https://review.openstack.org/#/c/284025/
16:45:18 <baoli> Sukhdev: thanks
16:46:30 <vdrok> sambetts, there is another one touching same code - https://review.openstack.org/#/c/287589/4
16:47:44 <sambetts> Different bugs too, but solving the same issue
16:47:55 <Sukhdev> I swear I saw another one over the weekend which touched the same code - can't seem to find the link now :-):-)
16:49:36 <Sukhdev> actually it was the same one - which vdrok posted - sorry
16:49:40 <lazy_prince> was it https://review.openstack.org/#/c/287589/
16:50:17 <Sukhdev> it is the same one
16:50:58 <Sukhdev> interesting!!!
16:51:27 <sambetts> Yeah its an interesting problem, with a fairly simple solution
16:52:00 <Sukhdev> Anything else?
16:52:12 * Sukhdev time check
16:52:51 <sambetts> 8 mins
16:53:06 <lazy_prince> I have one question..
16:53:18 <Sukhdev> lazy_prince : sure
16:53:53 <lazy_prince> with the generic switch driver being pursued by the community, are we still going with vendor specific mech driver..?
16:54:13 <lazy_prince> or will everyone start contributing to the genericSwitch Driver..?
16:55:00 <Sukhdev> lazy_prince: vendor drivers can not be replaced with genericSwitch driver
16:55:07 <Sukhdev> HW vs. SW
16:55:14 <lazy_prince> And is there any plan to support SDN based solutions etc in GenericSwitch Implementation..?
16:55:45 <lazy_prince> Well.. GenericSwitch has support for Cisco switches as of now..
16:56:36 <Sukhdev> lazy_prince: Does Cisco supports it - or stands behind it, if customers hit issues?
16:56:39 <lazy_prince> I just want to know the direction in which community is moving forward..
16:56:43 <sambetts> The Cisco nexus ml2 driver has support for the BM binding type too, and that is where we (Cisco) are supporting our hardware
16:57:09 <sambetts> Sukhdev: we've not been involved in the devlopment of that driver as far as I'm aware
16:57:51 <Sukhdev> I think it was implemented by Mirantis folks
16:57:59 <Sukhdev> to do some testing
16:58:17 <Sukhdev> lazy_prince: this can never replace vender drivers -
16:58:31 <lazy_prince> so correct me if I am wrong.. So the GenericSwitch will be only used and recommended for Gate and Devstack based tests..?
16:58:48 <Sukhdev> lazy_prince : I will think so
16:58:58 <lazy_prince> or is it going to be recommended for production env as well..
16:59:21 <vdrok> generic switch just does ssh into switch and runs some commands. any switch that supports the commands that it runs can be added
17:00:07 <Sukhdev> one could make it work for their specific environment/use case - but, it is not an answer to all
17:00:10 <lazy_prince> may be continue this discussion to netxt monday meeting.. since we are out of time..
17:00:25 <Sukhdev> yup - I will add to the agenda
17:00:32 <Sukhdev> Thanks folks
17:00:41 <Sukhdev> we are out of time and done!!
17:00:43 <Sukhdev> bye
17:00:49 <Sukhdev> #endmeeting