16:00:10 <jroll> #startmeeting ironic_neutron
16:00:14 <openstack> Meeting started Mon Aug 29 16:00:10 2016 UTC and is due to finish in 60 minutes.  The chair is jroll. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:16 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:17 <jroll> hi folks
16:00:18 <openstack> The meeting name has been set to 'ironic_neutron'
16:00:48 <jroll> turns out Sukhdev didn't make an agenda for this one, and neither did I
16:00:54 <jroll> I'll roughly follow the last one
16:01:10 <jroll> #link https://wiki.openstack.org/wiki/Meetings/Ironic-neutron#Meeting_August_22.2C_2016
16:01:14 <jroll> is anyone around?
16:01:45 <Sukhdev> I am here off and on - sitting in VMWorld keynote session
16:01:46 <mjturek> sorry I'm late o/
16:02:28 <jroll> \o
16:02:43 <jroll> #topic announcements
16:02:57 <jroll> security groups patch is now up, just waiting on a release note:
16:03:00 <jroll> #link https://review.openstack.org/#/c/361451
16:03:25 <jroll> pretty small, seems reasonable to try to get that in this week
16:03:34 <jroll> I can pick it up if Sukhdev is busy this week to help out
16:03:42 <jroll> portgroups api patches still in review
16:03:46 <jroll> any other announcements?
16:04:14 <mjturek> new patches related to portgroups are in the chain
16:04:38 <mjturek> adds a field for individual_mode so you can detect if fallback is possible
16:04:43 <devananda|OPS> o/
16:04:45 <jroll> cool
16:04:48 <jroll> so I've heard :)
16:05:08 <mjturek> so I don't think the nova patch will be necessary anymore
16:05:35 <jroll> nice
16:06:26 <jroll> so moving on
16:06:28 <jroll> #topic patches
16:06:38 <jroll> we kinda started this already, but seems like things are moving
16:06:47 <jroll> I see some CI failures on a few
16:06:52 <jroll> er, merge conflict
16:06:57 <jroll> https://review.openstack.org/#/q/project:openstack/ironic+branch:master+topic:bug/1526403+status:open
16:07:39 <jroll> anyone have something to chat about on this topic?
16:07:57 <jroll> otherwise I'd say keep reviewing
16:08:46 <jroll> there's a couple other ocata+ things on the agenda, I'd like to skip those and focus on newton things for now
16:08:48 <jroll> so
16:08:51 <jroll> #topic open discussion
16:09:12 <vsaienko1> jroll: I think they stuck in merge conflict today, I'll rebase them if mjturek do not mind
16:09:13 <jroll> (if you have strong objections to skipping over vlan-aware baremetal or interface attach/detach APIs, we can chat about them here)
16:09:23 <mjturek> vsaienko1: sure!
16:09:24 <jroll> vsaienko1: cool
16:09:29 <mjturek> vsaienko1: in regards to your individual_mode patches, think a nova side patch will be needed that leverages individual mode to see if there's a valid network possible?
16:10:02 <Sukhdev> jroll : I just pushed updated version of the patch with release notes - please have a look
16:10:36 <jroll> Sukhdev: cool
16:11:02 <vsaienko1> mjturek:  which exactly?
16:11:30 <mjturek> vsaienko1: sorry, I'm just referring to individual_mode in general
16:12:00 <mjturek> vsaienko1: if I remember correctly, it was added in response to the nova patch to fail if portgroups are used
16:12:23 <vsaienko1> mjturek: yes, sambetts requested to add it
16:12:58 <vsaienko1> jroll, Sukhdev: do we need to check that security group exists before starting deployment? https://review.openstack.org/#/c/361451
16:13:07 <jroll> vsaienko1: yes, I agree we should
16:13:27 <Sukhdev> jroll : not really - it is optional
16:13:54 <jroll> Sukhdev: if configured, we should check that it exists
16:13:57 <Sukhdev> jroll : if it is not specified, then the default group is used
16:15:03 <vsaienko1> Sukhdev: but what if ironic config contains wrong security group uuid? Deployment will fail when we will try to create a port in provisioning network
16:16:05 <Sukhdev> vsaienko1:  hmmm.. I see your point.
16:16:52 <Sukhdev> vsaienko1 : we do not check for the tenant network SG, then why for provisioning?
16:17:16 <jroll> nova verifies the tenant network SG, afaik
16:17:30 <vsaienko1> yeah, I think that nova should do it
16:17:39 <Sukhdev> jroll : right
16:18:13 <vsaienko1> Sukhdev: dtantsur add network validate() in his patch https://review.openstack.org/#/c/357983/ I think it is near close to be merged, you can use the same approach to validate SG
16:18:35 <jroll> but nova doesn't know about the provisioning SG, is the thing
16:19:13 <vsaienko1> jroll: I'm not sure but might be nova adds default SG when we do not specify it? or it is done on Neutron side?
16:19:28 <Sukhdev> jroll : if we want to check the SG, it has to be done in Ironic
16:19:41 <jroll> vsaienko1: idk
16:19:45 <Sukhdev> vsaienko1 : neutron does
16:19:47 <jroll> Sukhdev: right, please add that to the patch
16:20:49 <jroll> anything else for open discussion?
16:23:02 <jroll> alright, thanks for the short meeting everyone :)
16:23:05 <jroll> #endmeeting