21:00:39 <HenryG> #startmeeting networking
21:00:39 <openstack> Meeting started Mon Jun 13 21:00:39 2016 UTC and is due to finish in 60 minutes.  The chair is HenryG. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:43 <openstack> The meeting name has been set to 'networking'
21:00:46 <njohnston> o/
21:00:55 <HenryG> Our PTL is still goofing off on a beach somewhere.
21:01:02 <hoangcx> hi
21:01:08 <HenryG> Let's keep this short.
21:01:11 <fnaval> hi
21:01:19 <Sam-I-Am> hellloooo, and moo.
21:01:25 <hichihara> hi
21:01:29 <HenryG> The agenda for today:
21:01:35 <HenryG> #link https://wiki.openstack.org/wiki/Network/Meetings
21:01:46 <HenryG> #topic Announcements
21:01:55 <HenryG> The mid-cycle is August 17-19 in Cork, Ireland.
21:02:00 <HenryG> #link https://etherpad.openstack.org/p/newton-neutron-midcycle
21:02:08 <HenryG> Please add your name/nick to the etherpad if you plan to attend.
21:02:32 <HenryG> Any other announcements?
21:03:31 <HenryG> #topic blueprints
21:03:43 <HenryG> I have nothing specific, just raising awareness.
21:03:51 <HenryG> #link https://launchpad.net/neutron/+milestone/newton-2
21:03:59 <HenryG> Existing assignments for newton:
21:04:06 <HenryG> #link https://blueprints.launchpad.net/neutron/newton/+assignments
21:04:15 <HenryG> If you are an assignee or approver for a blueprint please work together to ensure the BP whiteboard is up-to-date.
21:04:28 <HenryG> N-2 is in about one month.
21:04:34 <HenryG> #link http://releases.openstack.org/newton/schedule.html
21:04:59 <HenryG> Does anyone have a specific BP or RFE to discuss today?
21:05:16 <kevinbenton> vlan-aware-vms
21:05:31 <HenryG> what about it?
21:05:38 <kevinbenton> there was a mailing list thread about changes to nova/os-vif
21:05:52 <kevinbenton> however, i wasn't aware of any required changes to those components
21:06:00 <kevinbenton> carl_baldwin: you started the thread, can you elaborate a bit?
21:06:05 * mestery walks in really late
21:06:20 <HenryG> #link http://lists.openstack.org/pipermail/openstack-dev/2016-June/097025.html
21:06:32 <kevinbenton> http://lists.openstack.org/pipermail/openstack-dev/2016-June/097025.html
21:06:38 <kevinbenton> whoops, HenryG beat me to it
21:07:03 <kevinbenton> if we do require a change to os-vif it's going to be a big risk to actually getting this feature completed
21:07:39 <carl_baldwin> kevinbenton: There has been a lot of discussion around it.
21:08:31 <carl_baldwin> kevinbenton: From Sean Mooney's reply, it looks it might need to be modified to create a (trunk) bridge if it doesn't exist.
21:09:26 <kevinbenton> carl_baldwin: i see, but even then we could workaround by creating the trunk bridge on the L2 agent automatically
21:09:31 <carl_baldwin> kevinbenton: Other proposals suggested a little more intrusive change, like plugging patch ports, etc.
21:09:57 <carl_baldwin> kevinbenton: I think the issue there is the race to plugin in to it.
21:10:33 <kevinbenton> carl_baldwin: ok. if that's the only bit that is needed then it's not too big of a deal
21:11:05 <kevinbenton> carl_baldwin: there are always backup options for that (let nova plug it into br-int and then we move it to a trunk bridge we create
21:11:08 <kevinbenton> )
21:11:25 <kevinbenton> if we can't get something into os-vif in time
21:11:32 <carl_baldwin> kevinbenton: Yeah.
21:11:51 <carl_baldwin> kevinbenton: I think the list made it apparent that the changes there should be minimized, especially in the short term.
21:11:59 <carl_baldwin> s/list/thread/
21:12:20 <kevinbenton> ok, that's all i had for that topic
21:12:38 <HenryG> thanks kevinbenton and carl_baldwin
21:12:58 <HenryG> #topic Bugs
21:13:14 <HenryG> Last week the bug deputy was ajo
21:13:35 <HenryG> (I think)
21:13:53 <HenryG> Anything to report, ajo ?
21:14:31 <HenryG> I guess it's night time in his time zone
21:14:41 <HenryG> Any critical or high bugs we should discuss?
21:15:40 <dougwig> does the dhcp numbering count?
21:16:04 <kevinbenton> no
21:16:22 <dougwig> no, it doesn't count, or no, you don't want a holy war on this fine monday?
21:16:31 <kevinbenton> no, i don't think it counts
21:16:38 <kevinbenton> it's not blocking anything
21:16:53 <kevinbenton> so from a bug perspective it's not high/critical
21:18:37 <HenryG> This week the deputy is dasm
21:18:41 <dasm> Aye, aye, captain!
21:18:49 <HenryG> The next unassigned week is July 4. Any takers? It should be a quiet week.
21:19:23 <dougwig> i'll be camping.
21:20:15 <HenryG> We'll ask for volunteers again at the next meeting.
21:20:20 <johndperkins> I'll do it
21:20:22 <Sam-I-Am> dougwig: what is this nature thing
21:20:36 <HenryG> johndperkins: great! thanks!
21:20:44 <dougwig> Sam-I-Am: it's where i play with my solar install and watch an ammeter all day.
21:20:47 <johndperkins> np
21:20:57 <Sam-I-Am> dougwig: you going to go vegan on me too?
21:21:08 <mlavalle> lol
21:21:14 <HenryG> #topic Docs
21:21:50 <HenryG> Sam-I-Am?
21:21:55 <Sam-I-Am> hey!
21:22:04 <HenryG> You're up!
21:22:12 <Sam-I-Am> still improving the networking guide
21:22:31 <Sam-I-Am> some folks have been working on existing content, such as subnet pools and address scopes
21:22:54 <HenryG> good to hear
21:22:55 <Sam-I-Am> i'm sort of working on restructuring the guide because too many things are ending up in 'advanced config'
21:23:04 <Sam-I-Am> so expect that section to go away
21:23:47 <Sam-I-Am> there will probably be a section of well-baked features including config, and another section for more experimental features
21:24:20 <Sam-I-Am> experimental means its either new, not well-tested, doesnt have sufficient docs, or isnt particularly easy to deploy
21:25:14 <Sam-I-Am> we'll see how it turns out... hopefully have a patch up this week, and if its too messy, take it to a list
21:25:20 <dasm> Sam-I-Am: "Feature Classification Framework"?
21:25:32 <njohnston> That sounds like a good differentiation; somewhat subjective, but highly useful
21:25:35 <kevinbenton> what does 'well-tested' mean in this case?
21:26:01 <Sam-I-Am> kevinbenton: deployable outside of devstack :)
21:26:20 <Sam-I-Am> as in, not "works in devstack"
21:26:38 <Sam-I-Am> or if its something that needs multi-node to be useful, testing in a multi-node env
21:26:55 <kevinbenton> right, so we have multi-node jobs now
21:26:59 <kevinbenton> so what i'm getting at
21:27:16 <kevinbenton> is if it has good testing coverage in tree
21:27:17 <Sam-I-Am> well, we do to some extent
21:27:26 <kevinbenton> is that good
21:27:59 <Sam-I-Am> i think it depends what it is, and might involve working with the developers to see how much it was tested outside of automation
21:28:06 <Sam-I-Am> like... did anyone scale it
21:28:56 <kevinbenton> so the concern i have is that this is going to become very subjective quickly
21:29:23 <kevinbenton> but we can always lay out guidelines later if there is a disagreement
21:29:37 <njohnston> +1
21:29:42 <Sam-I-Am> i dont really have answers. a lot of it is how thorough are the docs for actually deploying whatever it is
21:30:09 <Sam-I-Am> something with "set this flag in devstack" is not going to get out of the experimental section
21:30:49 <kevinbenton> yeah, i'm fine with the deployment docs requirement. that's pretty easy
21:30:54 <kevinbenton> to see what is required
21:30:57 <Sam-I-Am> i would like users of the guide to have some basic expectations of what they're deploying
21:31:44 <Sam-I-Am> either way, look for a simpler organization to the guide - intro content, standard config, experimental config, and some misc stuff.
21:31:59 <kevinbenton> makes sense
21:32:09 <dasm> Sam-I-Am | there will probably be a section of well-baked features including config, and another section for more experimental features
21:32:19 <dasm> Sam-I-Am: are you talking about Feature Classification Framework?
21:32:23 <dasm> https://review.openstack.org/#/c/324048/
21:32:24 <Sam-I-Am> right now i think everything going into 'adv config' is scaring people away from trying not-so-complex features
21:32:50 <kevinbenton> Sam-I-Am: +1
21:33:13 <Sam-I-Am> dasm: yeah, unless there's more to it
21:33:35 <HenryG> Thanks Sam-I-Am
21:33:44 <dasm> Sam-I-Am: there are some plans to move it to oslosphinx: https://bugs.launchpad.net/oslosphinx/+bug/1592128
21:33:44 <HenryG> One more docs thing ...
21:33:45 <openstack> Launchpad bug 1592128 in oslosphinx "Feature Classification Framework" [Undecided,New] - Assigned to Ankur (ankur-gupta-f)
21:34:14 <HenryG> sorry dasm
21:34:40 <dasm> HenryG: no problem. I just sent info and finished. Proceed
21:34:52 <HenryG> dasm: thanks for the info
21:34:56 <HenryG> One more docs thing ...
21:35:08 <HenryG> Note that the api-ref for Neutron is now in neutron-lib!
21:35:22 <HenryG> #link https://review.openstack.org/327510
21:35:29 <HenryG> Thanks to amotoki for that.
21:35:38 <HenryG> Now we can start correcting and fleshing our API reference docs.
21:35:50 <Sam-I-Am> this is using the new api stuffs?
21:35:55 <HenryG> yep
21:36:37 <Sam-I-Am> cool, glad to see that
21:36:43 <Sam-I-Am> so how has moving the api stuff in-tree helped?
21:36:47 <Sam-I-Am> or has it helped...
21:36:59 <HenryG> It just merged.
21:37:14 <dasm> nice job amotoki. Thanks
21:37:20 <HenryG> The advantage is mostly that it is now RST.
21:37:29 <Sam-I-Am> well, yeah
21:37:50 <Sam-I-Am> but i guess you can also include changes to it in code patches
21:37:59 <Sam-I-Am> so, less overhead?
21:38:07 <HenryG> right
21:38:43 <Sam-I-Am> cool. gives me some good feedback... and ideas.
21:39:06 <Sam-I-Am> out of curiosity, would having the networking guide in-tree increase contributions?
21:39:11 <Sam-I-Am> basically ops/user docs
21:39:32 <HenryG> Sam-I-Am: Not sure. Let's see how the api-ref goes.
21:40:00 <Sam-I-Am> HenryG: okey dokey
21:40:06 <Sam-I-Am> that's all from meeeeee
21:40:35 <HenryG> Thanks Sam-I-Am
21:40:38 <HenryG> Any other docs items?
21:41:19 <HenryG> #topic Open Discussion
21:42:05 <njohnston> How is the neutron-api project coming along?  I haven't heard much about it since the summit, but IIRC adoption of it was part of the requirements for eligibility to remain in the stadium.
21:42:20 <rtheis> FYI: OSC 2.6.0 was released with more network support (see https://etherpad.openstack.org/p/osc-neutron-support for status)
21:43:01 <rtheis> Also osc-lib is now available and can be used for neutronclient's OSC plugin
21:43:48 <hichihara> rtheis: Cool
21:43:51 <HenryG> njohnston: Last I heard from armax is he will use neutron-lib as the home for the neutron-API
21:44:35 <njohnston> HenryG: OK. So still pretty much future-tense at this point, you think?
21:45:01 <HenryG> njohnston: at this point, yes.
21:45:29 <njohnston> HenryG: Thanks!   That sets my expectations.
21:46:54 <HenryG> rtheis: thanks, and I would like to remind folks to get into the habit of using "openstack network ..."  instead of  "neutron ..."  now.
21:47:05 <Sam-I-Am> ^ this
21:47:14 <Sam-I-Am> well, networking service i guess
21:47:58 <HenryG> Sam-I-Am: I meant at the CLI
21:48:07 <Sam-I-Am> oh, yeah that too :)
21:49:48 <manjeets> HenryG: already to that habit of using openstack ** instead of neutron **
21:50:06 <HenryG> manjeets: +1
21:50:26 <Sam-I-Am> trying to convert all docs to using 'openstack' if possible
21:50:49 <HenryG> that must be painful
21:51:00 <Sam-I-Am> da
21:51:21 <twm2016> Hi everyone, I would like to introduce myself. My name is Trevor, I met some of you at the summit in Austin, and I will be *trying to actively contribute to "openstack networking" where I can. :)
21:51:46 <HenryG> Welcome twm2016 !
21:51:59 <hichihara> twm2016: welcome :)
21:52:07 <Sam-I-Am> hello
21:52:35 <njohnston> twm2016: Greetings!  I thought your nick was a reference to ye olde xwindows window manager 'twm'.
21:52:55 <twm2016> I'm looking forward to getting to know all of you. I hate my nick, I missed the naming convention when creating an account.
21:53:19 <twm2016> - I wanted to match my gerrit username. If you all must know its my initials plus the year I registered.
21:54:04 <twm2016> oh well, I hope some of you have seen my code reviews and found them useful. I look forward to getting to know all of you better
21:54:06 <HenryG> twm2016: freenode nicks are nothing. Just create another.
21:54:29 <njohnston> +1
21:54:45 <twm2016> HenryG: Thanks, I guess I'll do that ( action item? ) haha
21:54:54 <HenryG> OK, everybody gets 5 minutes back.
21:54:58 <HenryG> #endmeeting