17:00:17 <kiall> #startmeeting Designate
17:00:18 <openstack> Meeting started Wed Jun 17 17:00:17 2015 UTC and is due to finish in 60 minutes.  The chair is kiall. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:20 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:22 <openstack> The meeting name has been set to 'designate'
17:00:25 <kiall> Hey folks, who's around today?
17:00:27 <timsim> o?
17:00:30 <mlavalle> hi
17:00:34 <elarson> o)
17:00:36 <mugsie> o/
17:00:43 <jmcbride> o/
17:01:15 <kiall> #topic Action Items
17:01:54 <kiall> two open, both on me, one done.
17:01:54 <kiall> Update Neutron external DNS spec - done, will circle back
17:01:54 <kiall> Write up DB rework - still on my TODO list, apologies.
17:02:33 <kiall> (It's been one of those weeks, barely been at my desk all week)
17:02:42 <kiall> Moving on quickly before I get lynched
17:02:42 <kiall> #topic Bug triage (timsim - recurring)
17:03:01 <timsim> https://bugs.launchpad.net/designate/+bug/1463988
17:03:02 <openstack> Launchpad bug 1463988 in Designate "oslo.config version error in kilo" [Undecided,New]
17:03:35 * timsim thinks this is related to that pbr thing.
17:03:36 <kiall> I tried to repro that and couldn't? Anyone else?
17:04:00 <timsim> I think pglass did
17:04:02 <mugsie> timsim: i think you are right
17:04:20 <kiall> pglass: about?
17:04:32 <timsim> Depending on the environment, that pbr business still pops up
17:04:44 <pglass> yeah, the pbr thing pops up from time to time
17:05:05 <kiall> Oh, nice.. So it's one of those sometimes works sometimes fails? any clue what the trigger is?
17:05:35 <pglass> it should happen with stable/kilo all the time
17:06:06 <kiall> Oh, Humm.. Maybe I tested on the wrong branch.. If it's happening all the time, mark it High and we'll need to get it sorted soon :/
17:06:58 <timsim> I'll see if I can reproduce it again, and mark it.
17:07:18 <timsim> https://bugs.launchpad.net/designate/+bug/1464763
17:07:19 <openstack> Launchpad bug 1464763 in Designate "Incorrect status code on a PATCH /v2/pool/{id} ?" [Undecided,New]
17:07:26 <kiall> thanks, trying again on stable/kilo myself too.. I must have used thw wrong branch.
17:07:36 <pglass> there was a bug open for the pbr thing in heat, and someone working on solum here ran into that issue too
17:07:39 <mugsie> l1, high - assign to me
17:07:58 <kiall> I tend to just go with mugsie on (most) API things ;)
17:08:13 <kiall> pglass: any clue what the bug #'s were? if not, no worries
17:08:23 <mugsie> docs are right, code is wrong
17:08:40 <pglass> It put the heat bug in a comment on ours: https://bugs.launchpad.net/heat/+bug/1461592
17:08:42 <openstack> Launchpad bug 1461592 in heat kilo "stable\kilo is broken due to version conflict error" [Critical,Triaged] - Assigned to Kairat Kushaev (kkushaev)
17:08:50 <pglass> Idk if there are other bugs open
17:08:55 <timsim> Done mugsie
17:09:13 <kiall> pglass: ah, perfect.
17:09:57 <kiall> timsim: thats the lot of new stuff to look at?
17:10:04 <timsim> Yep. That's all
17:10:22 <kiall> Cool, moving on so..
17:10:22 <kiall> #topic Summit Session Review
17:10:22 <kiall> Almost done with this recurring topic ;)
17:10:30 <kiall> Neutron External Spec updated
17:10:31 <kiall> #link http://docs-draft.openstack.org/24/88624/4/check/gate-neutron-specs-docs/36e883b//doc/build/html/specs/liberty/external-dns-resolution.html
17:10:46 <kiall> and
17:10:49 <mlavalle> kiall: thanks, I will take a look
17:11:17 <kiall> #link https://review.openstack.org/#/c/88624/
17:11:46 <kiall> mlavalle: cool,there's a TODO in there around the network's new dns_domain property
17:12:00 <mlavalle> will look for it
17:12:03 <kiall> That's listed as part of the external spec, but is it not needed by the internal spec too?
17:12:30 <mlavalle> i'll take a look
17:12:47 <kiall> K - let me know and I'll update :)
17:13:18 <kiall> Any progress on the nova/neutron-internal specs? Hopefully can get these landed soon :)
17:13:33 <kiall> (specs, not code ;))
17:14:08 <mlavalle> kiall: on the nova side I refresehd it and got  a +1
17:14:14 <mlavalle> so it is moving along
17:14:27 <kiall> Sounds like the OpenStack specs process ;)
17:14:51 <mlavalle> on the neutron side https://review.openstack.org/#/c/88623/14 we got an objection on line 139
17:15:04 <mlavalle> kiall: carl_baldwin and I would like you to comment on it
17:15:05 <kiall> Okay, we should probably drop these on the Nova/Neutron weekly meet agends to see if we can move things forwards, assuming no negative reviews.
17:15:38 <kiall> mlavalle: ah, right.. Don't think I've been at my desk since you asked :) Will review right after this!
17:15:49 <mlavalle> kiall: thanks
17:16:07 <mlavalle> kiall: the other thing is that since late last week I am already coding the neutron side
17:16:20 <mlavalle> making good progress, working right now on the db
17:16:45 <kiall> Cool - looking forward to PS1 going up to test it out :)
17:16:55 <mlavalle> I want to push a WIP patchset this week to start getting early feedback
17:17:39 <kiall> Excellent, hopefully that will help ensure we have everything needed in the specs!
17:18:04 <kiall> Anything else on Designate/Nova/Neutron specs?
17:18:11 <mlavalle> nope, done
17:18:16 <kiall> (as always, silence will mean agreement.. )
17:18:56 <kiall> Okay - Another one, is https://review.openstack.org/#/c/188018/ - Ceilometer / Metering integration. I'd like to get ekarlso started on this one next week.. (I don't think he's at the IRC meet, it's late for him)
17:19:05 <mugsie> I think we are ikn good shape - that outstanding issue on the Nova one is the last hurdle for implemention
17:19:37 <kiall> Any objections other than the couple of nits I added? If not, I'll fix the nits and ask Endre to get started next week...?
17:20:01 <mugsie> if i find anything I will add it in the review later on
17:20:25 <kiall> jmcbride - Just spotted your comment on it, the timers will be different for the different tasks, but I'll call that out explicitly
17:20:52 <timsim> I didn't see anything on my initial look, but I'll give it another look, I think it's fine.
17:21:30 <kiall> Great, I'll push a nit-free version in the next few hours..
17:22:07 <kiall> Last spec I know of with open issues and (possibly) something to discuss is https://review.openstack.org/#/c/188019/ - ALIAS records, and it's interaction with CDN's etc etc
17:22:29 <kiall> timsim / jmcbride - did you guys come up with any possible solutions?
17:22:42 <jmcbride> no, I think we still need some more time.
17:23:07 <timsim> No really good solutions :/
17:23:10 <kiall> jmcbride: sure.. will leave it on back-burner for now
17:23:43 <kiall> (we may have one, but I don't think I could coherently explain it right now)
17:24:11 <kiall> (But - It doesn't really fix the issue for bind, only prevents having an API that doesn't work for bind)
17:24:33 <jmcbride> we'd love to here your approach - but if you'd prefer to draft it when you are "coherent", that would be great.
17:25:17 <mugsie> jmcbride: that doesnt happen very often
17:25:29 <mugsie> more coherent is all we can hope for :)
17:25:47 <jmcbride> :)
17:26:40 <kiall> I haven't got it clear in my head (yet).. The gist of it is allowing backends to say 1) I support A, AAAA, but not ALIAS etc and 2) allowing backends to "override the implementation" of a given RRType. e.g. flattening would be enabled for bind, or ALIAS would be disabled. For PowerDNS, we'd AXFR the ALIAS right to the server and let it just do it, for Akamai style things - placement is limited, and setting an ALIAS would trigger an
17:26:41 <kiall> API call to "do the right thing" via their APIs
17:27:26 <kiall> But - As I said, it's A) still incoherent in my head.. and B) doesn't really solve it for bind, just leaves the door open for whatever solution is found
17:28:08 <kiall> (I can see confused faces in Austin right now)
17:28:30 <timsim> lol. It makes sense, but like you say, doesn't quite solve the issue for BIND
17:29:13 <jmcbride> This sounds similar to that pluggable rr type note you mentioned previously.
17:29:40 <mugsie> jmcbride: it is exactly that :)
17:30:13 <kiall> jmcbride: it is, the orignal pluggable rrtype idea was a tad more limited though.. this extends it from minor details like the list of supported RR types we allow, and how the render them, to "Oh, here's a whole bunch of code needed to make this RRType + Backend combination actually work"
17:31:45 <kiall> Anyway - For bind - I'm thinking the only real solution for ALIAS+CDN'd stuff is a patch along the lines of the bind GeoIP one - https://code.google.com/p/bind-geoip/
17:32:22 <kiall> They (eventually) took that in-tree, so something similar for ALIAS support might be accepted too..
17:33:14 <timsim> Yeah one of the things we wanted to do was talk to ISC and get their thoughts.
17:33:19 <jmcbride> great pointer, thanks
17:33:46 <kiall> Yea, that would be the ideal answer.. ISC probably know how to do it sanely in bind ;)
17:34:28 <kiall> #topic Open Discussion
17:34:37 <kiall> Since we're drifting in that direction anyway :)
17:34:54 <kiall> Anyone have any off-agenda items?
17:35:24 <timsim> I was able to reproduce the oslo config issue in stable/kilo just now
17:36:16 <kiall> timsim: cool, any ideas if the fix Heat used works for us?
17:36:35 <kiall> something along the lines of https://review.openstack.org/#/c/185287/ - somewhere..
17:37:17 <timsim> Looked at that briefly, didn't see anything similar in Designate :/
17:37:41 <timsim> I'm sure I just missed it though
17:37:55 <kiall> Heh, Okay.. Well - at least we know it's actually an issue :) Will try repro+fix when I get home
17:39:24 <kiall> Okay - Anything else before we call it a day?
17:39:39 <mugsie> nope, nothing from me
17:39:56 <timsim> Not from me
17:40:33 <kiall> Okay, Let's call it so.. Thanks all, looking forward to seeing WIP progress on Nova/Neutron and Ceilo next week :)
17:40:33 <kiall> #endmeeting