17:00:19 <timsim> #startmeeting designate
17:00:20 <openstack> Meeting started Wed May 31 17:00:19 2017 UTC and is due to finish in 60 minutes.  The chair is timsim. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:24 <openstack> The meeting name has been set to 'designate'
17:00:24 <timsim> #topic Roll Call
17:00:30 <hieulq_> o/
17:00:31 <daidv> o/
17:00:34 <trungnv_> o/
17:00:34 <sonuk> o/
17:00:42 <timsim> Don't see any mugsie nicks in here
17:02:09 <timsim> #topic Bug Triage
17:02:42 <mugsie> o/
17:02:53 <hieulq_> hi mugsie
17:03:05 <mugsie> I am double booked today
17:03:12 <mugsie> so I will be sporadic
17:03:26 <timsim> https://bugs.launchpad.net/designate/+bug/1688621
17:03:27 <openstack> Launchpad bug 1688621 in Designate "Ability to have different Rabbit connections for Sink" [Undecided,New]
17:03:50 <timsim> Not really a bug, more a feature request
17:03:58 <mugsie> yeah
17:04:17 <timsim> I'll tell them to bp/spec if they're interested
17:04:39 <timsim> https://bugs.launchpad.net/designate/+bug/1690184
17:04:40 <openstack> Launchpad bug 1690184 in Designate "zones created via zone import don't have pool attributes" [Undecided,New]
17:04:42 <timsim> This is that terrible one
17:06:17 <mugsie> ya
17:06:36 <timsim> Critical I guess
17:07:03 <mugsie> definitly
17:07:11 <timsim> Don't think we have milestones...
17:07:13 <timsim> https://bugs.launchpad.net/designate/+bug/1691618
17:07:14 <openstack> Launchpad bug 1691618 in Designate "MQ messages flood in designate devstack master" [Undecided,New]
17:07:49 <timsim> No clue what's happening there
17:08:06 <timsim> I think I have seen more random oslo.messaging stuff over the last few months in the logs
17:08:31 <mugsie> yeah - someone needs to dig into it
17:08:31 <timsim> If the gate is still passing it's probably not a high priority. Medium?
17:08:34 <mugsie> med
17:08:46 <hieulq__> I will figure out after finishing the works in enable worker model
17:09:15 <mugsie> cool - thanks
17:10:22 <timsim> #topic Stable Backport Triage
17:10:27 <timsim> Been a while since we've done this http://paste.openstack.org/show/611122/
17:10:58 <timsim> Anyone see anything worth of backporting to stable branches?
17:11:32 <mugsie> nothing really
17:11:57 <timsim> yeah me either
17:12:07 <timsim> #topic Open Discussion
17:12:14 <timsim> Anyone want to discuss anything?
17:12:16 <trungnv_> Hi
17:12:22 <trungnv_> yes. I have
17:12:22 <hieulq__> one question to mugsie
17:13:06 <hieulq__> for enabling the worker model, is it possible to change grenade to realize our new config?
17:13:14 <mugsie> yes
17:13:21 <mugsie> I am not 100% what it is
17:13:24 <mugsie> but we can
17:13:38 <hieulq__> you recommend to create new grenade gate for worker model
17:13:57 <hieulq__> but I prefer configuring grenade to realize the new config
17:14:22 <hieulq__> so maybe we do not need new grenade gate for worker model only
17:14:28 <hieulq__> am I right?
17:15:57 <mugsie> I would prefer it
17:16:03 <mugsie> but I am OK now not having one
17:16:19 <hieulq__> thanks
17:16:32 <trungnv_> we are almost done with OVO (just zone object). please look at our PS https://review.openstack.org/#/c/464971/
17:16:50 <hieulq__> that's all from me
17:17:21 <trungnv_> we only issue with a functional testcase to pass gate.
17:17:47 <timsim> Did you all have any luck talking to your leaders about the proxy approach we discussed in Boston?
17:18:10 <hieulq__> no luck at all
17:18:42 <hieulq__> they still request us to implement ovo and other stuffs
17:18:52 <trungnv_> yep.
17:19:10 <timsim> How do you feel about that?
17:19:47 <hieulq__> not really comfortable
17:20:10 <hieulq__> but at least we should try in this cycle
17:20:21 <timsim> Well, alright then.
17:20:27 <timsim> Anything else from anyone?
17:20:46 <hieulq__> we are running ice house, so put a proxy into the ancient system could be a doom for us
17:21:52 <hieulq__> I think that's all from Fujitsu guys
17:22:21 <daidv> timsim, https://github.com/openstack/designate/blob/master/designate/objects/recordset.py#L142
17:22:30 <daidv> For this #TODO
17:22:49 <daidv> I think we can have easy implemention via OVO
17:23:00 <daidv> like this https://review.openstack.org/#/c/464971/13/designate/objects/fields.py
17:23:18 <daidv> with a PolymorphicObject.
17:23:41 <timsim> hm.
17:24:51 <timsim> My problem with our objects has always been: too complicated. It's very difficult to reason about when you get into it and try to actually use them. My fear with OVO, among other things, is that this is going to be even MORE complicated.
17:25:22 <timsim> For example, if by some miracle the functional tests are able to pass, I'll have almost no confidence that they're going to work going forward.
17:25:47 <timsim> Because it's going to take me 5 days of code review/messing around to even partly understand what's going on.
17:26:19 <timsim> I understand the approach of just trying to fit them in where the current objects work. But to get it into a mergeable state in the future, I fear you're going to have to rewrite how the objects work altogether.
17:26:49 <timsim> Which is what Graham mentioned originally. "Make these current objects better at what they do now" rather than "shoehorn ovo in here"
17:26:54 <timsim> Anyway.
17:26:59 <timsim> That's my opinion.
17:27:12 <timsim> Anything else?
17:28:20 <timsim> Alright, going once,
17:28:28 <timsim> twice
17:28:29 <trungnv_> yes. i have
17:28:33 <timsim> aha
17:28:38 <trungnv_> timsim, I have some PSs https://review.openstack.org/#/c/468088/ and https://review.openstack.org/#/c/468904/
17:28:58 <trungnv_> could you look at them?
17:29:41 <timsim> sure
17:30:03 <trungnv_> with OVO, we can try to get pass everything then we will improve whole code to get best solution for our codes.
17:30:23 <trungnv_> we also will test our code on our system.
17:30:48 <timsim> For https://review.openstack.org/#/c/468088/ I think a unit test is needed, +2'd the other one
17:32:27 <trungnv_> which remain PS?
17:34:07 <daidv> timsim, hmm, thanks. we will do both of them (improve testcase to ensure about some validation and try to migrate current objects to OVO)
17:34:23 <timsim> Cool
17:34:31 <timsim> Alright, I need to head out, anything else?
17:34:45 <trungnv_> that all for me.
17:34:48 <trungnv_> thanks
17:35:02 <daidv> me too, thanks.
17:35:42 <timsim> cool. Thanks all
17:35:44 <timsim> #endmeeting