16:01:48 <eglute> #startmeeting interopwg
16:01:49 <openstack> Meeting started Wed Nov 29 16:01:48 2017 UTC and is due to finish in 60 minutes.  The chair is eglute. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:50 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:52 <openstack> The meeting name has been set to 'interopwg'
16:01:55 <markvoelker> o/
16:02:00 <dims> o/
16:02:03 <eglute> #chair markvoelker hogepodge
16:02:04 <openstack> Current chairs: eglute hogepodge markvoelker
16:02:12 <eglute> #topic agenda
16:02:20 <eglute> Hello everyone!
16:02:22 <georgk> hi
16:02:30 <eglute> Here is today's agenda: #link https://etherpad.openstack.org/p/InteropVertigo.21
16:02:33 <hogepodge> hi
16:03:14 <mguiney> o/
16:03:46 <eglute> I will be traveling next week, Dec 6-13, so I won't be able to make next two meetings. I hope markvoelker and hogepodge will be able to cover, as it turns out i will be traveling during meeting time
16:04:01 <markvoelker> I should be here.
16:04:06 <eglute> thanks markvoelker
16:04:14 <eglute> #topic cycle name
16:04:15 <hogepodge> I'll be traveling next week, and unavailable on Wednesday.
16:04:32 <hogepodge> I'll also be on vacation for the remainder of the month of December, so again unavailable.
16:04:34 <eglute> Everyone, please enter your suggestion for new cycle name and vote for your top choice
16:04:43 <eglute> we will select the name at the end of this meeting
16:05:03 <eglute> #topic Vertical Program Update
16:05:18 <eglute> #link https://review.openstack.org/#/c/519732/
16:05:24 <eglute> not sure if georgk is around
16:05:31 <georgk> I am
16:05:32 <eglute> thank you georgk for all the patches
16:06:08 <eglute> they look good to me, but i am not NFV expert
16:06:15 <georgk> so, this patchset now includes initial scoring for the trunk port feature + the candidates from the etherpad
16:06:18 <eglute> so we really need everyone else to look at this as well
16:06:25 <eglute> thank you georgk
16:06:26 <georgk> I also added a few more candidates
16:06:28 <eglute> they look great
16:07:13 <eglute> if everyone else will review them, we can merge it soon
16:07:39 <georgk> yes, feedback is very welcome
16:07:59 <eglute> georgk you added some notes to etherpad- does that mean you will score additional capabilities?
16:08:24 <georgk> yes, I wanted to move on to other capabilities as well
16:08:48 <georgk> I looked at networking-bgp, but there are some issues with the tempest tests
16:09:18 <eglute> ok, does that mean you won't be adding those?
16:09:44 <georgk> question is if I should add this to the current patch or to a separate
16:10:04 <georgk> I prefer a separate one - just to keep things apart
16:10:16 <eglute> i think that will work
16:10:16 <georgk> or rather clearly separated
16:10:24 <eglute> do you need us to merge this one first?
16:10:32 <markvoelker> LEt's do separate.  Some of the others are going to be more controversial I think (like bgpvpn)
16:10:58 <georgk> well, please review it of course, I can create a dependent patch
16:11:17 <mguiney> o7
16:11:20 <eglute> sounds good georgk
16:11:24 <eglute> o8
16:11:56 <eglute> georgk also you have a comment about scenario tests
16:12:32 <georgk> yes. what is the policy here? do we typically only include API tests
16:12:51 <eglute> we dont have any capabilities right now that use scenario tests
16:13:08 <georgk> the background is that the API tests of the networkiing-bgp projects do not cover simple things like listing or showing stuff
16:13:21 <markvoelker> For the vertical programs I think some scenario tests may be appropriate.
16:13:22 <georgk> howwver, the scenario tests are more evolved
16:14:02 <eglute> i am ok with including scenario tests for vertical programs
16:14:05 <georgk> I reached out to the project to discuss if they can close the test gap
16:14:21 <georgk> no reply yet :-)
16:14:21 <markvoelker> That is to say: I would not exclude scenario tests from consideration.  The main difficulty is that they often test multiple capabilities, so you have to be a little careful...
16:14:38 <eglute> markvoelker that would be my main concern as well
16:15:03 <markvoelker> E.g. if it tests something that we're not proposing for the Guideline, passing the test means we've implicitly required more capabilities than the guideline lists.
16:15:24 <markvoelker> So, I'd say it's fine to propose them, we just need to do good due dilligence on them before approving.
16:15:50 <georgk> yes, because of these reasons I´d like to explain the background of the interop tests and check if they can implement separate tests for the simple capabilities
16:18:09 <markvoelker> That sounds fine to me.  Do you have specific tests you're considering now?  We can dig into those a bit to see if any are ok if so....
16:18:47 <georgk> markvoelker: scenario or api tests?
16:19:37 <markvoelker> Either that exist today. =)  E.g. if there are scenario tests today we can look at them and see if they're testing anything outside of what's be proposed for inclusion.
16:20:42 <georgk> ok, I´ll do another round of review of the test cases and then send out my results by email.
16:20:59 <georgk> is that ok?
16:21:12 <eglute> that would be great georgk
16:21:13 <eglute> thank you
16:21:18 <markvoelker> ++
16:21:30 <georgk> ok, will do
16:21:50 <eglute> thanks! anything else on this topic?
16:22:12 <markvoelker> In other related news, I've been picking at some stuff around EPA support and Aodh this week while tracking down some problems in a lab.  Should have a patch up soonish for those.
16:22:39 <eglute> tell us more markvoelker
16:22:49 <eglute> what is this for?
16:22:58 <markvoelker> For the NFV vertical program
16:23:19 <markvoelker> We'd cited those as candidates for the first release
16:23:44 <markvoelker> I'm actually poking at some neutron QoS things too but I'll defer those for a later release I think.
16:23:47 <kgarloff> So we'll have aodh covered in the NFV program?
16:24:09 <eglute> sounds good markvoelker thank you
16:24:17 <markvoelker> kgarloff: Well, at least we'll do a scoring exercise to see if anything merits inclusion.
16:24:26 <kgarloff> markvoelker: great
16:24:28 <georgk> one comment: in Denver, we said that the heat program is a candidate for a later point in time
16:24:58 <georgk> based on the fact that heat is used quite extensively in NFV, I was wondering if we should consider this also for inclusion
16:25:21 <kgarloff> georgk: heat was discussed to be an addon program for platform ... should it be in NFV-base?
16:26:01 <georgk> yes, like making the heat add-on also a component of the NFV program
16:26:21 <markvoelker> georgk: I think the main rationale for waiting on it was twofold: 1.) we were going to work on Heat as an add-on program first, and 2.) there's some limbo around tests for Heat right now while the TC sorts out opinions on where to put tests...they'd been asked int he past to move them into tempest, now that may not be as much a thing (or it yet might)
16:26:58 <markvoelker> So basically, I think the idea was to get an add-on program built for Heat and then make that part of the NFV vertical.
16:27:09 <georgk> markvoelker: ok, fine with me
16:27:21 <markvoelker> Nothing stopping us from doing them separately though, if we want...just seemed a bit duplicitous. =)
16:27:23 <kgarloff> We see significant heat usage on our public cloud (outside of NFV), so I agree that heat matters
16:28:08 <georgk> ok, so one thing at a time. sounds reasonable
16:29:01 <eglute> ok, anything else on this topic?
16:29:14 <eglute> #topic Microversions
16:29:18 <eglute> hogepodge any updates?
16:30:01 <hogepodge> eglute: none, this work is stalled out with me prepping for the summit and KubeCon
16:30:07 <eglute> cool, no worries
16:30:10 <eglute> #topic  2018.02 guideline
16:30:19 <eglute> markvoelker i think neutron is the only one missing
16:31:03 <markvoelker> Hah...got so wrapped up in the NFV stuff that I forgot to deal with that over Thanksgiving. =)  Ok, I'll add that to my queue for this week.
16:31:15 <markvoelker> I'm actually not sure I'll end up proposing anything though.
16:31:45 <eglute> markvoelker thank you! yeah, if you confirm no changes that works too
16:32:39 <markvoelker> (other than cleaning up the comments in the working doc about subnet pools since we actually added those)
16:32:39 <eglute> anything on the next guideline?
16:33:01 <eglute> cool
16:33:06 <eglute> there is a board meeting next week
16:33:22 <eglute> i will inform the board that the next guideline will be in february
16:33:47 <eglute> if there are any other updates i should announce, let me know
16:33:56 <eglute> #topic Add-on programs
16:34:47 <eglute> so there is a patch to add tc:approved release tag for designate https://review.openstack.org/#/c/521587/
16:35:01 <eglute> last meeting we ran out of time to fully discuss this
16:36:23 <mugsie> it is a simple step for the trademark program to actually add designate - the TC has to have blessed it
16:36:37 <mugsie> the other programs that were on the add on list were already in that tag afaik
16:37:24 * markvoelker is happy to see this and expects the measure to pass
16:37:31 <eglute> +1
16:37:48 <eglute> any other comments on the add on programs?
16:38:16 <eglute> there is also a testing patch:
16:38:19 <eglute> #link https://review.openstack.org/#/c/521602/
16:38:41 <eglute> thanks markvoelker for commenting on it
16:39:46 <eglute> that one seems to be more for TC to decide
16:40:07 <markvoelker> Others feel free to chip in too...the TC seems a bit divided on this at the moment, so your opinions and explanations of how things work in practice would be welcome I'm sure.
16:40:29 <eglute> ok i will respond later today, but i am in agreement with you
16:40:32 <mugsie> yeah, we had some discussion in the RefStack meeting yesterday as well
16:40:40 <eglute> i need to re-read all the comments again
16:40:59 <mugsie> my problem is some programs relying on a different interface, and the potential for breakage
16:41:09 <eglute> mugsie what was the consensus in the refstack meeting?
16:41:15 <mugsie> along with the need for increased review quality
16:41:20 <mugsie> eglute: none :(
16:41:59 <markvoelker> #link http://eavesdrop.openstack.org/meetings/refstack/2017/refstack.2017-11-28-17.02.html Refstack meeting logs
16:42:15 <eglute> ok... I will respond to that patch. thank you markvoelker
16:42:15 <tosky> during the meeting I pointed out that the breakage that happened in the past (which was reverted) was more an exception
16:42:57 <eglute> ok
16:43:57 <eglute> as for the proposed patch, i am ok with #2 or #3, I always liked #3
16:44:40 <eglute> everyone, please review and comment #link https://review.openstack.org/#/c/521602/
16:44:48 <eglute> any other discussion about it here?
16:45:35 <mugsie> eglute: what would your pushback for #1 be?
16:45:42 <markvoelker> I am a little interested in how the mechanics would play out for keeping things in a central repo, but I can comment on that once I get caught up with all the comments from while I was away
16:48:21 <eglute> mugsie i think having only option #1 is limiting. I think we should be a bit more flexible when it comes to test location. If a project has all its tests in a separate plugin already, making them move seems to be busy work
16:48:50 <eglute> but i need to catch up with all the comments as well
16:48:59 <mugsie> cool - I have a feeling this one will drag out a bit
16:50:00 <eglute> thank you for proposing the patch, I think it is great discussion and was much needed
16:50:01 <markvoelker> There have also been some concerns raised in the past about whether things like add-on programs were really "in scope" for tempest and whether QA would be able to handle all these tests...I think Andrea voiced those concerns in the review IIRC.
16:50:36 <tosky> that's the reason why tests for non-"core" components moved outside tempest.git
16:50:49 <mugsie> markvoelker: "in scope" for tempest is something that can be changed
16:51:10 <markvoelker> Sure, but reviewer bandwidth is more difficult to change. =)
16:51:22 <mugsie> it is 6 tests from us, I am not sure how many tests heat have but I cannot imagine it is huge
16:51:42 <mugsie> but, the discussion should be on the patch, so we have it recorded :)
16:51:57 <markvoelker> ++
16:52:01 <eglute> ++
16:52:10 <eglute> everyone, please comment on the patch
16:52:38 <eglute> if nothing else on this, we can return to the hardest problem of all
16:52:40 <eglute> naming
16:52:44 <eglute> #cycle name
16:52:53 <eglute> #topic cycle name
16:53:22 <hogepodge> The one hardest problem, naming things and off-by-one errors
16:53:46 <eglute> reminder, this is just for interop cycle name and really only affects the name of ehterpad mostly :)
16:53:51 <eglute> we have 5 suggestions right now
16:53:55 <eglute> https://etherpad.openstack.org/p/InteropVertigo.21
16:54:02 <eglute> please everyone add your votes
16:54:44 <eglute> right now Whistler is winning by 50%
16:54:58 <georgk> :-)
16:55:21 <eglute> ok, if no other votes, whistler it is!
16:55:46 <eglute> #topic open agenda
16:55:56 <eglute> we still have 5 minutes, anything else that we need to discuss?
16:56:11 <markvoelker> Quick note on PTG for those who didn't see it: https://twitter.com/tcarrez/status/935798883468431362
16:56:41 <eglute> is that a soccer stadium?
16:56:55 <kgarloff> Whaoh!
16:57:09 <markvoelker> More rugby nowadays, I think?  But yeah.
16:57:11 <hogepodge> Please let me know if you plan on attending.
16:57:13 <mugsie> eglute: I wouldn't call it a soccer stadium when you are here ;)
16:57:30 <eglute> mugsie yes, i grew up playing soccer
16:57:34 <mugsie> It is a gealic football stadium
16:57:35 <hogepodge> I have to respond to the PTG survey requesting resources. General head count and time needed.
16:57:45 <eglute> mugsie i mean football
16:57:49 <markvoelker> hogepodge: I'm planning to be there.
16:57:57 <kgarloff> Nice for us Europeans
16:58:05 <mugsie> It is a cool venue
16:58:10 <markvoelker> And will be happy to use a friendly game of football or rugby to settle tie votes...
16:58:13 <eglute> i think i will too.
16:58:38 <eglute> thanks everyone!
16:58:41 <eglute> #endmeeting