16:00:52 <markvoelker> #startmeeting defcore
16:00:52 <openstack> Meeting started Wed Jul 20 16:00:52 2016 UTC and is due to finish in 60 minutes.  The chair is markvoelker. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:53 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:55 <openstack> The meeting name has been set to 'defcore'
16:01:06 <eglute> o/
16:01:12 <markvoelker> #chair eglute
16:01:13 <openstack> Current chairs: eglute markvoelker
16:01:24 <hogepodge> o/
16:01:25 <gema> o/
16:01:28 <markvoelker> #chair hogepodge
16:01:28 <openstack> Current chairs: eglute hogepodge markvoelker
16:01:33 <luzC> o/
16:01:43 <markvoelker> #link https://etherpad.openstack.org/p/DefCoreLunar.12 today's agenda
16:02:06 <Rockyg> o/
16:02:12 <markvoelker> Ok, let's jump in....
16:02:20 <markvoelker> #topic DefCore Summer 2016 Sprint
16:02:23 <hogepodge> I have Catherine down for some items, but she is out today
16:02:44 <markvoelker> #link https://etherpad.openstack.org/p/DefCoreSummer2016Sprint Midcycle planning pad
16:03:18 <markvoelker> I think we're mostly in good shape here.  The additional attributes patch landed.
16:03:30 <markvoelker> hogepodge mentioned yesterday that the Foundation folks may want to join us Wednesday
16:03:38 <hogepodge> That's the plan
16:04:00 <markvoelker> hogepodge I think you said they'd work the Foundation update into the sessions you're running Wednesday?
16:04:16 <hogepodge> yeah
16:04:30 <markvoelker> Ok, nifty...that makes scheduling easier. =)
16:04:59 <eglute> hogepodge can you add whoever will be attending to the list on etherpad? so I can submit the names
16:05:40 * markvoelker sees hogepodge doing that now
16:05:40 <eglute> thank you!
16:06:03 <markvoelker> Ok, anything else we need to discuss today about the midcycle?
16:06:49 <eglute> i will send out dinner options doodle
16:07:08 <markvoelker> #action eglute will send out doodle poll for dinner
16:07:14 <markvoelker> thanks eglute
16:07:19 <eglute> if there is a restaurant you would like to see on the list, let me know!
16:07:42 <markvoelker> Ok then, moving right along...
16:07:54 <markvoelker> #topic Test Spec Proposal
16:08:00 <markvoelker> #link https://review.openstack.org/#/c/317531/ Test Spec Proposal
16:08:34 <gema> no updates, I have to put all the points in a few slides with the open questions so that we can have a productive discussion at the midcycle
16:08:44 <markvoelker> Excellent, thanks gema
16:08:57 <markvoelker> #info Test Spec work session is currently scheduled for 9am Tuesday
16:09:02 <Rockyg> dang.  still haven't gotten to it.  kep getting dragged onto other not openstack stuff
16:09:03 <eglute> thank you gema
16:09:06 <gema> yw
16:09:13 <eglute> i was reading through it, looks really good
16:09:15 <gema> Rockyg: get going, I need your stuff in there also
16:09:43 <Rockyg> yup.  also gotta book trip.
16:09:58 <Rockyg> that's how bad it is.
16:10:11 <markvoelker> Ok, anything else on the test spec today?
16:10:31 <markvoelker> #topic Tests with admin credentials update
16:10:49 <markvoelker> #link https://review.openstack.org/#/c/299491/
16:11:25 <markvoelker> I finally got some feedback on the Tempest patch for removing unnecessary admin credentials this week
16:11:40 <markvoelker> Not terribly specific, but I'lll iterate on it again this week
16:11:53 <eglute> luzC says they still require admin?
16:12:43 <markvoelker> eglute: the snat tests do (which is like two tests I think)
16:12:48 <luzC> not all, but some of them with the patch will be moved to an admin folder
16:12:50 <markvoelker> Those genuinely need admin
16:13:31 <luzC> yes... the guideline will still need some adjustments
16:13:49 <markvoelker> So basically (assuming the tempest patch lands), https://review.openstack.org/#/c/299491/ will need a new patchset
16:14:04 <markvoelker> Which flags the snat tests in 2016.01 (informational since they're not required in that Guideline)
16:14:09 <markvoelker> And drops them from next
16:14:43 <markvoelker> Make sense?
16:14:49 <eglute> yup
16:14:59 <luzC> yes
16:15:25 <markvoelker> I'm happy to iterate on that, but don't want to steal Catherine's thunder and there's no particular rush just yet.=)
16:15:34 <eglute> :)
16:15:38 <markvoelker> In the meantime I'll focus on the tempest patch some more
16:15:55 <markvoelker> As an FYI, I'll be out of the office next week so I'll be making every effort to land a new patchset on that one this week
16:16:15 <markvoelker> Feel free to add yourself to the review so you can check it out once I get something posted
16:16:29 <eglute> thank you markvoelker
16:16:30 <markvoelker> (or critique the existing patch)
16:16:45 <markvoelker> Anything else on this today?
16:17:22 <markvoelker> Moving on then
16:17:34 <markvoelker> #topic Interop challenge posted by IBM at Austin Summit update
16:17:43 <markvoelker> Catherine's not here to speak to this, but I'll summarize
16:18:36 <markvoelker> #link https://etherpad.openstack.org/p/interop-challenge-kickoff-meeting
16:18:55 <markvoelker> Basically, a bunch of companies are going to collaborate on an interoperability challenge
16:19:23 <markvoelker> THe goal is to run some real workloads (in the form of Heat templates, terraform configs, ansible plays, etc) on several vendor products
16:19:31 <markvoelker> And be able to report on the findings in Barcelona
16:19:53 <markvoelker> THis is complimentary to the work we do on Guidelines in that we tend to focus on the API layer while this will focus on the workload mobility layer
16:20:16 <markvoelker> IF your company is interested, add your contacts to that etherpad
16:20:34 <markvoelker> They'll be using the DefCore mailing list (and/or IRC channel) to communicate, so expect to see some messages
16:20:50 <markvoelker> Messages to the ML will be tagged [interop-challenge]
16:20:54 <markvoelker> (hopefully =p)
16:21:33 <markvoelker> Right now they're looking for candidate workloads, so feel free to suggest some
16:21:50 <markvoelker> Ok, I think that's a decent summary...if you need more info, you might check out this slide deck:
16:22:03 <markvoelker> #link https://drive.google.com/open?id=0B9udXv_O5_TsaktrSm5uNlAtVFZZWTZkck8xZ0ItcEprMlBj Interop Challenge Intro Deck
16:22:22 <markvoelker> Any questions?  Or stuff I missed, for those of you that were there for the kickoff?
16:22:32 <eglute> they want code contributions!
16:23:05 <eglute> thanks for the summary markvoelker
16:23:30 <Rockyg> good summary, markvoelker
16:23:46 <markvoelker> OK then, moving right along...
16:23:54 <markvoelker> #topic Flag/drop test_create_update_port_with_second_ip
16:24:09 <markvoelker> #link https://review.openstack.org/#/c/343122/ Flag/drop test_create_update_port_with_second_ip
16:24:40 <markvoelker> I spent some time last week reviewing next.json and going over some results from my own labs and the community results in refstack to see how things were shaping upo
16:24:58 <markvoelker> I identified one test that looks like it's not actually a good candidate for us and put up a patch to remove it
16:25:28 <markvoelker> Basically the test tries to assign multiple fixed IP's to a single Neutron port...which not all backends support
16:25:46 <markvoelker> Some clouds also limit fixed IP's per port in neutron.conf to 1 for security reasons (e.g. spoofguard)
16:25:55 <eglute> thank you markvoelker, looks good to me. if there are no other comments, i think it is ready to merge.
16:26:48 <markvoelker> Thanks eglute.
16:27:09 <markvoelker> Just FYI, I'm still picking through a few others that seemed to cause problems for some clouds, but so far they look mostly incidental
16:27:48 <markvoelker> Any questions/comments on this one?
16:28:01 <eglute> hogepodge just merged it :)
16:28:19 <markvoelker> Ok then, that sounds like a clear signal to move to the next agenda item. =)  Thanks hogepodge
16:28:39 <markvoelker> #topic Image sharing tests
16:28:43 <markvoelker> #link https://review.openstack.org/#/c/338609/1
16:29:22 <markvoelker> Catherine's not here today, but the crux of the issue here is basically that we've tried to steer clear of tests that end users may have a hard time running themselves so they can verify functionality
16:29:37 <markvoelker> There are a couple of tests relating to image sharing that require two sets of credentials
16:30:12 <gema> this was also discussed on the test spec, that we could have a set of tests that are business , and a set of tests that are for end users
16:30:21 <gema> and if someone wants to get certified has to pass both
16:30:31 <gema> else users can verify the ones that do not require multitenant
16:30:38 <gema> or multiuser
16:30:54 <gema> economy tests and business tests
16:30:56 <markvoelker> gema: exactly what I was about to point to. =)  My thought was to take this up at the midcycle during the test spec discussion.
16:31:08 <gema> yep, sounds good, it's on my list of issues
16:31:16 <Rockyg> ++
16:31:43 <markvoelker> We have time yet to discuss and either remove them or put them back in advisory for one more cycle if we need more time to flesh that out
16:32:10 <markvoelker> There are some other capabilities that might fall into a similar bucket (volume transfer comes to mind) so I don't think it's an isolated case
16:32:19 <gema> yep
16:32:51 <markvoelker> OK, any further discussion right now on this issue?
16:33:37 <markvoelker> #topic Longevity of aliases
16:34:02 <markvoelker> #link https://review.openstack.org/#/c/341175/ Add alias for rebuild_deleted_server test
16:34:20 <markvoelker> Basically Catherine brought up the question of how long we should persist aliases during discussion of a patch
16:34:51 <markvoelker> The patch in question here adds an alias to next.json that we already had in the previous two guidelines
16:34:55 <markvoelker> (but never added to next)
16:35:24 <markvoelker> If you use a modern Tempest, you shouldn't actually need it...but if you're looking at older results that were run on an older tempest, you'll see RefStack indicating failures
16:35:46 <hogepodge> I'm fine with always having them around
16:35:49 <markvoelker> So I'm of a mind that the cost of keeping aliases around is fairly trivial, and it's useful.
16:35:56 <eglute> +1
16:36:47 <markvoelker> Seems like that's generally palletable to folks.  I'm going to try to scan through this week and find any other aliases we might be missing in next, but I think we've got the majority
16:36:55 <markvoelker> (based on looking over some existing community results)
16:37:09 <eglute> thanks markvoelker
16:37:31 <markvoelker> Anything further on this topic?  Ready to move on?
16:38:07 <markvoelker> #topic Session submissions for Barcelona
16:39:01 <markvoelker> Several of you submitted session ideas for Barcelona that are of interest to DefCore/interoperability.
16:39:02 <markvoelker> Thanks!
16:39:20 <markvoelker> I've added the list from last week into the etherpad, so keep an eye out for those when voting takes place
16:39:31 <markvoelker> Feel free to add any we're missing
16:39:58 * markvoelker glances at clock and wonders if we might actually get through the agenda this week.....!
16:40:06 <markvoelker> Anything else on summit talks?
16:40:26 <eglute> hogepodge any word on working group sessions?
16:40:59 <hogepodge> eglute: not yet, I can check in. we need to be careful about how we get them, since working groups are generally managed by the UC, but we're a board group
16:41:45 <markvoelker> thanks hogepodge
16:42:10 <eglute> ok, we just need some time, preferably at least 2 hours
16:42:17 <eglute> thanks hogepodge
16:42:53 <markvoelker> Ok, moving on
16:42:57 <markvoelker> #topic Get / tests
16:43:58 <markvoelker> Not much to report here...we landed one for Neutron and there's an in-tree functional test for glance
16:44:13 <markvoelker> Per the TC resolution we either need a new tempest test or to move the existing one
16:45:54 <markvoelker> I'll see if I can follow-up with the glance folks on that, but no real rush since that capability wasn't proposed for 2016.08 anyway
16:46:01 <eglute> thanks markvoelker
16:46:18 <markvoelker> Ok, last item for the day:
16:46:23 <markvoelker> #topic Rename working group
16:46:57 <markvoelker> A while ago I put up an RFC patch to illustrate the sorts of things we'll need to update if we change the name of the group per the Board's direction
16:47:05 <markvoelker> #link https://review.openstack.org/#/c/327086/ RFC patch on renaming
16:47:34 <markvoelker> It's been noted that we may want to avoid the term "Work Group" since WG's are generally the purview of the UC nowadays
16:48:32 <hogepodge> Work Committee may be a way to indicate that we're a board thing, although it's a bit convoluted of a name
16:48:40 <eglute> it is
16:49:08 <eglute> that was the point of renaming to WG, to indicate that we are not strictly board committee
16:49:28 <Rockyg> interop board?
16:49:45 <Rockyg> but that might suggest lcted, so not grat
16:49:47 <eglute> we could, of course, choose to leave it as is.
16:50:33 <eglute> interoperability committee?
16:50:49 <markvoelker> I'd suggest folks leave their ideas in the gerrit review so we can centralize the discussion.
16:50:57 <eglute> ok!
16:51:02 <markvoelker> We can also use some fudge time to talk about it in San Antonio
16:51:09 <eglute> +1
16:51:14 <luzC> +1
16:51:14 <Rockyg> over lunch?
16:51:19 <markvoelker> (and the Foundation folks will be there too, and may have some opinions to add)
16:51:25 <gema> yeah, like that's gonna cut it
16:51:26 <gema> x)
16:51:32 <gema> naming things is tough
16:52:25 * markvoelker would make the obligatory joke about the two hard problems in computer science again, but feels he's beaten that horse to death a few times now
16:52:37 <gema> :)
16:52:44 <gema> markvoelker: this discussion will be EXP
16:53:04 <eglute> :D
16:53:51 <markvoelker> Ok gang, thanks for taking time away from your quest to catch all the Pokemon today for the meeting. =) Anything else before we close?
16:54:41 <markvoelker> Ok, in that case...
16:54:43 <markvoelker> #endmeeting