16:01:08 <markvoelker> #startmeeting defcore
16:01:10 <openstack> Meeting started Wed Sep 28 16:01:08 2016 UTC and is due to finish in 60 minutes.  The chair is markvoelker. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:14 <openstack> The meeting name has been set to 'defcore'
16:01:19 <markvoelker> #link https://etherpad.openstack.org/p/DefCoreLunar.19 Today's agenda
16:01:25 <Rockyg> o/
16:01:39 <catherine_d|1> o/
16:01:42 <markvoelker> #chair eglute_s
16:01:43 <openstack> Current chairs: eglute_s markvoelker
16:01:49 <eglute_s> o/
16:01:53 <markvoelker> #chair hogepodge
16:01:53 <openstack> Current chairs: eglute_s hogepodge markvoelker
16:02:02 <hogepodge> o/
16:02:17 <markvoelker> Hi everyone!  Let's dive in...
16:02:23 <markvoelker> #topic 2017.01 Guideline
16:02:41 <markvoelker> So hopefully everyone that's playing point for a particular project is readying scoring patches to submit
16:03:13 <markvoelker> I think I have the Glance and Neutron ones about ready to go and should get them up tomorrow
16:03:32 <markvoelker> Hopefully you've also all contacted the relevant PTL's to solicit their opinions
16:03:35 * eglute_s admits that markvoelker is way ahead of her
16:03:43 <luzC> o/
16:03:58 <markvoelker> If you need help, please let us know ASAP and we'll try to scramble you some cover
16:04:30 <markvoelker> Anyone have problems or updates they want to talk through now?
16:04:45 <eglute_s> Shamail told me he is not able to attend today's meeting, just fyi
16:04:49 <hogepodge> markvoelker: I probably need help, travel and illness have knocked me out flat
16:05:20 * eglute_s been sick the last week as well.
16:05:29 <markvoelker> hogepodge: Ok. Have you contacted the PTL yet?
16:05:47 <hogepodge> markvoelker: I haven't done anything. :-/
16:06:26 <markvoelker> hogepodge: Ok. I can probably throw in on that once I get the Neutron and Glance patches posted
16:07:04 <markvoelker> eglute_s: Are you good to go now, or do you need help too?
16:07:42 <eglute_s> I think i am good with swift, just waiting on notmyname, he said he has been traveling/busy. Have not looked at ceilometer
16:07:51 <notmyname> eglute_s: hi
16:07:59 <eglute_s> notmyname hi!
16:08:16 <notmyname> I looked. and then I realized it was more than a quick "check over this" thing ;-)
16:08:37 <eglute_s> notmyname that's why we reach out to PTLs :)
16:08:41 <eglute_s> you know your project best :)
16:08:46 <notmyname> so actually, I could use some guidance
16:09:01 <notmyname> (what's the link?)
16:09:34 <eglute_s> link to?
16:09:45 <eglute_s> #link https://github.com/openstack/defcore/blob/master/next.json
16:09:51 <notmyname> yeah, just found it :-)
16:09:52 <notmyname> https://github.com/openstack/defcore/blob/master/2016.08.json#L89
16:10:06 <notmyname> oh, mine is the current
16:10:08 <eglute_s> yes, thats the current one :)
16:10:19 <eglute_s> next is based on the current one
16:10:32 <eglute_s> thats the one we are looking to update
16:11:04 <notmyname> and you're looking to me to propose any changes? otherwise it stays the same?
16:11:13 <notmyname> well, except for advisory, right? those become require
16:11:15 <notmyname> d
16:11:25 <Rockyg> Any new features have to have been in since Liberty
16:11:46 <eglute_s> you dont have to propose changes directly to the next.json
16:11:47 <Rockyg> But also, new tests for existing capabilities
16:12:08 <eglute_s> just new capabilities we can score
16:12:21 <Rockyg> But, th test stuff comes after the scoring.
16:12:23 <notmyname> if the names are descriptive, the features under "advisory" have existed since circa 2011 (rough guess--point is, "a really long time")
16:12:23 <eglute_s> what Rockyg said. Tests must be in tempest though
16:12:44 <notmyname> yeah, I still don't like that ;-)
16:12:47 <eglute_s> well, the advisory status is recent
16:13:03 <notmyname> ok, so new capabilities. these are new user-visible features?
16:13:11 <eglute_s> notmyname i think we lost that battle for now
16:13:14 <eglute_s> correct
16:13:17 <eglute_s> non-admin
16:13:50 <notmyname> probably the biggest one I'd consider is something about storage policies, but I'd need some guidance there
16:14:06 <notmyname> you can set a storage policy on container create, but only if the admin has exposed that in the cluster
16:14:43 <notmyname> so it's a huge part of swift and a really useful feature, but not something that might be in all clusters
16:14:45 <eglute_s> does it have tests in tempest?
16:14:52 <notmyname> even if that cluster is running HEAD of master
16:15:29 <notmyname> not sure. for the reasons above, I'd guess no, but that's more easily solvable. what would you (defcore^Winterop) say about that sort of feature?
16:15:33 <eglute_s> if it is not widely used, then thats were scoring would come in. we score new capabilities on things like adoption
16:15:55 <eglute_s> i think it is worth scoring
16:16:08 <notmyname> how is "widely used" determined?
16:16:35 <hogepodge> notmyname: user surver sometimes
16:16:36 <notmyname> the 1k+ clusters I've got at swiftstack? ;-)
16:17:04 <notmyname> user survey participation is rather low. but that's a different discussion, probably :-)
16:17:11 <eglute_s> i think we usually look at the user survey...
16:17:13 <eglute_s> true!
16:17:36 <eglute_s> markvoelker other tips for determining adoption/usage?
16:17:48 <eglute_s> notmyname do you know if other clouds use this feature?
16:18:14 <notmyname> in general, I don't know. but don't take that as a comment on usage or not
16:18:28 <notmyname> ok. I'll look at what's available in tempest tests and leave a comment(?...where?)
16:18:32 <markvoelker> Product adoption (how many OpenStack products support it), client support, etc
16:18:46 <hogepodge> notmyname: we also rely on people telling us :-D
16:18:56 <markvoelker> Basically if you think it's reasonable to look at, suggest it and we'll take a look.
16:19:11 <eglute_s> notmyname ping me directly and i will submit a scoring patch
16:19:17 <eglute_s> unless you want to submit a scoring patch :)
16:19:29 <notmyname> yeah, I just don't want to say "you're not swift" if an op hasn't decided to use more than one policy in the cluster
16:19:39 <notmyname> eglute_s: ok, thanks. I'll ping you directly
16:19:43 <eglute_s> notmyname #link https://github.com/openstack/defcore/blob/master/working_materials/scoring.txt
16:19:45 <eglute_s> ok!
16:19:55 <notmyname> final question...
16:20:22 <notmyname> I think I'd prefer to stay out of required vs advisory, unless you need my commentary there. what are your thoughts?
16:20:34 <notmyname> (TBH, I'd call everything required)
16:21:10 <eglute_s> notmyname we start with advisory, then capability graduates to required next round
16:21:20 <eglute_s> basically, it has to be advisory before it becomes required
16:21:32 <notmyname> right, but that seems like a process for your team rather than something I directly comment on, right?
16:21:39 <eglute_s> correct
16:22:13 <notmyname> ok. so therefore I won't comment on required vs advisory and I'll only mention stuff that is in tempest tests that should be added to the list (which you can then put wherever is right)
16:22:17 <notmyname> sound good?
16:22:46 <eglute_s> new capabilities that will be added to the 2017.01 guideline will be added as advisory. for 2017.07 they will become required unless there are issues
16:23:11 <eglute_s> thanks notmyname! that very helpful
16:23:43 <markvoelker> Ok gang, any other scoring updates or problems to talk about today?
16:24:12 <markvoelker> Ok then, moving on...
16:24:18 <markvoelker> #topic Add neutron-lib to tc-approved-release (MERGED)
16:24:36 <markvoelker> #link https://review.openstack.org/#/c/371777/ Patch to make neutron-lib tc-approved-release
16:24:48 <markvoelker> #link http://eavesdrop.openstack.org/meetings/tc/2016/tc.2016-09-27-20.01.html TC meeting minutes where this was discussed
16:25:08 <markvoelker> Just a quick update on this since we talked about it last week: yesterday the TC merged that patch with unanimous vote
16:25:28 <markvoelker> This will allow us to consider adding neutron-lib as a designated section (or parts of it) in future Guidelines
16:25:51 <markvoelker> I'll save discussion on that for the upcoming scoring patch, but wanted to let folks know it landed
16:26:24 <markvoelker> FYI, I'm also looking at glance_store as a possibility for similar action (and someone--Doug I think?--brought that up in the TC meeting yesterday too by coincidence)
16:26:39 <markvoelker> IF there are similar library splits in your projects, you may want to think about that when you're doing scoring.
16:26:57 <markvoelker> Questions?
16:27:23 <markvoelker> Moving right along...
16:27:32 <markvoelker> #topic Clarifying that the TC only wants to consider changes to tc-approved-release that come from DefCore
16:27:44 <markvoelker> This was the other topic discussed yesterday, but hasn't landed yet.
16:28:09 <markvoelker> Basically this one says "if a project wants to be added to tc-approved-release, they talk to DefCore and DefCore proposes it to the TC"
16:28:55 <markvoelker> We discussed last week that we may want to draft up a little document that gives projects some guidance about how to go about doing that
16:29:11 <eglute_s> where there any concerns during yesterday's meeting?
16:29:22 * eglute_s quickly scans TC meeting log
16:29:26 <markvoelker> I've kind of backburnered that for the moment due to the need to get scoring done, but it might be a good thing to talk about it in Barcelona if I finish a draft before then
16:30:08 <markvoelker> eglute_s: Nothing major that I can recall...the gist was that the TC wanted to clarify the process, and I think that's probably quite reasonable.
16:30:59 <markvoelker> There was some back-and-forth about the DefCore->Interop Working Group name change, but that was fairly trivial. =)
16:31:06 <eglute_s> :0
16:31:08 <eglute_s> :)
16:31:33 <markvoelker> If anyone has concerns or comments, feel free to post those to gerrit since it hasn't landed yet.
16:31:44 <markvoelker> Anything further on this topic?
16:32:27 <eglute_s> I am good :)
16:32:37 <markvoelker> #topic Renames
16:33:07 <markvoelker> So Shamail isn't here today to talk about this, but just to highlight the to-do list that was drafted up
16:33:16 <markvoelker> #link https://etherpad.openstack.org/p/DefCore_Rename_Task_List Rename Task List
16:33:48 <markvoelker> We need to work on coordination of tasks so we don't have a lot of dangling bits.  IMHO that feels like a good Summit topic.
16:34:06 <eglute_s> agree
16:34:24 <markvoelker> If you can help with any of the tasks there, feel free to stick your name on the etherpad and ask questions.
16:35:10 <markvoelker> It's actually not a whole ton of legwork I think, particularly if it's well divvied up.  Just needs some coordination.
16:35:21 * notmyname raises hand for question
16:35:32 <markvoelker> notmyname: sure, go ahead
16:35:53 <notmyname> sorry, back to next guidelines, can wait if you need to continue current topic
16:36:12 <markvoelker> Actually I don't think there's much else to say on this since Shamail's not here, so go ahead. =)
16:36:28 <notmyname> heh, ok
16:36:40 <notmyname> so I'm looking at tempest and what's in liberty
16:36:50 <notmyname> that's what you said, right? tests that are in liberty?
16:37:17 <markvoelker> Well, the capabilities being tested need to be present in liberty
16:37:26 <notmyname> ah, ok
16:37:46 <notmyname> so liberty tempest doesn't matter. just tests exist now for features since liberty
16:38:13 <Rockyg> for the scoring, yeah.
16:38:14 <markvoelker> Right.  Most people run more modern versions of Tempest.  There's actually a tempest SHA in each guideline that was known to work at the time the Guideline was approved.
16:38:18 <notmyname> good. 'cause there's no reference to "liberty" tag or branches in tempest. most recent is grizzly, and after that just version numbers
16:38:31 <notmyname> ok, thanks
16:38:54 <markvoelker> Ok, moving on to summit stuff
16:39:01 <markvoelker> #topic Summit Planning
16:39:24 <markvoelker> It's almost time for Barcelona, and we've got one working session.  I've started a pad to solicit ideas for things we want to hash out
16:39:45 <markvoelker> We probably won't get to everything folks suggest due to the limited time, so try to focus on things that would benefit from facetime
16:39:59 <markvoelker> #link https://etherpad.openstack.org/p/DefCoreBarcelona Barcelona planning etherpad
16:40:03 <eglute_s> thanks markvoelker
16:40:29 <markvoelker> I put a few things in there to "seed the plot", but feel free to add more and we'll try to nail down an agenda after next week's meeting
16:41:03 <markvoelker> Oh, and as a reminder, our work session is:
16:41:12 <eglute_s> markvoelker what do you have in mind for adjusted guideline?
16:41:13 <markvoelker> #link https://www.openstack.org/summit/barcelona-2016/summit-schedule/events/16798/interop-defcore-working-group-work-session DefCore work session slot
16:42:20 <markvoelker> eglute_s: I don't think I have anything in mind yet.  But we do have a slightly abbreviated development cycle for Ocata, so it might be worth talking about whether we want to adjust the Guideline cadence accordingly
16:42:38 <markvoelker> Something to ponder and come to Barca with opinions/suggestions on I think. =)
16:42:45 <eglute_s> :)
16:42:51 <Rockyg> ++
16:43:03 <luzC> ++
16:43:07 <eglute_s> we already talked about 2017.01 being lighter than 2016.08
16:43:24 <eglute_s> are you thinking changing the schedule?
16:43:56 <markvoelker> Potentially, yes.  The development cycle has a slightly different schedule going forward, so it may make sense to adjust Guideline timing too.
16:44:14 <markvoelker> But like I said: it's an incomplete thought right now.
16:44:20 <Rockyg> I don't think the user survey timeframes will change, so we should consider what we trigger off of
16:44:40 <eglute_s> also, guideline schedule is in our process docs i believe
16:44:53 <eglute_s> i am ok changing it, just we need to keep that in mind
16:45:03 <Rockyg> Originally, DefCore was purposely not tid to relase schdule
16:45:16 <markvoelker> All good discussion points for Barcelona. =)
16:45:34 <Rockyg> Exactly.  What makes sense in light of....
16:46:17 <markvoelker> Ok, I think that's the last topic on the agenda today.
16:46:22 <markvoelker> #topic open discussion
16:46:30 <markvoelker> Anything else to bring up today folks?
16:46:40 <eglute_s> #link https://github.com/openstack/defcore/blob/master/doc/source/process/2016A.rst#expected-time-line
16:46:49 <eglute_s> for reading before Barcelona :)
16:47:52 <markvoelker> Going once...
16:48:05 <markvoelker> Going twice...
16:48:09 <eglute_s> thanks markvoelker! i am good :)
16:48:20 <markvoelker> Ok folks: get those scoring patches up!  See you next week.
16:48:24 <markvoelker> #endmeeting