16:01:02 <eglute> #startmeeting defcore
16:01:03 <openstack> Meeting started Wed Oct 19 16:01:02 2016 UTC and is due to finish in 60 minutes.  The chair is eglute. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:04 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:06 <openstack> The meeting name has been set to 'defcore'
16:01:13 <eglute> #topic agenda
16:01:24 <eglute> Hello Everyone! Please review agenda: #link https://etherpad.openstack.org/p/DefCoreLunar.21
16:01:29 <eglute> and update as needed
16:01:32 <shamail> Hi all
16:02:11 <eglute> Mark is unable to join today
16:02:15 <eglute> is hogepodge around?
16:02:20 <catherineD> o/
16:02:25 <gema> o/
16:02:34 <shamail> o/
16:02:55 <eglute> #topic guideline
16:03:21 <eglute> shamail, do you have an update on cinder?
16:04:05 <shamail> Yes, I plan to submit a patch later tonight.  I spoke with the PTL and we agreed that cinder-list-API-versions is probably worth scoring for this update.
16:04:22 <eglute> great, any other cinder changes that you see coming?
16:05:00 <shamail> The addition of V3 API does not require any updates yet since the tests are based on V2 and that API version will be around for a while (plus V3 is backwards compatible)
16:05:26 <eglute> ok... so that means we are not adding v3 then
16:05:32 <eglute> at least not yet
16:05:41 <shamail> Nothing else for this release since most of the other cinder capabilities require specific feature requirements or admin.
16:05:50 <eglute> thank you shamail!
16:05:58 <shamail> Yes, that is my recommendation for this update based on the conversation.
16:06:08 <shamail> You're welcome!  Sorry for being slow w/ the patch
16:06:11 <eglute> anyone have questions about cinder?
16:06:24 <hogepodge> o/
16:06:55 <eglute> glance is next on the list, Mark had signed up for that one, and i dont see a patch for it
16:07:01 <eglute> #chair hogepodge
16:07:01 <openstack> Current chairs: eglute hogepodge
16:07:16 <eglute> unless someone else has an update on glance?
16:07:25 <hogepodge> We shouldn't disallow they parts of v3 that are compatible with v2
16:07:36 <hogepodge> Sorry for being late
16:07:41 <eglute> hogepodge you mean for cinder?
16:07:55 <hogepodge> Yeah
16:08:12 <eglute> i dont think we are dissallowing, we are just not adding it yet. based on what shamail says it is not ready for scoring
16:08:35 <shamail> hogepodge: can you help me understand what that would mean?  We weren't planning on disallowing any parts.  The tests are still primarily V2 and the API is not going away soon so we weren't going to update to V3 for this guideline.
16:09:35 <hogepodge> But if a cloud is running v3 and passed testing I don't want to deny then there trademark because v2 is designated
16:09:50 <shamail> +1
16:10:08 <eglute> hogepodge how would you propose we handle that in the guideline?
16:10:21 <shamail> I have to double check the tests but do they support V3?
16:10:23 <eglute> add v3 with v2 tests?
16:10:35 <shamail> The feedback I got was V2 for now mainly
16:11:26 <hogepodge> If it's backwards compatible they should
16:11:51 <hogepodge> It's just the micro version problem again
16:11:52 * Rockyg sneaks in late
16:12:28 <pilgrimstack> HI from OVH
16:12:31 <eglute> should we discuss microversions during the summit? if it is not on the agenda yet?
16:12:39 <eglute> hello Rockyg and pilgrimstack!
16:13:25 <shamail> hogepodge: How should I handle the update based on your recommendation?
16:14:03 <hogepodge> I don't k ow
16:14:08 <hogepodge> Know
16:14:26 <catherineD> We can check again but I have not notice any volume tests with V3 ... so far only V1 and V2
16:14:55 <eglute> I added microversions to the summit agenda.
16:14:58 <shamail> catherineD: that's correct
16:15:09 <hogepodge> Update the designated section to say v3 compatible API is OK?
16:15:10 <shamail> I'll update as I proposed for now
16:15:18 <eglute> hogepodge i think that would work
16:15:29 <shamail> If we need to change after summit working session then we can make updates
16:15:37 <hogepodge> And if tests fail on v3 fix them to work on either
16:15:48 <eglute> shamail i think that will work
16:16:11 <eglute> hogepodge if someone comes to you and they run v3, we can def. expedite thigns
16:16:21 <eglute> whether it is flagging or adding new tests
16:18:07 <eglute> ok, anything else on cinder?
16:18:26 <eglute> since Mark is not here, does anyone else have update on glance?
16:19:01 <eglute> if not, Heat is next on the list
16:19:02 <eglute> catherineD, go ahead
16:19:06 <catherineD> right now in tempest config the default for cinnder v3 is false ... the defaults for v1 and v2 are true ... so for someone to purposely running v3 they need to turn this on
16:19:46 <catherineD> eglute: thx .... so we have a heat work session scheduled on Friday Oct 28
16:19:46 <eglute> catherineD good to know
16:19:54 <eglute> great, thank you
16:20:20 <eglute> everyone, if you can, please attend the session
16:20:22 <catherineD> the time an place a re in the etherpad ... please add that to your calendar ... we should be at the session
16:20:39 <catherineD> eglute: that is all for Heat
16:21:02 <eglute> thank you catherineD! I will send out calendar invite to that
16:21:03 <eglute> Keystone is next
16:21:07 <eglute> hogepodge do you have an update?
16:22:33 <hogepodge> No
16:22:51 <hogepodge> Mark had taken it over
16:23:05 <eglute> cool, was hoping you might have heard something :)
16:23:22 <eglute> no patch on keystone yet, I am sure mark will provide update later
16:23:38 <eglute> If there are no further comments on Keystone, Nova is next
16:23:54 <eglute> shamail, updates on nova?
16:24:05 <shamail> Hi eglute
16:24:08 <shamail> I submitted a patch
16:24:10 <eglute> #link https://review.openstack.org/#/c/385781/
16:24:12 <eglute> thank you!
16:24:47 <shamail> Mainly recommending suspend/resume and listing flavors
16:25:22 <eglute> i see there are 3 new capabilities
16:25:39 <eglute> I thought we already had flavors at some point
16:25:40 <shamail> list-api-versions was scored again… the results didnt change much (for adoption) but score went up slightly due to being sticky
16:26:17 <shamail> I checked the existing guideline and couldn’t see it.  Did it removed at some point?
16:26:32 <eglute> let me see
16:26:35 <Rockyg> that one might be a good one no matter the score.  Discovery is really important
16:27:24 <eglute> it was there in 2015.05 https://github.com/openstack/defcore/blob/master/2015.05.json#L125
16:28:22 <eglute> but it is not in the next one
16:28:24 <eglute> odd
16:28:29 <shamail> So list-flavors is the only non-admin operation re: flavors
16:28:34 <Rockyg> we need some way to let devs know about apis we think need tests or improvements to tests so they can be included, also
16:28:37 <shamail> was it removed because of admin for others?
16:29:18 <eglute> i cannot find it. it should be listed as removed in the next one, but it is not
16:29:25 <Rockyg> shouldn't be, but since it's removed, we should have more history on it
16:29:55 <eglute> what Rockyg said. I will dig after the meeting
16:30:07 <eglute> unless someone remembers
16:30:08 <shamail> Thanks
16:30:30 <shamail> I think its an important one to include unless we can recall the reason why it was removed
16:31:11 <Rockyg> ++
16:31:12 <eglute> shamail i agree. i am sure we can find the history, unless it was accidental omission
16:31:29 <eglute> anything else on nova?
16:31:30 <shamail> thanks
16:31:41 <shamail> Nope :)
16:32:22 <eglute> thank you shamail!
16:32:28 <eglute> Neutron is next!
16:32:47 <eglute> Please review: https://review.openstack.org/#/c/381393/
16:33:16 <eglute> mark added networks-list-api-versions and networks-subnet-pools-CRUD
16:33:40 <eglute> as well as neutron-lib to designated sections.
16:33:55 <eglute> if no comments on it today, i would like to merge it
16:35:20 <eglute> and i mean comments on the patch :)
16:35:24 <hogepodge> eglute: can you wait on it for just a day or so?
16:35:30 <eglute> yes of course
16:35:49 <hogepodge> I had someone ping me about neutron las week but I couldn't adrress their concerns. I just don't want to miss something. I'll catch up with them today
16:36:22 <eglute> hogepodge sure thing. i am just trying to move things along if there are no concerns, but if they are there, I will wait
16:37:02 <eglute> if no other feedback on neutron, we can move on to swift
16:37:16 <eglute> #link https://review.openstack.org/#/c/384660/
16:37:19 <eglute> huge patch there.
16:37:29 <eglute> notmyname provided some great feedback
16:37:44 <eglute> including re-organizing things, adding tests, and renaming capabilities
16:37:52 <eglute> i have a question on renaming capabilities
16:38:12 <eglute> right now i renamed them in the patch, but we have no way of tracking renames
16:38:21 <eglute> besides the tests staying the same
16:38:28 <eglute> or am i missing somehting
16:38:49 <Rockyg> I think you're right
16:39:15 <eglute> so should i hold off on renames?
16:39:21 <eglute> hogepodge what do you think
16:39:32 <Rockyg> so at a mimimum, we might want specific patches that are just the renames
16:39:52 <hogepodge> eglute: I need to take a close look at the patch
16:39:59 <eglute> ok
16:40:14 <hogepodge> eglute: I'd hold off for a bit, maybe until we can talk in person next week.
16:40:15 <Rockyg> that way, you can search for old name in reviews to find the new name
16:40:22 <eglute> i would appreciate everyone taking a look
16:40:31 <eglute> and i can wait until next week
16:40:44 <eglute> i scored several new capabilities for swift
16:40:55 <eglute> i would appreciate people looking at the scores as well
16:42:25 <eglute> additionally, lots of test changes for existing capabilities. Swift PTL provided those. We do not have a process for adding new tests, but i think it should be ok to add them, correct?
16:42:27 <notmyname> I just added myself as a reviewer on that patch. is there something you'd like me to specifically look at?
16:43:17 <eglute> notmyname please take a look at the scoring and for those that didnt score high enough to make the cut, you may want to argue that should be scored higher
16:43:32 <catherineD> eglute: we need to add adding new test process to DefCore ... I don't think it is very clear at the moment
16:43:33 <notmyname> heh. that would assume I've ever been able to understand the scoring ;-)
16:44:12 <eglute> notmyname I would be happy to work with you on that. it is extremely subjective :)
16:44:20 <notmyname> eglute: ok, thanks
16:44:22 <eglute> catherineD i agree...
16:44:22 <Rockyg> notmyname, convert it to a table and it's much easier
16:44:55 <notmyname> Rockyg: no, the problem is understanding "[1,1,1] [1,1,1] [1,1,0] [1,1,1] [1] [91]*" etc. I'll work with eglute on it
16:45:05 <eglute> notmyname: https://github.com/openstack/defcore/blob/master/working_materials/scoring.txt#L12
16:45:09 <eglute> see if that helps
16:45:16 <catherineD> I remember at the earlier days we discussed that the tests have to be in Tempest for 2 cycles for the test stability ... not sure whether that is still the criteria
16:45:56 <eglute> notmyname and here are the core criteria: https://github.com/openstack/defcore/blob/master/doc/source/process/CoreCriteria.rst
16:46:06 <Rockyg> right.  those used to all be part of a table (really a spreadsheet) and the columns with headings made *much* more sense
16:46:41 <eglute> catherineD i think you are right, but that would be hard to track for renames
16:47:19 <catherineD> we can use  the id to track rename tests ... if it is a new id it is a new test
16:48:06 <catherineD> currently, we use test id to determine whether they are aliases
16:48:07 <eglute> catherineD true
16:48:34 <eglute> catherineD would you help me define a process for adding new tests? i think we all agree that we need more tests for capabilities
16:49:26 <catherineD> eglute: sure we can work together on that ..
16:49:30 <Rockyg> let's work on that at the summit..can we add to etherpad?
16:49:41 <eglute> Rockyg yes, I will add it
16:49:43 <catherineD> Rockyg: +1
16:49:54 <eglute> catherineD thank you, I appreciate it!
16:49:56 <catherineD> more on the policy side
16:50:40 <gema> there is a test that needs an empty tenant to run
16:50:42 <notmyname> I think I checked that the tests were in liberty, but it would be good to have confirmation. thanks
16:51:04 <eglute> gema you mean for swift?
16:51:12 <gema> oh, no, sorry
16:51:18 <gema> I thought we were takling in general about tests
16:51:22 <gema> I will wait till the ned
16:51:23 <gema> end
16:51:41 <gema> (it is for compute)
16:52:08 <eglute> gema thanks for bringing it up
16:52:11 <catherineD> gema: I was talking about test in general  (policy ) :-)
16:52:22 <gema> catherineD: ack
16:52:35 <gema> eglute: tempest.api.compute.servers.test_list_servers_negative.ListServersNegativeTestJSON.test_list_servers_by_limits_greater_than_actual_count <- this tests
16:52:38 <gema> test
16:53:05 <eglute> so seems like that test should be removed?
16:53:07 <gema> it makes assumptions about how many servers the tenant has
16:53:10 <gema> I think so
16:53:14 <eglute> gema would you submit a patch?
16:53:21 <gema> eglute: will do
16:53:26 <eglute> thank you gema
16:53:43 <eglute> anything else on swift?
16:53:57 <eglute> I would appreciate everyone looking at it
16:54:27 <eglute> next, renames
16:54:33 <eglute> #topic Renames
16:54:34 <catherineD> eglute: will do
16:54:39 <eglute> shamail hogepodge any updates?
16:55:16 <hogepodge> Not here. I think shamail had sent some patches up
16:55:41 <shamail> Yes
16:55:51 <shamail> I have created a new room #openstack-interop
16:55:56 <eglute> shamail do you have links?
16:56:22 <shamail> I will need to work with one of the founders of our current channel to finish the rename process
16:56:29 <shamail> eglute: both patches merged
16:56:36 <eglute> ah ok
16:56:43 <shamail> #openstack-defcore no longer has bots, they have moved to #openstack-interop
16:56:43 <eglute> i think i created this channel
16:56:59 <shamail> Awesome
16:57:06 <shamail> I will work with you on the final item
16:57:07 <eglute> err, not this one, but the openstack-defcore
16:57:21 <shamail> I wanted to update the meeting entry in eavesdrop too
16:57:27 <eglute> thanks shamail
16:57:32 <shamail> but I realized that it points to our wiki page for the agenda
16:57:43 <eglute> right, we need to coordinate things
16:57:47 <hogepodge> everyone should join that channel now, since we're talking about it
16:57:50 <eglute> i am guessing after the summit
16:57:51 <shamail> No one has signed up to update the wiki
16:58:13 <shamail> yes eglute, that is probably the time frame for updating our wiki at this point
16:58:13 <eglute> i can update the wiki,
16:58:22 <shamail> I should be able to help with wiki move afterwards
16:58:22 <catherineD> so we should join #ropenstack-interop not #openstack-defcore ?
16:58:27 <shamail> yes
16:58:32 <eglute> wait wait
16:58:42 <eglute> lets not make the official switch yet
16:58:48 <shamail> eventually (with the steps I need help with), if someone enters openstack-defcore, they will be redirected automatically
16:58:54 <eglute> lets switch during/after summit
16:58:57 <hogepodge> ah, oh, don't join it. sry
16:59:09 <shamail> (http://docs.openstack.org/infra/system-config/irc.html)
16:59:15 <eglute> thanks shamail
16:59:15 <catherineD> I think we should do some announcement with date before we switch
16:59:25 <eglute> i think we want to switch all the things at once
16:59:31 <eglute> catherineD agreed
16:59:35 <shamail> Last step is “Renaming an IRC channel”, I will need to grant you op on the new channel eglute
16:59:43 <eglute> thanks shamail
16:59:49 <eglute> we are out of time...
17:00:00 <eglute> please add things to https://etherpad.openstack.org/p/DefCoreBarcelona
17:00:08 <eglute> and please review Refstack patch
17:00:19 <eglute> Please review https://review.openstack.org/#/c/382653/   we have users currently failed for this removed test.
17:00:27 <catherineD> that is a DefCore patch ...
17:01:01 <eglute> ah ok, it was listed after refstack and i didnt click it
17:01:03 <eglute> everyone review that patch :)
17:01:04 <catherineD> we need help to have it merged to skip a tempest removed test ... RefStack user are failing for that reason now
17:01:17 <catherineD> eglute: my mistake :-)
17:01:29 <eglute> and we are out of time... over in the #openstack-defcore channel to continue, i will be tehre
17:01:36 <eglute> thanks everyone!
17:01:38 <eglute> #endmeeting