20:01:42 <ttx> #startmeeting tc
20:01:43 <openstack> Meeting started Tue Sep  3 20:01:42 2013 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:01:45 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:01:47 <openstack> The meeting name has been set to 'tc'
20:01:54 <ttx> Our agenda:
20:01:59 <ttx> #link https://wiki.openstack.org/wiki/Governance/TechnicalCommittee
20:02:09 <ttx> I suspect we'll probably to next week
20:02:17 <ttx> err
20:02:18 <jd__> o/
20:02:21 <ttx> s/probably/overflow/
20:02:25 <hub_cap> :)
20:02:37 <ttx> #topic Marconi incubation request: final discussion
20:02:42 <flaper87> \o/
20:02:48 <ttx> #link http://lists.openstack.org/pipermail/openstack-dev/2013-August/014076.html
20:02:52 <markmcclain> o/
20:02:57 <ttx> #link https://wiki.openstack.org/wiki/Marconi/Incubation
20:02:58 <russellb> oh, i'm proxying for markmc
20:03:09 <russellb> btw.
20:03:11 <ttx> russellb: 2 votes! power to you !
20:03:15 <russellb> yeah!
20:03:24 <ttx> I think last week we covered Marconi scope questions, like:
20:03:33 <ttx> - Marconi is a queuing service itself, not a provisioning service for some other queue software
20:03:45 <ttx> - Marconi shall provide alternatives to MongoDB on the storage backend, but ideally keep the user-facing API(s) consistent
20:03:52 <flaper87> We put together a Q&A section in MArconi's incubation page: https://wiki.openstack.org/wiki/Marconi/Incubation#Raised_Questions_.2B_Answers
20:04:04 <ttx> flaper87: just read that, it's pretty nice
20:04:10 <ttx> I wanted us to discuss integration points with other projects.
20:04:16 <flaper87> +1
20:04:21 <ttx> Obviously you will integrate with Keystone, I suspect you'll appear in Horizon at some point too...
20:04:22 <kgriffs> sure thing
20:04:34 <flaper87> ttx: we already support keystone
20:04:36 <ttx> Your wiki page mentions running on Nova servers, behind "OpenStack load balancers" -- Neutron LBaaS stuff, or somethign else ?
20:04:57 <kgriffs> re Horizon, I could see it being helpful to see queue stats, create/delete queues, a few management things like that, so
20:05:14 <kgriffs> ttx: LBaaS
20:05:29 <ttx> kgriffs: What about using Heat vs. using Nova directly ?
20:05:40 <russellb> and it doesn't use nova directly does it?
20:05:49 <russellb> are you saying that it's possible to deploy on top of nova?
20:05:53 <russellb> as opposed to specific integration?
20:06:13 <kgriffs> sorry, that blurb may be a bit misleading
20:06:38 <russellb> just like you could run glance on nova, or whatever
20:06:54 <kgriffs> while we expect that the majority of deployments will utilize LBaaS and IaaS marconi's design does not preclude running elsewhere
20:07:03 <flaper87> yeah, I'd say so, yes!
20:07:12 <flaper87> what kgriffs said!
20:07:21 <russellb> ok.
20:07:28 <russellb> so, it's not specific integration
20:07:39 <ttx> can't wait until Ironic makes all that even more confusing
20:07:39 <russellb> just that you expect that it will be common to deploy that way
20:07:45 <kgriffs> jright
20:07:47 <flaper87> As per Ceilo, we're planning to integrate w/ it for states and billing. I started digging into how it could be done
20:07:47 <kgriffs> right
20:07:49 <russellb> cool
20:07:55 <annegentle> kgriffs: has marconi been tested with neutron's lb?
20:07:56 <kgriffs> ideally you would also hook into heat/autoscale
20:07:57 <flaper87> stats*
20:08:08 <flaper87> annegentle: it hasn't
20:08:09 <kgriffs> but thats an operation thing and not tightly coupled to marconi
20:08:17 <russellb> yeah, so maybe there's some integration work with the deployment program (tripleo)
20:08:17 <notmyname> /late
20:08:23 <russellb> some integration that could be done, i mean.
20:08:29 <flaper87> russellb: totally
20:08:35 <russellb> as in, help develop the heat templates to get this stuff deployed
20:08:39 <ttx> notmyname: we were just slowly getting more excited.
20:08:52 <notmyname> :-)
20:08:59 <notmyname> just caught up with the buffer playback
20:09:00 <ttx> Sidenote: if accepted, Queue service will become a new "program" with the Marconi project itself being in incubation for integration in the common release
20:09:09 <ttx> The "mission statement" for the program is mentioned on the wiki page linked above
20:09:17 <flaper87> ttx: +1
20:09:27 <ttx> you might want to review it as part of your decision -making process
20:09:37 <jeblair> hi. i'd like to encourage marconi to target integration with devstack, tempest, and the devstack-gate tests early; and i'd like to encourage the tc to view that as a goal for promotion from incubation.
20:10:00 <russellb> jeblair: +1
20:10:01 <ttx> jeblair: yes, we intend to be a lot more blocking on that
20:10:05 <flaper87> FWIW, the integration with other services and tools like devstack is part of our roadmap for our incubation process
20:10:17 <flaper87> meaning, we want to get all that done before graduating
20:10:23 <ttx> jeblair: especially with projects that will be aware of that condition from incubation day 0
20:10:23 <kgriffs> +1
20:10:42 <flaper87> (at least we discussed that, FWIW)
20:10:57 <ttx> More questions on Marconi before we vote ?
20:10:58 <jeblair> cool; it's pretty easy to run alternate devstack tests/environments now, so a project can start running those kinds of tests before it joins the shared devstack-gate queue with the rest of the common release projects
20:11:00 <annegentle> kgriffs: so this is where I get confused. As an admin user, I could go to Horizon to see my OpenStack deployment's queue?
20:11:04 <flaper87> jeblair: +1
20:11:30 <kgriffs> russellb: re heat templates, I think it only makes sense for the Marconi team to contribute there
20:11:30 <russellb> annegentle: nah, this is different from the queues used as a part of the infrastructure
20:11:32 <annegentle> kgriffs: as opposed to some cool stock quote queue I made up as a web dev?
20:11:42 <flaper87> annegentle: yup
20:11:46 <annegentle> russellb: ok that's what I thought, we'd really need to map out the Horizon integration then
20:11:52 <ttx> Personally I consider a queue service to be pretty essential in a IaaS+ setting
20:12:01 <ttx> I'd like the focus to be first on providing a featureful and performant solution...
20:12:05 <flaper87> ttx: Agreed
20:12:06 <ttx> ...rather than on supporting multiple "transports" and "backends".
20:12:11 <annegentle> flaper87: I think the answer is "no?" no?
20:12:13 <annegentle> flaper87: :)
20:12:15 <ttx> With that caveat noted, I'm ready to vote
20:12:19 <oz_akan_> ttx: +1
20:12:21 <russellb> ready here too
20:12:22 <flaper87> ttx: that's our goal
20:12:40 <flaper87> annegentle: right
20:12:42 <flaper87> :D
20:12:45 <annegentle> flaper87: ok, whew
20:12:51 <annegentle> ttx: ok good to go too
20:12:59 * flaper87 bites his nails
20:13:06 <mikal> Heh
20:13:10 <ttx> flaper87: note that it's everyoen goal, and then vendors start contributing weird stuff
20:13:18 <russellb> heh
20:13:27 <kgriffs> quick question
20:13:27 <russellb> yeah, don't want to get distracted by too much of that ...
20:13:31 <jgriffith> ttx: +1
20:13:32 <ttx> (no, i won't give examples :)
20:13:32 <annegentle> ttx: you're too funny
20:13:51 <flaper87> ttx: lol
20:13:55 <kgriffs> would you see an alternative backend (e.g., MySQL) as a requirement for successful incubation?
20:14:23 <notmyname> IMO, no
20:14:24 <annegentle> kgriffs: is successful incubation graduation to integrated?
20:14:25 <russellb> i'd like to say no ... but the licensing of mongodb probably means we should consider it
20:14:31 <ttx> kgriffs: I think some people see a non-AGPL deployemnt option as a key requirement
20:14:31 <kgriffs> yes, graduation
20:14:36 <russellb> because i think a lot of people are going to hate that
20:14:40 <russellb> for graduation
20:14:45 <kgriffs> ok
20:14:49 <flaper87> ok
20:14:52 <ttx> kgriffs: but personally I would not block on that
20:15:02 <russellb> i would be tempted to personally ...
20:15:05 <flaper87> awesome feedback, thanks
20:15:05 <zaneb> russellb: didn't we burn that bridge with ceilometer? ;)
20:15:07 <russellb> well, maybe i wouldn't block
20:15:09 <russellb> heh
20:15:12 <kgriffs> thanks for the clarification, makes sense
20:15:22 <russellb> i mean if people are upset enough, they should help add what they want, right?  :-)
20:15:29 <russellb> but i expect it to be an "issue" with some folks.
20:15:40 <gabrielhurley> russellb: seconded
20:15:50 <ttx> kgriffs: so maybe I wouldn't require for graduation, but would be a good feature to add before the first integrated release
20:15:50 <russellb> cool
20:16:04 <ttx> (i.e. during the J cycle once you graduated)
20:16:07 <kgriffs> ttx: cool beans
20:16:15 <russellb> that's like *forever* from now right?
20:16:21 <kgriffs> heh. :D
20:16:22 <ttx> more questions ? or all TC members ready to vote ?
20:16:28 <dolphm> ttx: i thought there was a licensing issue? no?
20:16:32 <mikal> I'm ready
20:16:38 <russellb> ready
20:16:47 <annegentle> what's ceilometer using now?
20:16:59 <zaneb> dolphm: no, the client is permissively licensed
20:17:00 <ttx> dolphm: no. There is a licensing issue with one deployment option of marconi
20:17:06 <vishy> o/
20:17:17 <russellb> vishy: just in time for the best part!
20:17:26 <mordred> o/
20:17:30 <jd__> annegentle: mongodb or sql
20:17:40 * mordred scans scrollback real quick
20:17:41 <ttx> vishy, mordred: you might want to file questions before we vote ?
20:17:51 <annegentle> jd__: thanks
20:17:57 <mordred> <--- "some people see a non-AGPL deployemnt option as a key requirement" +100
20:17:58 <vishy> sorry i'm late, just got through airport security
20:18:05 <mordred> vishy: I'm in an airport too! :)
20:18:19 <ttx> mordred: I was trying to proxy you informally
20:18:30 <mordred> I would say non-agpl deployment option should be a requirement for graduation
20:18:35 <mordred> ceilometer has one
20:18:47 <flaper87> ok
20:18:58 <russellb> definitely more important than a new transport IMO
20:19:00 <ttx> mordred: they can use sqlite, does it count ?
20:19:03 <notmyname> licensing is a separate issue than "alternate backends", IMO.
20:19:05 <ttx> :P
20:19:09 <flaper87> ttx: lol
20:19:10 <gabrielhurley> sqlite--
20:19:20 <notmyname> back to the first question of it being a requirement for graduation
20:19:23 <russellb> json blob text files
20:19:28 <mikal> Well, at least one distro uses sqlite as their default
20:19:32 <mikal> For nova glance etc
20:19:39 <gabrielhurley> mikal: I don't consider that to be any better
20:19:44 <russellb> mikal: and that's just bonkers
20:19:47 <gabrielhurley> distro--
20:19:55 <mikal> I didn't say it was a good idea, just observing...
20:19:55 <dolphm> also worth reiterating that graduation is also decided by the future TC, not us
20:20:02 * notmyname likes sqlite
20:20:03 * mordred just doesn't want to see 'install mongodb' be a requirement for anyone wanting to run it 'for real'
20:20:08 * mordred likes sqlite too
20:20:11 <ttx> dolphm: apply for reelection!
20:20:19 <dolphm> ttx: just sayin'
20:20:31 <kgriffs> mordred: makes total sense to me
20:20:32 <ttx> mordred, vishy: you ready to vote ?
20:20:34 * mordred does not know enough about how it uses sqlite to know if it's sensible like swift, or whether it's a silly choice like a nova + sqlite would be
20:20:41 <notmyname> ...just a funny discussion when a typical swift deployment can have millions of sqlite DBs
20:20:47 <dolphm> we had a conversation previously about setting terms for graduation on a project going into incubation
20:20:54 <kgriffs> mordred: sqlite is just for localhost dev/test environments
20:21:02 <ttx> notmyname: you're using it in a unique way arguably :)
20:21:03 <mordred> notmyname: I think your architecture makes sense with sqlite - you don't have a central sqlite
20:21:03 <kgriffs> (not production!)
20:21:14 <mordred> kgriffs: right. so I think that for me it's a graduation requirement
20:21:19 <mordred> and I say that
20:21:31 <gabrielhurley> I would be +1 for calling it a graduation requirement, fwiw
20:21:33 <mordred> just because I think we shoudl be explicit about things we expect you to work on in incubation
20:21:36 <lifeless> oh look, a meeting.
20:21:40 <flaper87> FWIW, we were already planning to implement a rel-based backend, it was a matter to know when
20:21:47 <kgriffs> I'm cool with that
20:21:50 <mordred> cool
20:22:01 <flaper87> so, having it for graduation makes that decision easier
20:22:03 <flaper87> :D
20:22:10 <notmyname> mordred: gabrielhurley: "it" being "an alternate storage backend" or "it" being "something not agpl licensed"?
20:22:11 <ttx> gabrielhurley: having graduation requirements is all about learning from our past mistakes after all
20:22:20 <kgriffs> we sort of expected folks would want a couple deployment options, in any case
20:22:24 <mordred> notmyname: "something not agpl licensed"
20:22:32 <dolphm> notmyname: +1 for simply "not agpl licensed"
20:22:38 <gabrielhurley> notmyname: I'd go with "something not agpl licensed which is already common in OpenStack deployments"
20:22:39 <mordred> notmyname: ability to install marconi in a real production manner without having to install something AGPL
20:22:45 <gabrielhurley> I don't like the increasing number of projects increasing the requisite number of underlying services a deployer has to support.
20:22:48 <kgriffs> gabrielhurley: +1
20:22:49 <ttx> gabrielhurley: +1
20:22:52 <notmyname> ok, so if they ripped out mongo and replaced it with something else, but not "pluggable", that's ok
20:22:53 <mordred> gabrielhurley: +100
20:22:56 <gabrielhurley> we shouldn't be requiring 4 different databases.
20:23:00 <mordred> yes. I would be fine with that
20:23:04 <markwash> gabrielhurley: +1
20:23:06 <mordred> although I doubt they will make that choice :)
20:23:07 <annegentle> gabrielhurley: fo shizzle
20:23:22 * mordred ready
20:23:33 <notmyname> mordred: perhaps, but ttx (and others agreed) that the priority should be on functionality rather than plugins
20:24:01 <dolphm> #ready
20:24:05 <russellb> #set
20:24:10 <markwash> #abstain
20:24:11 <ttx> notmyname: unfair! I was thinking ZeroMQ transport and RabbitMQ backend when I said that
20:24:15 <ttx> :)
20:24:25 <notmyname> ttx: ah, sorry for the misrepresentation
20:24:42 <gabrielhurley> markwash: lol
20:24:45 <hub_cap> lol
20:24:53 <ttx> notmyname: no, you're right. And I stand corrected.
20:24:57 <ttx> #startvote Accept Queue service as a new program with the Marconi project in incubation during the Icehouse cycle? yes, no, abstain
20:24:59 <openstack> Begin voting on: Accept Queue service as a new program with the Marconi project in incubation during the Icehouse cycle? Valid vote options are yes, no, abstain.
20:25:00 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
20:25:06 <notmyname> #vote yes
20:25:07 <russellb> #vote yes
20:25:09 <mikal> #vote yes
20:25:10 <annegentle> #vote yes
20:25:10 <ttx> #vote yes
20:25:14 <markmc_by_proxy> #vote yes
20:25:15 <shardy> #vote yes
20:25:17 <dolphm> #vote yes
20:25:18 <markmcclain> #vote yes
20:25:20 <mordred> #vote yes
20:25:27 <gabrielhurley> #vote abstain
20:25:31 <ttx> russellb: who needs gerrit when we have NICKCHANGE
20:25:37 <russellb> ttx: yes!
20:25:38 <jd__> #vote yes
20:25:43 <vishy> #vote abstain
20:25:45 <ttx> 30 more seconds
20:25:51 <markwash> #vote abstain
20:26:04 <jgriffith> #vote abstain
20:26:17 <ttx> #endvote
20:26:18 <openstack> Voted on "Accept Queue service as a new program with the Marconi project in incubation during the Icehouse cycle?" Results are
20:26:19 <openstack> yes (11): ttx, notmyname, markmc_by_proxy, jd__, annegentle, russellb, mikal, mordred, shardy, dolphm, markmcclain
20:26:20 <openstack> abstain (4): gabrielhurley, markwash, vishy, jgriffith
20:26:30 <ttx> For the abstain votes: is that no opinion? not really convinced ?
20:26:38 <jgriffith> not convinced
20:26:39 <russellb> or haven't had enough time to investigate?
20:27:14 <ttx> jgriffith: fair enough
20:27:15 <jgriffith> Only real argument I see is "Amazon has it"
20:27:17 <russellb> curious which part you're not convinced of though ... it's fit for openstack?  goals?  etc?
20:27:22 <zaneb> jgriffith: not convinced of the need, or not convinced of this approach?
20:27:27 <gabrielhurley> I have nothing against marconi, I just haven't really been convinced that it should be an integrated project under a new Program.
20:27:38 <jgriffith> goals, value, what the implementation is really going to look like etc etc
20:27:42 <gabrielhurley> jgriffith: +1
20:27:47 * markwash got last minute cold feet. .
20:27:53 <hub_cap> lol
20:28:03 <ttx> id devanada back from Burning man ?
20:28:06 <markwash> generally positive
20:28:07 <ttx> is*
20:28:08 <dolphm> (fwiw, i was nearly an abstain, due to concerns about scope creep into notifications)
20:28:15 <ttx> devananda: around ?
20:28:17 <vishy> ttx: just because i missed the first part of the meeting
20:28:25 <ttx> vishy: perfect reason!
20:28:27 <russellb> i'm a little worried about scope creep in transports, even, but i took a leap of faith, heh
20:28:28 <hub_cap> aww ttx u dont wanna do trove first ;)
20:28:36 <jgriffith> haha
20:28:40 <hub_cap> NobodyCam is maybe around in devananda's stead?
20:28:40 <ttx> just making sure I haven't missed something obvious
20:28:51 <gabrielhurley> If marconi does eventually become integrated I'm all for it, fwiw
20:28:54 <kgriffs> we can definitely curtail that scope creep, FWIW. :D
20:29:06 <ttx> hub_cap: ironic should be fast. I'm trying to delay yours so that you can get heat-trusts in :P
20:29:12 <gabrielhurley> lol
20:29:13 <hub_cap> HAH ttx
20:29:13 <zaneb> lol
20:29:26 <hub_cap> ttx: thats a bug afaic
20:29:29 <hub_cap> in trove
20:29:29 <jgriffith> hub_cap: code faster.....
20:29:36 <hub_cap> not a feature, wink wink
20:29:37 <ttx> unfortunately nobody from ironic is around yet...
20:29:40 <russellb> kgriffs: yep, i have faith
20:29:49 <ttx> hub_cap: so you go now
20:29:51 <notmyname> ttx: default deny request?
20:29:52 <ttx> #topic End-of-cycle graduation review: Trove
20:29:58 <flaper87> Thanks guys
20:30:03 <hub_cap> lol notmyname glad im here!
20:30:19 <ttx> hub_cap: Could you describe your current status and progress ?
20:30:22 <hub_cap> so fwiw, the heat optional provisioning is under review and passing tests
20:30:25 <hub_cap> sure ttx
20:30:33 <hub_cap> we are integrated into horizon (thx gabrielhurley +crew)
20:30:46 <hub_cap> we have gone from 2 companies to about 8 with at least 4 actually coding daily
20:30:48 <dolphm> kgriffs: i think my concerns weren't valid... you're clearly venturing into notifications, but the long term scope seems reasonably well defined
20:31:05 <hub_cap> weve cleaned up a good bit of tech debt
20:31:15 <ttx> hub_cap: I think there were two main concerns I think wrt graduation: Heat integration and Tempest integration
20:31:25 <gabrielhurley> hub_cap: ++
20:31:31 <hub_cap> ttx: we decided it woudl be just plain wrong to push tempest on me
20:31:33 <ttx> If those are not addressed yet we at least need to be convinced that they will be addressed soon enough to not adversely impact the Icehouse cycle
20:31:34 <hub_cap> a few wks ago
20:31:50 <hub_cap> heat is under review, w 2 caveats
20:31:50 <shardy> hub_cap: can you link the heat provisioning patch pls?
20:31:56 <hub_cap> https://review.openstack.org/#/c/44530/
20:31:59 <hub_cap> 1) trusts
20:32:17 <hub_cap> 2) resize isint using heat, because we rely on the confirm resize nova step to do some "stuff"
20:32:43 <hub_cap> #2 id like to address by not necessarily using the nova resize, but a smarter cluster-like resize
20:32:44 <ttx> hub_cap: like I just said, i wouldn't block graduation on that... we just need to be convinced this is top prio and will not be a problem during Icehouse
20:32:48 <hub_cap> if possible
20:32:52 <hub_cap> ttx: its my #1 past heat
20:32:53 <shardy> so trusts *should* land really really soon, hopefully tomorrow
20:32:58 <hub_cap> woo shardy
20:33:14 <hub_cap> i really want to do tempest integration really really bad
20:33:15 <jeblair> how about CI integration?
20:33:18 <hub_cap> i want official gating tests
20:33:22 <shardy> hub_cap: https://review.openstack.org/#/c/43380/
20:33:23 <zaneb> hub_cap: but your use case should work now, even without trusts
20:33:33 <ttx> jeblair: that's what I mean by tempest integration
20:33:36 * mordred is specifically less worried about tempest and trove due to existence of trove ci stuff - I know the work exists, so porting it in in icehouse should be doable
20:33:45 <hub_cap> zaneb: it does when i send the password
20:33:56 <jeblair> mordred: it's existed for a long time but still isn't being run in the context of openstack ci
20:33:59 <hub_cap> yes mordred, but i still really want to make all new tests go thru tempest
20:34:02 <mordred> jeblair: agree
20:34:07 <jeblair> it's being run as a third party test system
20:34:12 <hub_cap> i dont know if we want to do that jeblair/mordred
20:34:12 * mordred thinks all of trove ci needs to be integrated
20:34:13 <zaneb> hub_cap: there was a patch to remove that requirement... I thought it got merged
20:34:17 <mordred> hub_cap: we do
20:34:17 <hub_cap> id rather start making tempest tests
20:34:21 <jeblair> which is just a little weird for an official openstack project
20:34:23 <vishy> hub_cap: can the current version be configured to work using full vms instead of openvz containers.
20:34:30 <hub_cap> mordred: id be happy to work w/ you
20:34:34 <hub_cap> vishy: default is not ovz
20:34:38 <hub_cap> its whatever devstack uses
20:34:47 <hub_cap> trove is virt agnostic
20:34:52 <mordred> hub_cap: the question is whether or not we give you a pass on not having done that yet :)
20:34:52 <vishy> hub_cap: corollary: so it might be possible to do it via docker instead?
20:35:07 <vishy> (for example)
20:35:10 <shardy> hub_cap: bug #1217617 will allow most use-cases except autoscaling work with token auth and no trusts
20:35:11 <uvirtbot> Launchpad bug 1217617 in heat "Always requiring password on create is too restrictive" [High,In progress] https://launchpad.net/bugs/1217617
20:35:12 <hub_cap> mordred: well sure, but i dont remember that being a requrement a few mo' ago
20:35:20 <hub_cap> vishy: if docker works w/ nova then hell yes
20:35:20 <ttx> jeblair: the "no graduation without integration tests" condition was added a bit recently, while the trove folks were already struggling to get our "heat integration" condition in
20:35:28 <hub_cap> +1 ttx
20:35:35 <hub_cap> trust me ;) its my highest prio
20:35:37 <mordred> hub_cap: yup. but that's before we realized that letting things graduate without ci integration was getting us into a bad place as a project
20:35:38 <vishy> hub_cap: interesting. Do you make use of cinder volumes?
20:35:39 <hub_cap> past getting heat in
20:35:41 <jeblair> ttx: well, this is your show, but honestly, it's not like devstack or tempest or the openstack ci system has been a secret
20:35:44 <zaneb> hub_cap: actually, my bad. you are waiting on this: https://review.openstack.org/#/c/44400/
20:35:58 <mordred> hub_cap: because you _do_ have some CI, I can be convinced that making it an icehouse task might not be terrible
20:35:59 <hub_cap> zaneb: yessir
20:36:00 <ttx> jeblair: so for that last time I'd settle with "it's my top priority once heat is in, which should be in a few days"
20:36:16 <hub_cap> mordred: should be pretty easy honestly i bet we could get it knocked out in a few days
20:36:31 <gabrielhurley> sounds acceptable to me. there's clearly a plan here.
20:36:46 <ttx> jeblair: but I tend to agree with you it should have been worked on even while not being a strict condition.
20:37:22 <hub_cap> ya we got it running a while ago and havent really mucked w/ it. id gather we could get it running before the summit tho
20:37:55 <hub_cap> its really just running a script on a vm
20:37:57 <hub_cap> thats all we do :P
20:38:07 <hub_cap> and you have the whole "prov a vm" stuff worked out
20:38:08 <mordred> hub_cap: no, I think we're missing each other
20:38:20 <jgriffith> vishy: yes it does, but hub_cap will have to provide details
20:38:21 <hub_cap> mordred: im always missing you /me sniffles
20:38:36 <jgriffith> at least last time I looked
20:38:46 <hub_cap> vishy: yes we do (Sry)
20:38:47 <mordred> hub_cap: we're not interested in infra running additional things that aren't related to devstack-gate
20:39:08 <mordred> hub_cap: we're interested in integrating the things that your script does into the normal scripts
20:39:10 <hub_cap> mordred: ah ic. i dont blame you! then it can be retrofitted id imagine :)
20:39:13 <ttx> hub_cap: is there any missing feature in your havana release that you would definitely want to add during the Icehouse cycle ?
20:39:32 <ttx> like a blatant feature gap ?
20:39:36 <hub_cap> well tempest, buit that not a feature
20:39:45 <ttx> hub_cap: apart from the two already-mentioned
20:39:45 <hub_cap> ttx: i have wants but nothing is missing
20:39:53 <hub_cap> config mgmt is gonna be nice
20:39:56 <hub_cap> and so is clustering
20:40:02 <hub_cap> and replication/resizes
20:40:08 <hub_cap> but those arent missing
20:40:16 <hub_cap> they just arent done yet
20:40:19 <mordred> hub_cap: does troveclient work with normal OS_* env vars yet?
20:40:22 <mordred> oh! the plane is moving
20:40:23 * mordred runs away
20:40:37 <russellb> ha.
20:40:46 <hub_cap> HA mordred not yet but i can knock that out fast
20:40:51 <mordred> please do
20:40:59 <hub_cap> aboslutely
20:41:02 <NobodyCam> gah thought today was monday :(
20:41:04 <hub_cap> ill put it as a bug in rc1
20:41:13 <hub_cap> cuz it is a bug to me
20:41:22 <ttx> hub_cap: you mentioned 4 companies working daily on the project now... but then you seem to be the only one working on graduation-essential stuff like CI and heat integration ?
20:41:38 <ttx> does that mean they contribute tactical features ?
20:41:38 <hub_cap> heh.. well yes
20:41:47 <hub_cap> basically ya
20:42:18 <hub_cap> mirantis/cern made rhel work, and are making the dev env a bit nicer before they jump in full force w/ features
20:42:19 <ttx> when you say "you", is that "rackspace" or really just you ? Worrying about bus factor
20:42:34 <hub_cap> i prefer the lotto factor ttx
20:42:50 <hub_cap> im not the only person working on public facing interests
20:43:01 <hub_cap> i just took it upon myself to tackle the graduation stuff
20:43:12 <hub_cap> horizon got done via rax/cern/mirantis together
20:43:20 <hub_cap> devstack is a hp thing thats happening
20:43:33 <hub_cap> so other teams/companies are working on non tactical features
20:43:47 <hub_cap> well non $tactical$ features
20:44:00 <ttx> OK. On my side I can report that trove has been following the release schedule for a few milestones already and it worked really well
20:44:13 <hub_cap> yes thank you ttx for the help w/ all that
20:44:16 <ttx> so no objection from a release management perspective
20:44:24 <hub_cap> its nice to be in the know at the begin of the process
20:44:28 <hub_cap> rather than after graduation
20:44:46 <hub_cap> i liked a full cycle of milestones + rc's (when that comes), ff, etc..
20:45:02 <ttx> maybe other programs could chime in (docs, QA, infra) -- although jeblair already raised his QA/Infra concerns
20:45:25 <ttx> annegentle: is Docs fine with trove being integrated in Icehouse ?
20:45:31 <hub_cap> docs http://docs-draft.openstack.org/30/44530/3/check/gate-trove-docs/cef4dbc/doc/build/html/
20:45:41 <ttx> annegentle: I guess that would be a prio 2 (non core) project ?
20:45:44 <annegentle> so for a docs perspective, they have done the work we've outlined with openstack processes
20:45:48 <hub_cap> also we have a mirantis fellow who has created docs for a existing install
20:45:57 <hub_cap> thats under review
20:46:33 <hub_cap> https://review.openstack.org/#/c/44668/
20:46:40 <ttx> OK, I'm ready to vote -- other questions ? Anyone wanting a delay before we vote ?
20:46:58 <ttx> it's our first end-of-cycle graduation review so the rules are a bit fresh
20:47:03 <hub_cap> :)
20:47:14 <hub_cap> horray /me is a guinea pig
20:47:21 <hub_cap> oh nm
20:47:26 <hub_cap> just this cycle lol
20:47:29 <russellb> #ready
20:47:29 <ttx> I guess we could do it over two meetings if that gives people time to look into the code with more details
20:47:37 <russellb> #ornot
20:47:40 <hub_cap> HA
20:47:51 <ttx> although I don't expect that many people reading random code those days
20:47:54 * jgriffith is still ready
20:48:01 <russellb> ttx: quite a busy week ahead ...
20:48:02 <hub_cap> w/ the review rush ttx ;)
20:48:06 * gabrielhurley is ready
20:48:07 <dolphm> russellb: ++
20:48:14 <mikal> ready
20:48:20 <russellb> steady!
20:48:28 <hub_cap> go?
20:48:32 <jgriffith> :)
20:48:38 <russellb> -2
20:48:40 <russellb> kidding.
20:48:46 <jgriffith> russellb: slap!
20:48:57 <hub_cap> ouch i got a big red X from russellb
20:48:58 <ttx> I'm giving it another minute to make sure everyone is fine with voting now
20:49:02 <jgriffith> :)
20:49:09 <hub_cap> hes good at giving those ;)
20:49:34 <annegentle> fine with voting now
20:49:38 <jd__> same
20:49:59 <ttx> #startvote Graduate Trove to be integrated during the Icehouse cycle? yes, no, abstain
20:50:00 <openstack> Begin voting on: Graduate Trove to be integrated during the Icehouse cycle? Valid vote options are yes, no, abstain.
20:50:01 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
20:50:07 <russellb> #vote yes
20:50:09 <jgriffith> #vote yes
20:50:11 <markwash> #vote yes
20:50:12 <shardy> #vote yes
20:50:13 <markmc_by_proxy> #vote yes
20:50:13 <ttx> #vote yes
20:50:13 <gabrielhurley> #vote yes
20:50:16 <markmcclain> #vote yes
20:50:27 <mikal> #vote yes
20:50:33 <dolphm> #vote yes
20:50:35 <ttx> 30 more seconds
20:50:45 <vishy> #vote yes
20:50:48 <jd__> #vote yes
20:50:50 <notmyname> #vote abstain
20:50:51 <annegentle> #vote yes
20:51:11 <ttx> #endvote
20:51:13 <openstack> Voted on "Graduate Trove to be integrated during the Icehouse cycle?" Results are
20:51:14 <openstack> yes (13): ttx, vishy, markmc_by_proxy, jd__, annegentle, shardy, russellb, markwash, mikal, gabrielhurley, dolphm, jgriffith, markmcclain
20:51:16 <openstack> abstain (1): notmyname
20:51:21 <dolphm> no mordred?
20:51:28 <gabrielhurley> he got on a palce
20:51:29 <hub_cap> mordred: is on a plane
20:51:30 <ttx> mordred: on a plae now
20:51:32 <gabrielhurley> *plane
20:51:37 <hub_cap> snakes not included
20:51:47 <hub_cap> thx everyone!!
20:51:48 <ttx> #topic End-of-cycle graduation review: Ironic
20:51:53 <ttx> hub_cap: congrats!
20:51:54 <russellb> hub_cap: nice work
20:52:02 <ttx> still no ironic dude around ?
20:52:05 <zaneb> congrats hub_cap
20:52:09 <NobodyCam> me
20:52:16 <ttx> NobodyCam: hi!
20:52:20 <NobodyCam> deva is driving right now
20:52:29 <ttx> NobodyCam: So my understanding is it's pretty obvious this one should not graduate, since devananda himself said it's not ready
20:52:38 <ttx> ...and it does not really work yet...
20:52:42 <NobodyCam> correct
20:52:49 <ttx> So the question is more... should it continue to be incubated or should it be dropped because it's not going anywhere ?
20:53:06 <NobodyCam> it is going forward!
20:53:08 <annegentle> oh I hope we keep incubating
20:53:16 <vishy> not going anywhere?
20:53:23 <mikal> I would be opposed to dropping it from incubation
20:53:29 <mikal> Although I also don't have the time to work on it...
20:53:40 <ttx> vishy: s/because/if/
20:53:58 <ttx> vishy: just making sure we still all want it in incubation.
20:54:00 <russellb> ironic - 72 code reviews in the laast 30 days
20:54:02 <vishy> i haven't been keeping up with the status
20:54:04 <russellb> so still moving
20:54:05 <notmyname> ttx: perhaps a better way to phrase it is "should it be dropped and reapply when it is ready"
20:54:07 <markwash> stupid english lacking subjunctive
20:54:12 <russellb> 462 reviews in the last 90 days
20:54:17 <russellb> so, quieter in the last month
20:54:21 <mikal> russellb: do you have stats on how many active devs?
20:54:25 <lifeless> well, burning man
20:54:28 <russellb> lifeless: yeah
20:54:28 <NobodyCam> yes we are moving forward we had a lot vacations last couple of weekes
20:54:32 <russellb> mikal: just reviewers
20:54:37 <russellb> http://russellbryant.net/openstack-stats/ironic-reviewers-90.txt
20:54:42 <ttx> notmyname: yes, better. That said I'm very fine with projects taking more than one cycle to incubate properly
20:54:55 <gabrielhurley> do you think it'll take one more cycle to be ready? two? three?
20:55:02 <dolphm> mikal: 24?
20:55:05 <ttx> I would be glad to set a precedent, actually
20:55:09 <notmyname> ttx: one of the issues being discussed now is that graduation isn't "winning" just as not graduation isn't "losing"
20:55:22 <dolphm> mikal: ah, nvm
20:55:22 <notmyname> ttx: (being discussed in context of the "what is core" stuff)
20:55:26 <mikal> dolphm: wow, more than I expected
20:55:35 <russellb> define active
20:55:50 * mordred back
20:56:07 <ttx> mordred: trove was accepted, discussing ironic now
20:56:12 <russellb> http://stackalytics.com/?release=havana&metric=commits&project_type=openstack&module=ironic&company=&user_id=
20:56:16 <russellb> ironic contributors ^
20:56:27 * mikal looks
20:56:33 <ttx> So I'm fine with voting for keeping Ironic in incubation now
20:56:54 <mordred> do we need to vote on that?
20:56:56 <ttx> not sure we really need more time to discuss it, but was can also wait until next week
20:56:58 <russellb> looks like the project could use some more heavy contributors beyond devananda
20:57:11 <jgriffith> russellb: +1
20:57:18 <ttx> mordred: maybe ? It's the first cycle we do end-of-cycle reviews :)
20:57:25 <mordred> I was sort of looking forward to seeing the incubation process for this last until it's ready
20:57:30 <ttx> mordred: I'm fine with not voting too.
20:57:48 <russellb> probably only needs a vote if anyone suggests that it be dropped
20:57:54 <russellb> i think the default answer should be that it stays
20:57:56 <mordred> ttx: it is currently incubated - nothing automatically removes that - I think it's just still incubated unless someone wants to move that we un-incubate it
20:58:02 <mordred> what russellb said
20:58:02 <ttx> mordred: we need some way to remove dead stuff from incubation, but that can be ad-hoc rather than a regular review
20:58:14 <ttx> russellb: +1
20:58:37 <mordred> ttx: I think someone will propose we drop it if it's dead
20:58:48 <ttx> #info Ironic stays in incubation for the Icehouse cycle
20:58:51 <mordred> yikes. laggy plane wifi
20:58:57 <russellb> dang, productive meeting :)
20:59:07 <jgriffith> mordred: first world problems
20:59:07 <ttx> #topic Open discussion
20:59:08 <dolphm> mikal: since june, 19 authors (including jenkins and a dupe)... so 17 by my count http://paste.openstack.org/raw/45694/
20:59:19 <ttx> all: You might have noticed that we're not using Gerrit to vote yet
20:59:23 <markwash> jgriffith: lol
20:59:28 <ttx> The repository is almost set up, blocking on bug 1219731
20:59:30 <uvirtbot> Launchpad bug 1219731 in openstack-ci "Populate tech-committee* groups in Gerrit" [Undecided,New] https://launchpad.net/bugs/1219731
20:59:39 <ttx> Then we can review/accept the initial commit of reference documents at:
20:59:43 <ttx> #link https://review.openstack.org/#/c/44489/
20:59:50 <ttx> Anything else, anyone ?
21:00:39 <ttx> markmc from France !
21:00:48 <russellb> markmc: oh snap!
21:00:49 <markmc> yah :)
21:00:50 * mordred thinks everyone is great - and also you all might have blinky lights around your head in mmy mind
21:01:18 <ttx> ok then, next meeting
21:01:21 <ttx> #endmeeting