16:01:35 <VW> #startmeeting Large Deployments Team February 2016 Meeting
16:01:36 <openstack> Meeting started Thu Feb 18 16:01:35 2016 UTC and is due to finish in 60 minutes.  The chair is VW. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:37 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:40 <openstack> The meeting name has been set to 'large_deployments_team_february_2016_meeting'
16:02:01 <VW> hello, folks
16:02:27 <mdorman> o/
16:02:58 <belmorei_> hi
16:04:04 <VW> yay!  belmoreira is back
16:04:19 <VW> #topic Manchester recap
16:04:47 <belmoreira> I think is on me :)
16:04:50 <VW> for those of us who couldn't make it, here are the notes from the LDT session there - https://etherpad.openstack.org/p/MAN-ops-Large-Deployment-Team
16:05:00 <belmoreira> https://etherpad.openstack.org/p/MAN-ops-Large-Deployment-Team
16:05:02 <VW> yes, and special thanks to belmoreira for running that session
16:05:26 <openstackgerrit> Merged openstack/osops-tools-monitoring: Client compatibility fix  https://review.openstack.org/281167
16:05:57 <belmoreira> We were around 10 in the room
16:06:41 <belmoreira> and operators that attend the session didn't particularly have a large installation
16:06:48 <VW> hmm
16:07:00 <mdorman> well if you “feel” large...
16:07:08 <VW> seemed like a lot of interested in cells though
16:07:15 <belmoreira> but they expect to grow
16:07:23 <VW> how we set them up, decisions made, etc
16:08:02 <belmoreira> most of the ops there were interested to know when make sense to run cells
16:08:50 <belmoreira> mdorman yes, one of the problems is to define large
16:09:42 <VW> yeah - I suppose we could try to come up with a deffinition
16:09:55 <VW> but seems like we'd inadvertently leave someone out
16:10:00 <belmoreira> the session moved more to describe CERN/Rackspace experience with cells
16:10:06 <VW> or imply they didn't fit
16:10:43 <belmoreira> VW I dont think we gave that impression
16:10:53 <VW> no, I don't think we did
16:11:02 <VW> was saying if we tried to define  "large" we might
16:11:17 <mdorman> i feel like the right recommendation now is to wait for v2.   and then it’s kind of a moot question of cells vs. not cells…  just a question of when’s the right time to scale out to more.
16:11:38 <VW> yeah, but that is dragging on :\
16:11:49 <belmoreira> people was simple curious how to run cells and learn/hear others for when they need to scale there infrastructure
16:12:47 <mdorman> VW:  true.
16:13:04 <VW> John G. (johnthetubaguy) gave and update on V2 in one of the general sessions, didn't he belmoreira
16:13:12 <belmoreira> yes
16:13:44 <belmoreira> unfortunately I couldn't join this session
16:14:41 <VW> ah - bummer
16:14:49 <belmoreira> the cells v2 update session basically trigger a lot of curiosity in cells and all those questions
16:15:17 <mdorman> makes sense
16:16:09 <VW> Yeah - I'm getting more and more interested in how we help land V2
16:16:09 <belmoreira> I think it was interesting
16:16:44 <VW> got woken up last night specifically because of a delay in meta data between the cell and the top level
16:16:55 <VW> ready for cells issues to be "mainstream" :)
16:16:57 <belmoreira> great to have Robert from Rackspace in the room to give his experience
16:17:01 <logan-> i have the same curiosity as an operator with a small deployment, lookign at growth.. things like when the cutoff is where cells make sense, whether the transition is worthwhile or if you think you might need them deploy with it right away? (probably not with the v1/v2 stuff?). I don't want to interrupt the LDT meeting with small deployment stuff but I am glad to see that touched on in the etherpad :)
16:17:34 <VW> I have to step a way for five minutes
16:17:38 <VW> keep talking...
16:18:01 <yankcrime> johnthetubaguy's short statement on cells v2 was basically yes it'll be in mitaka as the default, but it'll only support a single cell
16:19:04 <belmoreira> Ops will need to deal with a new DB
16:19:18 <belmoreira> it's already there in Liberty but was not used
16:20:07 <yankcrime> http://www.slideshare.net/johnthetubaguy/nova-update-openstack-ops-midcycle-manchester-feb-2016
16:21:24 <yankcrime> http://lists.openstack.org/pipermail/openstack-dev/2016-January/084545.html (reference from https://etherpad.openstack.org/p/MAN-ops-Nova)
16:25:10 <mdorman> any other big news from the meetup?
16:25:51 <VW> back - sorry
16:26:02 <belmoreira> there was a status update in Live Migration
16:26:08 <belmoreira> https://etherpad.openstack.org/p/MAN-ops-Live-migration
16:26:19 <VW> great to hear, thanks for sharing, yankcrime
16:26:49 <belmoreira> and i this was interesting: https://etherpad.openstack.org/p/MAN-ops-Top-10
16:27:22 <VW> I need to get up to speed on Live Migration.  From my perspective, it's a thing.  We use it all the time
16:27:33 <VW> but doesn't seem to be so for a lot of folks
16:28:09 <belmoreira> most of the work in live migration is done
16:29:00 <VW> looks like Rob - one of our engineers was in the conversation.  See a bunch of notes about our tooling in that etherpad
16:29:29 <belmoreira> I need to leave in 5 min.
16:29:41 <VW> ah, I see belmoreira
16:29:54 <VW> any other non-mid cycle stuff you want to cover?
16:30:35 <belmoreira> mdorman do you know more about the network segmentation spec
16:30:57 <VW> #topic Network Segmentation
16:31:03 <belmoreira> it's still in review :)
16:32:02 <VW> ah review... :)
16:32:08 <mdorman> i don’t have much new info, other than that carl is trying to push it forward.
16:32:28 <VW> I know that Brian S from here was helping.  I have not gotten to sit down with him on it yet, though
16:32:41 <mdorman> i think the neutron midcycle is in the next week or two.  then klindgren and i were going to meet up with carl to see where things are at (all three of us are local here in denver)
16:32:55 <VW> nice!
16:33:05 <VW> I'll try to catch Brian before the midcycle then
16:33:08 <mdorman> i re-read the spec a few weeks ago and it’s quite a bit different than the original, but there was nothing that stood out to me that was a major red flag
16:33:32 <mdorman> it’s pretty watered down.   even once this spec is approved and implemented, we’re still quite a ways off from getting everything that was in the ask.  but it’s a step forward for sure.
16:34:11 <belmoreira> mdorman yes
16:34:27 <belmoreira> I need to leave...
16:36:06 <VW> yeah, mdorman - I'm trying to haraass some folks here about pushing more of quark into ml2
16:36:16 <VW> uphill battle at the moment
16:36:31 <VW> but trying to bridge the gaps between or public and private clouds
16:36:37 <VW> this and cells v2 are a big part of that
16:37:45 <mdorman> yeah, it’s all related
16:38:24 <VW> any other stuff you wanted to dive deep into, mdorman?
16:39:49 <mdorman> i think that covers any of our concerns at the moment.   klindgren couldn’t join this morning b/c his internet is getting fixed right now, wah wah
16:40:46 <VW> gotcha
16:41:03 <VW> that's about it for me too
16:41:21 <VW> was mainly interested in getting caught up on mid-cycle
16:41:29 <mdorman> yup same here
16:41:45 <VW> cool - then we will wrap up early unless anyone has objections
16:42:16 <mdorman> i’m good.
16:42:44 <VW> cool
16:42:52 <VW> well, thanks for attending!
16:43:55 <VW> #endmeeting