14:01:23 <ihrachys> #startmeeting networking
14:01:24 <openstack> Meeting started Tue Apr 12 14:01:23 2016 UTC and is due to finish in 60 minutes.  The chair is ihrachys. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:01:25 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:27 <openstack> The meeting name has been set to 'networking'
14:01:28 <johnsom> o/
14:01:28 <korzen> hello
14:01:32 <ihrachys> hello my friends!
14:01:40 <njohnston> \o/
14:01:43 <reedip__> |o|
14:01:48 <HenryG> o/
14:01:52 <rtheis> o/
14:02:00 <mestery> o/
14:02:04 <ajo> ~o~
14:02:07 <amuller> (^.^)
14:02:12 <ihrachys> #topic Announcements
14:02:16 <ski3> hey!
14:02:18 <ihrachys> first thing first...
14:02:22 <ihrachys> Mitaka is finally released
14:02:25 <ihrachys> woohoo
14:02:33 <ajo> yeah :D
14:02:33 <dasm> uff.
14:02:40 * regXboi pulls the cord for the balloon drop
14:02:45 <blogan> hi
14:02:50 <reedip__> Thank God :)
14:03:02 <ihrachys> but Newton is already open for a while, so it's not like a huge change for most of you
14:03:08 <ihrachys> still, stable/mitaka is now open
14:03:15 <ihrachys> and we can land backports there. so welcome.
14:03:31 <ihrachys> #topic Summit planning
14:03:53 <ihrachys> as you all probably know, we will have some design summit sessions for neutron
14:04:11 <dasm> ihrachys: 2 days. am i right?
14:04:18 <ihrachys> kinda
14:04:26 <ihrachys> it's one whole day and then another half
14:04:28 <ihrachys> #link http://lists.openstack.org/pipermail/openstack-dev/2016-April/092049.html
14:04:36 <dasm> mhm. gotcha
14:04:40 <ihrachys> ^ armax posted a draft schedule, take a look
14:04:52 <ihrachys> note that Tue will be dedicated to cross-project sessions.
14:05:08 <ihrachys> so for neutron specific matters it's Wed and part of Thu
14:05:12 <ihrachys> and then Fri for meetups
14:05:18 <ihrachys> in case someone is still alive
14:05:22 <amuller> There's less pure Neutron slots this time around, and more slots for Neutron stadium projects
14:05:51 <amuller> We're probably gonna organize a few Friday sessions, I suspect testing, vlan aware vms (That one is a tradition we must uphold)
14:05:53 <salv-orlando> amuller: stadium? what stadium? ;)
14:06:04 <salv-orlando> the stadium is dead. Long live the stadium!
14:06:08 <ihrachys> there will be another for upgrades, ...
14:06:13 <ajo> Fri probably will have a QoS meetup too :)
14:06:13 <ihrachys> anyhow.
14:06:25 <ihrachys> if you go to the schedule, you can see etherpad links there for each session
14:06:32 <ihrachys> most of them are half baked or empty
14:06:34 <Hunter> #openstack-meeting
14:06:52 <ihrachys> but folks that will lead those sessions are working on filling them in with relevant stuff
14:07:04 <mhickey> hello
14:07:07 <amuller> agendas / abstracts should be done this week
14:07:14 <ihrachys> last but not least, register for the official openstack party, it requires RSVP :)
14:07:32 <ajo> ihrachys, link?
14:07:38 <ihrachys> ^ yeap. what amuller told.
14:08:14 <ihrachys> #link https://www.eventbrite.com/e/stackcity-austin-a-community-festival-for-stackers-tickets-24174378216
14:08:33 <ihrachys> I guess we will plan for neutron gathering too, but I will leave it to the next meeting.
14:08:35 <ajo> ihrachys++
14:08:51 <ihrachys> ok, moving on to the next topic which is...
14:09:01 <ihrachys> #topic Gate status
14:09:10 <dasm> one more event. sunday: https://www.eventbrite.com/e/hold-my-beer-while-i-register-mini-brewery-tour-of-austin-tickets-24244360535
14:09:38 <ihrachys> dasm: thanks
14:09:42 <ihrachys> a lot happened with gate lately
14:09:51 <ihrachys> 1. some of you could notify api job is non voting
14:10:08 <ihrachys> that's because we import from tempest, and there were breaking changes there
14:10:21 <amuller> about api job, right now it's non-voting so we can make some changes to project-config
14:10:22 <ihrachys> amuller and armax are hopefully on top of getting the job back in shape and voting
14:10:24 <amuller> We have this patch: https://review.openstack.org/#/c/274023/
14:10:32 <amuller> and 3 patches linked in the review comments for that patch
14:10:42 <amuller> we should get those in this week then flip the switch back to voting
14:10:47 <amuller> and we'll finally use the tempest plugin interface, which is neat
14:10:55 * ihrachys highly encourages reviewers to jump into those reviews.
14:11:08 <ihrachys> it's everyone's gate, right? :)
14:11:26 <ihrachys> once we get the plugin, we don't import from tempest, and all goes well, and we get back to voting.
14:11:32 <ihrachys> another thing to note on gating is...
14:11:48 <ihrachys> 2. a lot of fixes for fullstack/functional jobs landed lately
14:12:00 <ihrachys> that will hopefully make the jobs less fragile
14:12:08 <ihrachys> it's still ongoing
14:12:33 <ihrachys> again, amuller and jlibosva are on top of that. I encourage everyone to help them getting things done.
14:12:39 <amuller> fullstack is in good shape, it's non-voting so that's less of a concern to stability. The functional job is voting and currently has a very high failure rate.
14:12:46 <ihrachys> for what I know, it's both reviews and debugging time that are needed right now.
14:12:52 <amuller> This is the top issue: https://bugs.launchpad.net/neutron/+bug/1567668
14:12:53 <openstack> Launchpad bug 1567668 in neutron "Functional job sometimes hits global 2 hour limit and fails" [High,In progress] - Assigned to Jakub Libosvar (libosvar)
14:12:54 * dasm is looking at grafana and indeed. failure rate dropped a little
14:12:57 <dasm> http://grafana.openstack.org/dashboard/db/neutron-failure-rate?panelId=5&fullscreen
14:13:20 <amuller> jlibosva is currently working on gathering more logs so we have a chance to understand what's wrong
14:13:44 <amuller> it starts from this patch https://review.openstack.org/#/c/303053/
14:13:48 <amuller> currently 4 patches in that chain
14:14:08 <ihrachys> amuller: why WIP on the bottom?
14:14:10 <amuller> right now it's just logging stuff, jlibosva correct me if I'm wrong but we currently don't have a solid lead on what is actually causing the issue
14:14:23 <amuller> ihrachys: just haven't updated the commit message, it's good to go on my end
14:14:36 <jlibosva> The latest known thing is that ovsdb native interface gets stuck probably when creating bridge
14:14:46 <ihrachys> other patches have +W but they are waiting for the first one to proceed
14:14:57 <jlibosva> the reason why job gets stuck as well is probably due to cleaups
14:15:06 <jlibosva> as they get stuck on the very same call as job
14:15:11 <jlibosva> s/job/test/
14:15:13 <ihrachys> jlibosva: amuller: do we need someone to help you guys?
14:15:22 * jlibosva looks at otherwiseguy
14:16:23 <ihrachys> ok make sure you have the needed help, and reach out.
14:16:28 <amuller> ihrachys: we could always use help. People can look at the final patch in the series, it reproduced the issue with logs
14:16:36 <amuller> ihrachys: I've been unable to reproduce this locally, which is always fun...
14:16:52 <ihrachys> link to logs?
14:17:02 <jlibosva> http://logs.openstack.org/94/303594/2/check/gate-neutron-dsvm-functional/1f86f1d/
14:17:10 <ihrachys> thanks jlibosva
14:17:22 <ihrachys> ok, on bright side of the gate...
14:17:24 * ajo hates when he's not able to reproduce locally
14:17:33 <ihrachys> usual tempest jobs are pretty stable lately
14:17:40 <ihrachys> #link http://grafana.openstack.org/dashboard/db/neutron-failure-rate?panelId=4&fullscreen
14:18:02 <ihrachys> and that's a good time to add more jobs
14:18:13 <ihrachys> so first thing, dvr job is now voting!
14:18:40 <carl_baldwin> In the gate queue no less.
14:18:45 <ihrachys> second thing, grenade multinode is on its way too, I expect it to be sorted out before summit
14:18:46 <amuller> cheers to the DVR team
14:19:04 <ajo> +++ congrats
14:19:30 <ihrachys> yeah, thanks everyone for getting it to where it is.
14:19:38 <ihrachys> it was a long way.
14:19:48 <ihrachys> #topic Bugs
14:20:02 <ihrachys> previous week amotoki served as the deputy
14:20:08 <ihrachys> and blogan is this week deputy
14:20:18 <ihrachys> amotoki: anything worthy?
14:20:20 <blogan> yes
14:20:29 <amotoki> i think the deputy of the previous week was reedip
14:20:42 <ihrachys> oh sorry, I misread
14:20:51 <amotoki> but these weeks were quiet
14:20:55 <blogan> i saw reedip doing bugs, but i thought armax said it was you last week
14:20:56 <ihrachys> reedip: any specific bugs that need team attention?
14:21:47 <ihrachys> blogan: yeah, that's why I assumed
14:22:12 <ihrachys> hm, it does not seem like reedip is ready for an update. I will reach him later.
14:22:27 <ihrachys> blogan: thanks for volunteering for the week.
14:22:29 <blogan> ill try to hit him up too see if there's anything i need to know
14:22:40 <ihrachys> thanks
14:22:50 <ihrachys> we would need volunteers for next weeks.
14:22:52 <blogan> ihrachys: its about time i do it :)
14:23:10 <ihrachys> it may be hard to get someone on the summit week, so I highly encourage folks that don't go to step in
14:23:33 <ihrachys> but I will leave that to be sorted on the next meeting.
14:23:48 <dasm> ihrachys: armax will deal with it ;)
14:23:52 <ihrachys> finally, the main part of the meeting, which is...
14:24:00 <ihrachys> #topic Newton blueprints
14:24:10 <ihrachys> #link https://launchpad.net/neutron/+milestone/newton-1
14:24:30 <ihrachys> the last time we were going thru the list, but have not completed it
14:24:59 <amotoki> #link https://blueprints.launchpad.net/neutron/newton/+assignments
14:24:59 <ihrachys> we should check all BPs to see whether new assignees are needed, and whether we want those to retarget.
14:25:23 <ihrachys> amotoki: thanks, that's useful
14:26:07 <carl_baldwin> I'm looking to target https://blueprints.launchpad.net/nova/+spec/vm-boot-with-unaddressed-port for Newton.
14:26:18 <ihrachys> the last time we ended at fwaasv2 and vlan-aware-vms.
14:26:32 <carl_baldwin> It turns out there is some synergy with it and routed networks' need for deferred IP address assignement.
14:26:47 <ihrachys> it's nova?
14:27:15 <ihrachys> I guess that's something to discuss with nova folks
14:27:17 <carl_baldwin> ihrachys: Maybe I got the wrong one, I thought there was a Neutron verison.
14:27:19 <carl_baldwin> version.
14:27:25 * carl_baldwin goes looking.
14:27:49 <ihrachys> is it https://blueprints.launchpad.net/neutron/+spec/vm-without-l3-address ? we will get to it later.
14:28:00 <ihrachys> ok first thing is.
14:28:01 <ihrachys> #link #link https://blueprints.launchpad.net/neutron/+spec/neutron-lib
14:28:27 <ihrachys> dougwig: wanna give an update on the spec?
14:28:35 <carl_baldwin> ihrachys: Yep, that's it.  Sorry.
14:28:35 <ihrachys> sorry, bp
14:28:50 <ihrachys> looking at the gerrit topic
14:28:51 <ihrachys> #link https://review.openstack.org/#/q/topic:bp/neutron-lib,n,z
14:28:58 <ihrachys> there is not much to review.
14:29:38 <ihrachys> mestery: since you are the approver, can you help with that?
14:29:42 <HenryG> Well, I have some patches
14:30:00 <mestery> ihrachys: Looking
14:30:00 <amotoki> I think most reviews have no bp/neutron-lib topic
14:30:03 <amotoki> https://review.openstack.org/#/q/project:openstack/neutron-lib+status:open
14:30:11 <amuller> ihrachys: Is there a point we can say the neutron-lib blueprint is accomplished? Otherwise we don't need a blueprint anymore to track it
14:30:30 <ihrachys> amuller: exactly my question, you beat me. when are we DONE?
14:30:41 <HenryG> No
14:31:20 * ihrachys tries hard to relate answer to a question
14:31:24 <HenryG> We have a session at the summit for neutron-lib
14:31:57 <HenryG> There we will outline/discuss goals
14:32:19 <HenryG> And maybe figure out how to decide when to close the BP
14:32:19 <ihrachys> HenryG: ok, fair enough. let's postpone.
14:32:30 <HenryG> Because it will go on forever
14:32:51 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/service-subnets
14:33:11 <ihrachys> I saw haleyb respinned the spec.
14:33:28 <haleyb> ihrachys: i need to update it again
14:33:47 <ihrachys> ok as long as we have active interest.
14:33:47 <haleyb> i saw your comment on tags, we decided against that at the mid-cycle
14:34:08 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/subnet-onboard
14:34:34 <ihrachys> carl_baldwin: is it on your radar? do you have cycles?
14:36:00 <ihrachys> hm. ok let's assume the answers are yes and yes.
14:36:08 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/enginefacade-switch
14:36:23 <ihrachys> HenryG: akamyshnikova: you handle that right? we don't have a spec or smth, so it's a matter of review.
14:36:40 <carl_baldwin> ihrachys: yes and yes
14:36:58 <ihrachys> carl_baldwin: as I thought ;)
14:37:05 * carl_baldwin was delayed while focusing on the radar screen for a moment.
14:37:32 <reedip__> hi
14:37:34 <ihrachys> HenryG: how huge is the work there?
14:37:51 <HenryG> ihrachys: We are working on a mini-spec listing the steps needed
14:38:03 <amotoki> I see WIP patch has been posted on engine facade. it says step-1, so it woud be nice if the milestone is clarified.
14:38:10 <akamyshnikova> ihrachys, there is one patch on review and more to come :)
14:38:36 <ihrachys> akamyshnikova: do we have an estimation of the number of steps to take?
14:39:53 <akamyshnikova> ihrachys, it will be two or three steps, but the last one can contain a lot of patches, I'm still looking onto this
14:40:14 <ihrachys> ok, thanks. let's wait for the spec.
14:40:19 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/dhcp-options-per-subnet
14:40:29 <ihrachys> that one is rather dead
14:40:41 <ihrachys> there is a spec: https://review.openstack.org/247027
14:40:47 <ihrachys> but it's not actively worked on
14:40:59 <amotoki> yes, the author was not active lately..
14:41:09 <ihrachys> I guess unless the author gets back to work, we will realistically drop it from the target.
14:41:38 <ihrachys> anyone interested to revive the work?
14:42:19 <amotoki> i will check the author about his bandwidth
14:42:27 <ihrachys> I guess the answer is no. so I will discuss the drop with armax later.
14:42:38 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/bgp-spinout
14:43:22 <ihrachys> I believe we are on track with it
14:43:31 <ihrachys> there are infra/governance patches in flight
14:43:33 <ihrachys> https://review.openstack.org/#/c/268726/
14:43:51 <ihrachys> vikram_: I assume you will get it to completion in N1-N2
14:44:10 <vikram_> ihrachys: sure
14:44:29 <vikram_> ihrachys: Need help for the repo creation approval
14:44:32 <ihrachys> carl_baldwin: you had concerns on that spin-off?
14:44:49 <ihrachys> oh it was just about Mitaka.
14:45:03 <carl_baldwin> ihrachys: right
14:45:10 <carl_baldwin> I'm +1 now
14:45:28 <vikram_> some minor polishing required to make armax happy ;)
14:45:42 <vikram_> will discuss and publish the next PS soon
14:46:02 <ihrachys> thanks. moving to the next one which is...
14:46:08 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/keystone-v3
14:46:19 <dasm> here
14:46:50 <ihrachys> dasm: spec re-proposed for N, right?
14:46:53 <dasm> ihrachys: i've seen your comments. will try to address it asap. i talked to HenryG and he's also planning to send some updates on this.
14:46:58 <dasm> ihrachys: yes.
14:47:12 <dasm> if there is still interest on get rid of legacy code, i'm in.
14:47:12 <ihrachys> cool. I guess it's safe to say there will be progress on that one in N1.
14:47:49 <dasm> definitively
14:47:59 <ihrachys> HenryG: please work with dasm to get spec ready, so that dasm can do coding
14:48:08 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/vm-without-l3-address
14:48:14 <ihrachys> carl_baldwin: that's the one you wanted to raise.
14:48:27 <ihrachys> it's in Slow progress right now
14:48:46 <carl_baldwin> ihrachys: I've assigned it to me.  I'm going to pick it up, at least for the hard part which is the port deferral.
14:49:11 <carl_baldwin> So, I guess I'll be spinning the spec very soon.
14:49:45 <ihrachys> carl_baldwin: thanks a lot, it's in good hands now :) ajo, make sure you provide feedback on the spec.
14:49:53 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/ml2-ovs-qos-with-dscp
14:50:00 <ihrachys> hm. how is it still a thing?
14:50:08 <ihrachys> haven't we landed it already?
14:50:13 <njohnston> I guess it's just waiting for the fullstack tests to merge
14:50:19 <ihrachys> aha
14:50:20 <njohnston> that's all that is left
14:50:21 <ihrachys> link?
14:50:42 <njohnston> https://review.openstack.org/#/c/288392/
14:50:57 <ihrachys> ok I will take a look after the meeting.
14:51:07 <njohnston> and those are waiting on the next release of python-neutronclient so we can get a version that has the client dscp code for gate tests
14:51:20 <ihrachys> amotoki: are we ready to release client?
14:51:39 <amotoki> ihrachys: will check soon. i don't see any problesm now.
14:51:55 <ihrachys> amotoki: then let's cut it off and allow qos folks to proceed with fullstack
14:52:02 <ihrachys> amotoki: let's discuss it after the meeting.
14:52:03 <amotoki> ihrachys: sure
14:52:06 <njohnston> thanks
14:52:13 <ihrachys> ok cool. and the last one is...
14:52:14 <ihrachys> #link https://blueprints.launchpad.net/neutron/+spec/vmware-vpnaas-service-driver
14:52:21 <ihrachys> I am not sure why it's in neutron.
14:52:31 <ihrachys> is it because vpnaas does not split drivers out of tree?
14:52:41 <blogan> ihrachys: if its like lbaas that is correct
14:52:55 <blogan> and im pretty sure it is
14:53:18 <ihrachys> the patch in question is really tiny
14:53:18 <amotoki> i think we can do either: in-tree and out-of-tree
14:53:19 <ihrachys> https://review.openstack.org/#/c/303226/3
14:53:27 <ihrachys> which makes me wonder why we need a bp for that :)
14:53:28 <ihrachys> but ok
14:53:45 <ihrachys> amotoki: right. we could at least encourage new drivers to go out of tree.
14:53:55 <ihrachys> as we did in neutron while working on the split.
14:54:09 <ihrachys> but probably vpnaas team has a different view on that.
14:54:23 <ihrachys> anyhow, I don't see a problem with tracking the driver with BP.
14:54:31 <ihrachys> and that's the whole list we currently have.
14:54:43 <amotoki> correct. it is better to be discussed in the future of *aas session and clarified.
14:54:55 <ihrachys> if you have more stuff that you want to target to N, and you have time and resources to handle it, please reach drivers team or PTL.
14:55:11 <ihrachys> and that's the agenda I had for today.
14:55:14 <ihrachys> #topic Open discussion
14:55:21 <ihrachys> anything that folks would like to raise now?
14:55:49 <reedip__> hi
14:55:50 <hichihara> I saw reedip__
14:55:56 <hichihara> :)
14:56:10 <reedip__> I found one bug of interest
14:56:11 <reedip__> https://bugs.launchpad.net/neutron/+bug/1568202
14:56:14 <openstack> Launchpad bug 1568202 in kolla mitaka "linux bridge agent gets out of sync with plugin" [Critical,Triaged]
14:56:22 <reedip__> which I thought might be important
14:56:32 <ihrachys> reedip__: hi. you served as a bug deputy right? please sync with blogan who is serving on this week.
14:56:40 <reedip__> Yes, I will
14:56:41 <ihrachys> reedip__: hm, this bug seems kolla only
14:56:48 <ihrachys> reedip__: there is a patch for kolla
14:56:59 <ihrachys> they just forgot to bundle brtcl into their containers
14:57:08 <ihrachys> so I don't see how it's neutron
14:57:32 <ihrachys> I remove neutron from the bug
14:58:06 <hichihara> ihrachys: +1
14:58:08 <ihrachys> ok, let's call it a day.
14:58:08 <ihrachys> #endmeeting