15:00:12 <mlavalle> #startmeeting neutron_l3 15:00:12 <openstack> Meeting started Thu Sep 8 15:00:12 2016 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:16 <openstack> The meeting name has been set to 'neutron_l3' 15:00:17 <jschwarz> \o/ 15:00:20 <njohnston> o/ 15:00:20 <carl_baldwin> o/ 15:00:27 <tidwellr> hi 15:00:54 <mlavalle> #chair carl_baldwin tidwellr 15:00:55 <openstack> Current chairs: carl_baldwin mlavalle tidwellr 15:01:06 <johnbelamaric> hello 15:01:28 <mlavalle> Today's agenda is here: 15:01:28 <haleyb> hi 15:01:36 <mlavalle> #link https://etherpad.openstack.org/p/neutron-l3-subteam 15:02:00 <mlavalle> #topic Announcements 15:02:38 <mlavalle> As we all know, RC-1 deadline is next week: 15:02:51 <mlavalle> #link https://releases.openstack.org/newton/schedule.html 15:03:36 <mlavalle> Please also note that the PTL self nomination period is also next week, in case anyone is interested 15:04:02 <carl_baldwin> mlavalle: maybe you? ;) 15:04:14 <mlavalle> nope :-) 15:04:35 <mlavalle> Any other annoucements? 15:05:04 <carl_baldwin> Not from me 15:05:19 <mlavalle> ok, let's dive in.... 15:05:26 <mlavalle> #topic Bugs 15:05:37 <mlavalle> First in the list is https://bugs.launchpad.net/neutron/+bug/1609540 15:05:39 <openstack> Launchpad bug 1609540 in neutron "Deleting csnat port fails due to no fixed ips" [High,In progress] - Assigned to Carl Baldwin (carl-baldwin) 15:06:05 <carl_baldwin> So, we won't see the message unless we bump up the level 15:06:10 <mlavalle> As noted in the bug itself, I checked with infra, and debug level information is not captured in logstash 15:06:17 <carl_baldwin> I'll post a patch for that today. 15:06:41 <carl_baldwin> mlavalle: what level do they index to search? 15:07:05 <clarkb> info and greater 15:07:19 <carl_baldwin> clarkb: thanks 15:07:32 <carl_baldwin> I'll bump it to info 15:07:47 <mlavalle> Thanks carl_baldwin 15:07:50 <carl_baldwin> mlavalle: that's all 15:08:18 <mlavalle> Next one is https://bugs.launchpad.net/neutron/+bug/1596075 15:08:19 <openstack> Launchpad bug 1596075 in neutron "Neutron confused about overlapping subnet creation" [High,In progress] - Assigned to Kevin Benton (kevinbenton) 15:08:20 <clarkb> for a couple reasons. The big one is we can't index all the data at debug level. There is just too much. But also if something is important enough to be indexed that probably means its more than debug 15:08:47 <mlavalle> clarkb: thanks for the clarification! 15:08:53 <carl_baldwin> clarkb: understood. It makes perfect sense. 15:09:34 <mlavalle> regarding the bug, I can see 2 fixes were proposed to it 15:09:41 <mlavalle> over the past 6 hours 15:10:08 <mlavalle> thanks kevinbenton! 15:10:41 <mlavalle> So I guess we need to wait for the fixes to be deployed by the bug reporter ans let us know what the status is 15:10:54 <mlavalle> any other comments? 15:11:42 <mlavalle> ok, let's move on 15:11:52 <mlavalle> Next one is https://bugs.launchpad.net/neutron/+bug/1599329 15:11:53 <openstack> Launchpad bug 1599329 in neutron "Potential regression on handing over DHCP addresses to VMs" [High,In progress] - Assigned to Miguel Lavalle (minsel) 15:12:15 <mlavalle> I spent some time yesterday with this one 15:12:34 <mlavalle> as of last evening, I could see 28 ocurrences over the previous 7 days 15:13:14 <mlavalle> I can see instances not receiving leases from dhcp 15:13:43 <mlavalle> But I couldn't connect those failures with specific events in the dhcp agent log 15:14:05 <mlavalle> So I will continue digging over the next few days 15:14:17 <mlavalle> Any comments, advice or guidance? 15:15:57 <mlavalle> ok, let's move on 15:16:04 <mlavalle> https://bugs.launchpad.net/neutron/+bug/1610483 15:16:05 <openstack> Launchpad bug 1610483 in neutron "Pluggable IPAM rollback mechanism is not robust" [High,Confirmed] - Assigned to Aliaksandr Dziarkach (aliaksandr-dziarkach) 15:16:28 <mlavalle> johnbelamaric: any comments on this one? 15:17:18 <johnbelamaric> mlavalle: no update right now, I will talk to him in 45 mins. 15:17:43 <mlavalle> johnbelamaric: Thanks for taking ownership of this one! 15:18:14 <mlavalle> Next one is https://bugs.launchpad.net/neutron/+bug/1619312 15:18:16 <openstack> Launchpad bug 1619312 in neutron "dvr: can't migrate legacy router to DVR" [High,Confirmed] 15:18:28 <jschwarz> this is still looking for an owner :< 15:18:38 <mlavalle> yeap, how about you? 15:19:11 <jschwarz> mlavalle, I probably won't get to it in the next few weeks, so I'd rather not 15:19:19 <jschwarz> I have work travels coming up 15:19:27 <mlavalle> Thanks 15:19:41 <mlavalle> Any other comments from anyone? 15:20:01 <haleyb> mlavalle: i guess when dvr things fall through the cracks they go to me 15:20:38 <haleyb> so i'll take it over 15:20:42 <mlavalle> haleyb: thanks! 15:21:50 <alraddarla_> I joined late and I am new so I apologize if this is not where I should be asking...Do you all know if neutron has the ability for to use the l3 scheduler to pin a tenant to a particular l3 router? 15:22:50 <jschwarz> alraddarla_, I think you can use the AZScheduler for that. Lets take it to #openstack-neutron ? 15:23:12 <mlavalle> Last bug in the list today is: https://bugs.launchpad.net/neutron/+bug/1522980 15:23:13 <openstack> Launchpad bug 1522980 in neutron "L3 HA integration with l2pop assumes control plane is operational for fail over" [High,In progress] - Assigned to venkata anil (anil-venkata) 15:23:38 <jschwarz> mlavalle, Anil is on PTO I believe so that's why there hasn't been progress on that during the last few days 15:23:46 <alraddarla_> Sounds good. I will ask there. Thank you jschwarz 15:23:54 <jschwarz> I'm not sure when he's due back 15:24:31 <mlavalle> jschwarz: thanks. Yeah, carl_baldwin left some comments in the patchset. So let's wait for him to come back 15:24:53 <jschwarz> we'd still like it to go in for RC1 though :) 15:25:23 <mlavalle> Any other bugs from the team? 15:25:51 <carl_baldwin> mlavalle: jschwarz: I'll take a look. 15:25:58 <jschwarz> carl_baldwin, thanks 15:26:35 <mlavalle> If not, let's move on 15:26:47 <mlavalle> #topic Routed Networks 15:27:06 <mlavalle> carl_baldwin: any updates for this team? 15:27:44 <carl_baldwin> I haven't done much, to be honest. We plan to talk about plans for Ocata next Tuesday. 15:27:57 <carl_baldwin> I am working on docs, but very slowly. 15:28:12 <mlavalle> ok, that's good, let's move on 15:28:28 <carl_baldwin> I got a review from john-davidge on the docs patch. More input would probably be helpful. 15:29:04 <carl_baldwin> #link https://review.openstack.org/#/c/356013/ 15:29:44 <mlavalle> Thanks 15:29:48 <mlavalle> moving on 15:30:01 <mlavalle> #topic BGP Dynamic Routing 15:30:06 <tidwellr> hi 15:30:58 <tidwellr> no updates this week, we're ready to release Newton 15:31:03 <liuyulong> hi 15:31:34 <liuyulong> tidwellr, https://bugs.launchpad.net/neutron/+bug/1614866, this bug 15:31:36 <openstack> Launchpad bug 1614866 in neutron "BGP: Failed to withdraw route while dissocating floating IP" [High,In progress] - Assigned to LIU Yulong (dragon889) 15:31:44 <tidwellr> liuyulong: I know you've got some bug fixes outstanding, I'm trying to work in reviews, my apologies 15:32:16 <liuyulong> tidwellr, thank you very much. 15:32:51 <tidwellr> liuyulong: I would love to get those fixes in, so I'll spend some time today 15:33:07 <tidwellr> but if we can't get them in for Newton, I think we'll still be OK 15:33:20 <carl_baldwin> Could I help with reviews? 15:33:44 <liuyulong> carl_baldwin, yes, please 15:33:49 <liuyulong> tidwellr, OK 15:33:57 <tidwellr> carl_baldwin: that would be great, I'm drowning with other things 15:33:59 <carl_baldwin> Do you have links handy? 15:34:12 <liuyulong> tidwellr, mitaka may need that fix. 15:34:21 <tidwellr> https://review.openstack.org/#/c/357716/ 15:34:45 <liuyulong> ^ 15:35:13 <liuyulong> carl_baldwin, and neutron side: https://review.openstack.org/#/c/357722/ 15:35:14 <tidwellr> carl_baldwin: this is the neutron-dynamic-routing fix, the issue is that a FIP withdrawl can sometimes take a whole re-sync cycle because of a bug 15:35:50 <carl_baldwin> liuyulong: thanks 15:35:51 <tidwellr> carl_baldwin: we'd like to not hit this bug and have the FIP route withdrawn immediately rather than waiting for the re-sync 15:36:17 <carl_baldwin> tidwellr: ack 15:36:36 <tidwellr> liuyulong: I'm not sure if this is a backportable change, but it's worth a look 15:36:54 <tidwellr> it involves the RPC interface, so we need to tread lightly with the backport 15:37:14 <carl_baldwin> yep 15:38:07 <tidwellr> liuyulong: thanks for being persistent on this fix, anything else to add? 15:38:42 <tidwellr> if not, I don't have anything else to add 15:39:40 <mlavalle> Thanks tidwellr 15:39:47 <mlavalle> Let's move on, then 15:39:54 <mlavalle> #topic FWaaS 15:40:03 <mlavalle> njohnston: you are up 15:40:22 <njohnston> FWaaS is now using L3 agent extensions, delivered in Newton 15:40:26 <njohnston> #link https://review.openstack.org/#/c/355576/ 15:40:35 <njohnston> heavily tested manually 15:41:06 <njohnston> so I think the main remaining bit of work is just the L3 agent extensions fullstack test 15:41:08 <njohnston> #link https://review.openstack.org/#/c/345631/ 15:41:41 <njohnston> which involces faking out stevedore to load a dummy plugin. I don't know how to do this, and I'm still looking for the best way to do it, reading Stevedore's own tests 15:42:14 <njohnston> I also figured I'd reach out to the other *-aas teams to ask them about them integrating with L3 agent extensions 15:42:39 <njohnston> to advocate for it with them, and provide limited assistance with it 15:42:50 <liuyulong> tidwellr, I have a bad networking at home, sorry for late reply. Only needs pay more attention is that cache check change in dr-agent side. 15:43:19 <liuyulong> https://review.openstack.org/#/c/357716/6/neutron_dynamic_routing/services/bgp/agent/bgp_dragent.py 15:43:23 <liuyulong> tidwellr, ^ 15:43:31 <tidwellr> liuyulong: ack 15:44:18 <njohnston> I think that's it for my update. But I'd be interested: which *-aas type teams do you think I should approach to see if I can help them transition to L3 agent extensions? VPNaaS? 15:44:44 <mlavalle> njohnston: my bet would be the LBaaS team 15:45:14 <mlavalle> start with blogan 15:45:22 <njohnston> mlavalle: Will do, thanks! 15:46:25 <mlavalle> ok, moving on 15:46:43 <mlavalle> #topic Conversion to pluggable IPAM 15:46:58 <mlavalle> This topic is mostly done, AFAIK 15:47:16 <carl_baldwin> Yes. 15:47:20 <mlavalle> The only pending point is updating docs 15:47:28 <carl_baldwin> We have an open doc bug. 15:47:36 <mlavalle> any comments on this, johnbelamaric? 15:47:39 <carl_baldwin> #link https://bugs.launchpad.net/neutron/+bug/1588984 15:47:40 <openstack> Launchpad bug 1588984 in openstack-manuals "Improve IPAM documentation" [Medium,Triaged] 15:47:46 <carl_baldwin> I was just discussing with johnbelamaric 15:48:10 <johnbelamaric> mlavalle: one change for this has merged, but I need to do a release note too 15:48:30 <johnbelamaric> mlavalle: also, I am thinking we may want to include a "Known Drivers" section with links to those drivers 15:49:12 <johnbelamaric> mlavalle: AFAIK, the drivers out there are Infoblox, Calico, Romana. Anyone know of any others? 15:49:12 <mlavalle> johnbelamaric: so I'll keep this topic in the agenda, so we can revisit in future meetings 15:49:30 <carl_baldwin> johnbelamaric: we might want to ask Sam-I-Am about what happened with the docs and what to do. 15:49:47 <johnbelamaric> carl_baldwin: ok, thanks 15:49:54 <carl_baldwin> johnbelamaric: I don't know of any others. 15:50:33 <johnbelamaric> k 15:51:04 <mlavalle> johnbelamaric: thanks for the update! 15:51:37 <mlavalle> #topic Open Agenda 15:51:58 <mlavalle> Any other topics that anyone in the team wants to bring up? 15:53:06 <mlavalle> Ok, I think this is it 15:53:25 <mlavalle> Thanks for your attendance and updates. Have a nice rest of the week! 15:53:33 <mlavalle> #endmeeting