15:00:08 #startmeeting neutron_l3 15:00:10 Meeting started Thu Oct 5 15:00:08 2017 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:11 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:13 The meeting name has been set to 'neutron_l3' 15:00:19 Hey 15:00:37 hi 15:00:49 Hi 15:01:47 Swami: ping 15:02:29 hi 15:02:33 #topic Announcements 15:02:33 hi 15:03:03 The Q-1 milestone is almost here 15:03:22 October 16 -20 is my old memory still serves me ;-0 15:04:12 We have the Sydney summit also coming soon after, November 6th - 8th 15:04:35 I'll be arriving Sunday 5th in the morning and flying back Thursday 9th 15:05:08 mlavalle: I have to be in another meeting. So my responses will be delayed. 15:05:28 mlavalle: I will be arriving in Sydney earlier on Nov2nd. 15:05:40 And finally I want to thank all the team members who volunteered to be bug deputies 15:05:43 I am yet to get my visa approval. 15:06:07 with your great response, each one of us will be on duty only once in almost 4 months 15:06:35 any other announcements from the team? 15:07:05 ok, moving on 15:07:11 #topic Bugs 15:07:25 mlavalle: thanks 15:07:55 Let me start with the dvr bugs 15:08:10 I have not seen any new bugs pop for DVR this week. 15:08:17 \o/ 15:08:21 But let us discuss the status of the existing bugs. 15:08:35 mlavalle: haleyb: thanks for your help in getting the backlog bugs reviewed. 15:08:46 I've tried to dutifully review all the relatded patchses 15:09:14 #link https://bugs.launchpad.net/neutron/+bug/1717597 15:09:15 Launchpad bug 1717597 in neutron "Bad arping call in DVR centralized floating IP code" [High,In progress] - Assigned to Brian Haley (brian-haley) 15:09:38 #link https://review.openstack.org/#/c/504252/ 15:09:47 This patch is not lucky enough to clear the gate. 15:09:49 i was just taking a look at that review and why it got the -2 in the gate 15:09:56 It always fails with the volume test. 15:10:18 It has been there for more than 2 weeks now. 15:10:34 recheck... 15:10:54 well, belive it or not, this is progress 15:11:00 Let us wait for the recheck. 15:11:18 at least we are not being stopped my the move to zuulv3 15:11:30 this volume issue I think is an old friend 15:11:46 not only for DVR patches but for Neutron in general 15:11:46 mlavalle: yes 15:11:57 mlavalle: for a while we had this test skipped. 15:12:09 moving on 15:12:12 #link https://bugs.launchpad.net/neutron/+bug/1716829 15:12:13 Launchpad bug 1716829 in neutron "Centralized floatingips not configured right with DVR and HA" [High,In progress] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 15:12:36 #link https://review.openstack.org/#/c/503530/ 15:12:52 This patch was all ready to merge and had a merge conflict, so I did a rebase yesterday. 15:13:09 But it seems that this also hits the volume error. I need to recheck again. 15:13:49 The next one in the list is 15:13:51 #link https://bugs.launchpad.net/neutron/+bug/1718788 15:13:52 Launchpad bug 1718788 in neutron "DVR: Migrate centralized unbound floatingip to the respective host when the port is bound" [High,In progress] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 15:14:34 #link https://review.openstack.org/#/c/505324/14 15:14:52 This also failed zuul on rebase. So I will run a recheck on it. 15:15:16 yeah, that is zuulv3 15:15:40 mlavalle: great zuulv3 :) 15:15:44 The next one is 15:15:44 we need to run it through our trusty zuulv2 15:15:54 #link https://bugs.launchpad.net/neutron/+bug/1718345 15:15:55 Launchpad bug 1718345 in neutron "ml2_distributed_port_bindings not cleared after migration from DVR" [Medium,In progress] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 15:16:10 I have a patch for review 15:16:20 #link https://review.openstack.org/#/c/508808/ 15:16:48 This is still work in progress. I got some reviews from haleyb and anilventa on this patch. 15:16:48 So it is moving on. 15:17:00 I might have to add another functional test to validate the migration and ports on it. 15:17:59 #link https://bugs.launchpad.net/neutron/+bug/1717302 15:18:00 Launchpad bug 1717302 in neutron "Tempest floatingip scenario tests failing on DVR Multinode setup with HA" [High,Confirmed] 15:18:21 This is the one bug, that i have not found out the root cause. I am still investigating this bug. 15:18:35 Once I have some handle on it or able to reproduce locally I will push in a patch for it. 15:19:01 #link https://bugs.launchpad.net/neutron/+bug/1716401 15:19:02 Launchpad bug 1716401 in neutron "FWaaS: Ip tables rules do not get updated in case of distributed virtual routers (DVR)" [Undecided,New] - Assigned to Reedip (reedip-banerjee) 15:19:29 mlavalle: as we discussed last about this issue, I have not got any message or update back from reedip on this bug. 15:19:44 So might be better to bring it up in the FWaaS meeting next week. 15:19:53 I will do that 15:20:27 HoloIRCUser^^^^ 15:20:46 mlavalle: thanks 15:20:47 reedip was under this nick in the fwaas channel earliertoday 15:20:59 but I think he is gone 15:21:09 I will follow up in the fwaas meeting 15:21:16 mlavalle: thanks that would help. 15:21:20 which now is right before this one 15:21:48 I have one patch that I am working on for the Fullstack testing. That is still work in progress. 15:22:19 #link https://bugs.launchpad.net/neutron/+bug/1716401 15:22:20 Launchpad bug 1716401 in neutron "FWaaS: Ip tables rules do not get updated in case of distributed virtual routers (DVR)" [Undecided,New] - Assigned to Reedip (reedip-banerjee) 15:22:30 Any early reviews on this patch would help. 15:22:44 mlavalle: that's all I had for today's bug discussion 15:22:49 mlavalle: back to you 15:22:59 Swami: thanks for the update 15:23:36 earlier today I spent time going through bugs 15:24:15 I closed one anssigned to janzian 15:24:22 no hist for a long time 15:24:26 hits 15:24:33 mlavalle: thanks 15:25:01 and I'll be working on this one https://bugs.launchpad.net/neutron/+bug/1610483 15:25:02 Launchpad bug 1610483 in neutron "Pluggable IPAM rollback mechanism is not robust" [High,In progress] - Assigned to Aliaksandr Dziarkach (aliaksandr-dziarkach) 15:25:42 haleyb and I made an agreement with kevinbenton about removing the rollback mechanism 15:25:48 I'll follow up with this 15:26:03 any other bugs? 15:26:30 ok, moving on 15:26:43 #topic Multiple port bindings 15:27:03 I would like to add to our weekly agenda a topic on multiple port bindings 15:27:42 mlavalle: i have just reuploaded the spec on the nova side https://review.openstack.org/375580 with the changes you requested 15:27:49 eventhough it doesn't fit neatly in "L3", it may have impact on the work that Swami did coordinating with nova 15:27:59 in the case of migrations 15:28:19 mlavalle: i will take a look at the spec and see if there is any impact on our existing code path. 15:29:09 Swami: yeah, let's take a few minutes every Thursday to coordinate status with sean-k-mooney and ralonsoh, in case we need to fix something 15:29:35 sean-k-mooney: perfect, I will take a look again and hopefully +1 it 15:29:58 mlavalle: sure makes sense 15:30:29 in setting up a dev enviornment i belive there may be a bug in the ovs l2 agent currently. where vxlan tunnels between ovs and linux bridge are nolonger created correctly 15:30:54 on my side I am already re-working the code in the service plugin that was coded previouly and moving it to the ML2 15:31:25 I should push something, at least WIP, by the end of the weekend 15:32:19 anything else to say about this? 15:32:32 mlavalle: is there a reiview up currently for that that i should follow? or will you be pushing with a new change id? 15:32:50 sean-k-mooney: there is a review and I will reuse it.... hang on 15:33:34 #link https://review.openstack.org/#/c/414251/ 15:34:37 ok, moving on 15:35:17 davidsha: do you want to talk today about flows based L3 routers? 15:36:11 Sure, I have another patch ready to go up, I'm just looking into developing some tests for the patch at the moment. 15:36:25 #topic flows based L3 routers 15:36:59 I've taken Swami's input and broken the dvr_local_bridge_router.py into smaller mixins to make it easier to review. 15:37:11 for those interested, this is the patch 15:37:19 #link https://review.openstack.org/#/c/472289/ 15:37:41 yup, that's it 15:38:18 davidsha: thanks that would help. 15:38:32 is the jenkins failure due to the code in the patch? 15:39:15 and: you haven't pushed the latest version, right davidsha? 15:39:32 I don't believe so, but I'll look into it. 15:39:58 maybe push the latest code and see if we pass jenkins 15:40:20 mlavalle: ack 15:40:21 let's save some of your brin ycle 15:40:33 brain cycles^^^ 15:40:39 always appreciated :D 15:41:25 I am still working on getting Floating IP working as well. 15:41:26 im not sure why the dvr multinode tests are marked as failed http://logs.openstack.org/89/472289/4/check/gate-grenade-dsvm-neutron-dvr-multinode-ubuntu-xenial/4dae1e1/logs/testr_results.html.gz 15:41:34 looks like tempest passed 15:42:30 anything else on this topic? 15:43:02 Not for the moment, I'll get the next patch set up for it asap 15:43:33 thanks for working so hard on this, davidsha. it is a nice new option :-) 15:43:46 #topic Open Agenda 15:44:01 any other topics we should discuss today? 15:44:42 Just on testing 15:45:09 The L3 tests are all going to be moved to Fullstack tests to lower the number of jobs is it? 15:45:41 Or am I mistaken? 15:45:51 No I don't think so 15:46:05 is that from the PTG summary? 15:46:24 No, I thought I'd read it discussed somewhere. 15:46:36 no, we haven't said that 15:46:43 we like our Tempest 15:46:50 right Swami? ^^^^ 15:47:09 *Silence* 15:47:17 mlavalle: yes that's my understanding 15:47:36 fwiw.... he was distracted LOL 15:47:58 mlavalle: the reason is since we don't have enough resources in the gate to provide more nodes, it was decided for multinode or multi-agent configuration we should be using the fullstack. 15:48:03 Ok, thanks for clearing that up1 15:48:12 mlavalle: sorry I am in a different meeting as well. 15:48:17 I know 15:48:28 thanks for answering 15:49:10 anything else? 15:49:23 davidsha: so i think the point you were raising was that instead of having another tempest job configuration for dvr with openflow that would be tested via fullstack is that correct 15:50:24 sean-k-mooney: yes that is the intent. Since davidsha's work would be a different agent_type=dvr_bridge, it would be great to use the fullstack. 15:50:36 ++ 15:51:19 we chould have an experimental job for this also but while its an experimental feature i dont think it need to be andded to the standard integration set 15:52:03 I think this is what we said in the drivers meeting a few weeks ago 15:52:51 yep i think fullstack is a good fit for this also 15:53:00 cool 15:53:47 ok team, thanks for attending 15:53:59 Thanks! 15:54:02 see you next week, same time same channel 15:54:11 #endmeeting