15:01:02 #startmeeting neutron_l3 15:01:03 Meeting started Thu Jul 21 15:01:02 2016 UTC and is due to finish in 60 minutes. The chair is tidwellr. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:07 The meeting name has been set to 'neutron_l3' 15:01:34 #chair carl_baldwin mlavalle 15:01:35 Current chairs: carl_baldwin mlavalle tidwellr 15:01:44 o/ 15:01:58 #topic Annpuncements 15:02:32 neutron mid-cycle is coming up 15:03:42 any more annoncements? 15:04:32 if not, let's dive in 15:05:00 #topic Bugs 15:05:20 working from the bottom of the list 15:05:47 oh, by the way the agenda is here https://etherpad.openstack.org/p/neutron-l3-subteam 15:06:02 https://bugs.launchpad.net/neutron/+bug/1603162 15:06:02 Launchpad bug 1603162 in neutron "IP deallocation failed on external system with pluggable IPAM" [High,New] - Assigned to Pavel Bondar (pasha117) 15:06:58 hi 15:06:58 pavel_bondar: any progress? 15:06:59 pavel_bondar carl_baldwin: This looks new, any updates? 15:07:15 not much progress yet on my side 15:07:17 yeah, pavel_bondar filed it a few minutes ago 15:07:37 mlavalle: that is another one 15:07:42 ok 15:07:52 one that I filed is: https://bugs.launchpad.net/neutron/+bug/1605277 15:07:52 Launchpad bug 1605277 in neutron "[IPAM] 'Internal' ipam driver does not allow to delete all pools on subnet update" [Undecided,In progress] - Assigned to Pavel Bondar (pasha117) 15:08:29 as for 1603162, no progress yet, but I will pick it as my next task 15:08:59 pavel_bondar: let me know if you want help 15:09:54 pavel_bondar: thanks for the update on these 15:10:05 carl_baldwin: sure, as I got for now this ticket is about writing UT to cover rollback case and emulating db transaction breakage 15:10:12 is it correct? 15:10:19 pavel_bondar: yes 15:10:43 carl_baldwin: ok, will try to do soon 15:11:25 that is it from my side about this ticket 15:11:47 https://bugs.launchpad.net/neutron/+bug/1600530 15:11:47 Launchpad bug 1600530 in neutron "Part 2 of engine DB facade breaks metadata agent" [High,In progress] - Assigned to Henry Gessau (gessau) 15:13:04 looks like a fix is already in 15:13:07 I think that one should be closed now 15:13:18 ok 15:13:36 I'll move it to fix released 15:14:01 mlavalle: thanks, I moved it off the agenda 15:14:13 https://bugs.launchpad.net/neutron/+bug/1599329 15:14:13 Launchpad bug 1599329 in neutron "Potential regression on handing over DHCP addresses to VMs" [High,In progress] - Assigned to Kevin Benton (kevinbenton) 15:15:32 looks like kevinbenton and ajo are on this one 15:15:55 I haven't low at the proposed fix yet. 15:16:55 well, it seems more like they are researching it 15:16:55 *looked 15:17:08 kevinbenton proposed a fix but was then abandoned 15:17:09 mlavalle: could be 15:17:41 I see https://review.openstack.org/#/c/344545/, but it has a sticky -2 while someone does their due diligence 15:17:43 and ajo reported some research results just 15 minutes ago 15:20:25 ok, anything else to add? 15:20:26 if not we can move on and take an action item to follow this one 15:20:27 https://bugs.launchpad.net/neutron/+bug/1596075 15:20:27 Launchpad bug 1596075 in neutron "Neutron confused about overlapping subnet creation" [High,In progress] - Assigned to Kevin Benton (kevinbenton) 15:21:31 looks like we have a fix for this one 15:22:18 yeah, but kevinbenton is asking the reported to do some things, right above the merge 15:22:45 so I say let's keep as is and keep an eye on it 15:22:54 the fix might be partial 15:23:19 ah, partial fix 15:23:23 I would like to hear back from the reporter 15:23:40 definitely 15:23:56 keep watching this one then 15:25:14 ok, I think we've worked through the bugs. Are there any I missed that need to be discussed? 15:25:27 Yes I want to add one 15:25:37 https://bugs.launchpad.net/neutron/+bug/1604370 15:25:37 Launchpad bug 1604370 in neutron "functional: test_legacy_router_ns_rebuild is unstable" [High,New] 15:25:53 I just raised its importance to high 15:26:22 there are 17 ocurrences over the past 7 days 15:26:49 so it warrants further investigation 15:26:58 I will dig in later today 15:27:07 malvalle: thanks, I've added that to the etherpad 15:27:26 thanks 15:27:37 mlavalle: thanks for staying on top of it 15:27:44 any more bugs to discuss? 15:28:19 one that I filed today: https://bugs.launchpad.net/neutron/+bug/1605277 15:28:19 Launchpad bug 1605277 in neutron "[IPAM] 'Internal' ipam driver does not allow to delete all pools on subnet update" [Undecided,In progress] - Assigned to Pavel Bondar (pasha117) 15:28:42 this one was found during analysing test failures for migration to pluggable ipam 15:28:56 * john-davidge was just reading that one 15:29:30 I have proposed fix for it: https://review.openstack.org/#/c/345498/ 15:29:36 pavel_bondar: good find 15:29:55 I was starting to look in to that failure yesterday myself 15:30:11 Will look at your fix. 15:30:27 carl_baldwin: great, thanks 15:30:45 fix is pretty trivial by the way 15:31:11 that is probably it about it 15:31:13 pavel_bondar: thanks for the update, we like easy fixes :) 15:31:24 by all means 15:31:27 :) 15:31:48 #topic Routed Network Segments 15:32:23 carl_baldwin mlavalle: what do you have cooking at the moment? 15:32:57 I have a plan now with nova to get the deferred ip fix to merge. Discussed it with them yesterday. 15:33:06 ++ 15:33:45 It is official that we won't have ip availability aware scheduling in Newton. 15:34:11 :(, good progress was still made though 15:34:39 Yes, we've done a ton of work 15:36:45 carl_baldwin: thanks, anything else? 15:37:12 Just keep plugging away. We'll have something workable in Newton. 15:37:43 carl_baldwin: sounds good 15:38:01 #topic FWaaS 15:38:08 Hi 15:38:28 So the first part of the agent work is complete and merged: https://review.openstack.org/329701 15:38:34 thanks to all who helped with that 15:39:05 but that was just making the l2 code generalized and available to the l3 agent 15:39:30 the second step is creating the intermediate classes necessary for an l3 agent extension to run: L3 agent extension implementation: https://review.openstack.org/337699 15:39:55 I am starting to test that out now 15:40:09 I will file a follow-up patch with a fullstack test to really test it out 15:40:22 The fullstack test creates a dummy l3 agent extension and loads it and then sends some router add/update/delete messages to the agent. The extension communicates back over a TCP socket to say "I was called with X payload" so the test can verify the extension was called correctly. 15:40:51 but I don't want to hold up 337699 while I work out the complexities of that 15:41:40 njohnston: ack 15:41:45 There are a couple of people working on the actual FWaaS l3 agent extension (mfranc213 and yushiro) and that is coming along. 15:41:56 I am still optimistic that we will make Newton in good form 15:42:38 njohnston: very nice, thanks for the update. Is there any help you need at the moment? 15:43:31 No, nothing per se... I think once I can get the l3 agent extension to really work, then I think it's just left for things to start plugging into it. 15:43:54 Perhaps in a week I will be ready for reviews on 337699 15:44:19 njohnston: great 15:44:31 njohnston: anything to share with the team? 15:44:38 *else 15:44:43 Nope, that's it got me 15:44:47 s/got/for/ 15:45:04 njohnston: :) 15:45:20 topic BGP Dynamic Routing 15:45:49 ^^ #? 15:45:59 oh yeah 15:46:05 :) 15:46:07 Tidwellrm 15:46:09 https://review.openstack.org/#/c/337662/ 15:46:11 #topic BGP Dynamic Routing 15:46:28 this patch need some help review 15:46:38 Hi tidwellr 15:46:45 steve_ruan: I saw that this morning 15:47:09 it had subnetpools in the title so it caught my attention :) 15:47:09 this patch has a small change about subnetpool event 15:47:17 thanks 15:47:47 steve_ruan: I can give that a look 15:48:10 steve_ruan: how is everything else going, any updates to share? 15:48:33 still working on evpn advertisement 15:48:38 almost finished 15:48:46 will submit soon 15:49:26 steve_ruan: I see the code starting to be proposed, how is the spec doing? I haven't seen any updates for a while 15:49:50 tidwellr: I think the spec is in decent shape. Needs more reviews. 15:50:09 there is no more doubt about the spec, right? 15:51:02 mickeys: my apologies, I'm drowning here. I missed the updated patch set almost 2 weeks ago 15:51:23 #action tidwellr to look over https://review.openstack.org/#/c/322654/ 15:52:42 I was able to get the tempest tests in neutron-dynamic-routing cleaned up and running in the check jobs 15:53:22 I'm now more comfortable with code being proposed against the project now, the next step is to just make that job a voting job 15:54:06 mickeys steve_ruan: anything else to add? 15:54:15 no 15:54:19 Nothing else from me 15:54:35 #topic Open Discussion 15:55:01 The Service Subnets patch has been updated since yesterday https://review.openstack.org/#/c/337851/ 15:55:13 Thanks to everyone for the reviews 15:55:28 john-davidge: sweet, I didn't realize that was up for review 15:55:53 tidwellr: It's in pretty good shape now I think 15:56:20 Once the positive reviews start appearing I'll get started on the follow-up for the port creation logic 15:56:24 john-davidge: thanks 15:56:41 john-davidge: As I recall it was pretty close. I'll have a look today. 15:57:08 carl_baldwin: Thanks Carl, your reviews have been very helpful! 15:57:25 anything else to discuss? 15:58:34 Thanks, tidwellr ! 15:58:39 alright, let's call it a meeting then 15:58:47 thanks everyone! 15:58:54 #endmeeting