14:05:39 #startmeeting neutron_l3 14:05:41 Meeting started Wed Oct 2 14:05:39 2019 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:05:42 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:05:44 The meeting name has been set to 'neutron_l3' 14:05:55 #chair slaweq 14:05:56 Current chairs: haleyb slaweq 14:06:21 hi 14:06:39 I think liuyulong is on PTO this week 14:06:40 slaweq: i added you as chair in case i disappear, am in another meeting at the same time 14:06:46 haleyb: sure 14:06:49 #topic Announcements 14:08:55 Train RC1 is out, if there are any critical things needed in it raise it with the PTL 14:09:32 #link https://etherpad.openstack.org/p/Shanghai-Neutron-Planning 14:09:51 add any items/comments there for the PTG 14:10:19 any other announcements? 14:10:41 next week is final week for rc2 if we would need anything 14:10:51 just FYI :) 14:11:22 ack, i don't think there are any L3 critical things to get in it 14:12:12 #topic Bugs 14:12:17 me neighter but I'm saying just in case something would jump in last minute 14:13:09 i know there was a new bug mlavalle wanted the team to look at, let me find it 14:14:26 https://bugs.launchpad.net/neutron/+bug/1845360 14:14:27 Launchpad bug 1845360 in neutron "ARP advertisement issue with DVR" [Undecided,New] 14:15:23 but I don't understand this 14:15:29 i did not confirm it, but theoretically i think it can happen? 14:15:39 the 2nd port is a new one 14:16:10 and, if this 2nd port has another FIP, this situation should not happen 14:16:35 ah ok, now I understand 14:16:37 ralonsoh: i think there is just one port, it's just unbound initially 14:16:38 haleyb: maybe stupid question, but why FIP needs to be configured in snat namespace for unbound port? 14:16:50 that's the point 14:16:53 I don't think so 14:17:11 in DVR, the FIP is "assigned" in the DVR router 14:17:24 but just when you have a port bound in this compute node 14:17:41 ("assigned" actually only sending GARP) 14:18:10 slaweq: it's a good question, and probably what the bug is, wonder if it even happens in a legacy router 14:18:45 in legacy router you have the FIP in only the network node 14:18:56 haleyb: in legacy routers, there is no garps send when FIP is configured 14:18:56 regardless of the port the FIP is assigned 14:18:59 right? 14:19:04 we do 14:19:09 ralonsoh: right, but should we configure the FIP if the port isn't bound? 14:19:17 but the FIP is assigned to the GW port in the router 14:19:31 ralonsoh: ok, that's tru 14:19:34 in DVR the FIP is nowhere but we send the GARP 14:19:35 *true 14:19:44 the point is: 14:19:46 in DVR 14:20:05 in this situation, the FIP is assigned to the SNAT? 14:20:07 I dont; think so 14:21:43 does anyone want to take the bug and confirm it? 14:22:05 not now, sorry 14:22:08 maybe next week 14:22:33 I also don't have cycles this week probably 14:22:38 I can take a look 14:22:52 thx tidwellr :) 14:22:57 tidwellr: great! 14:23:09 and hi btw :) 14:23:18 * tidwellr slinks in to the meeting 14:24:21 onto the next bug... 14:24:25 https://bugs.launchpad.net/neutron/+bug/1845557 14:24:25 Launchpad bug 1845557 in neutron "DVR: FWaaS rules created for a router after the FIP and VM created, not applied to routers rfp port on router-update" [High,Confirmed] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 14:24:41 https://review.opendev.org/#/c/686029/ was proposed 14:25:09 this is a rocky-only change 14:26:05 https://bugs.launchpad.net/neutron/+bug/1844915 14:26:05 Launchpad bug 1844915 in neutron "Duplicate packets with two networks connected by router" [Undecided,Incomplete] 14:26:40 this is older than last week 14:27:55 there's a lot of info and commands there 14:28:22 i'm fine leaving in the 'backlog' until someone has time to look into it 14:29:49 https://bugs.launchpad.net/neutron/+bug/1846198 14:29:49 Launchpad bug 1846198 in neutron " packet loss during active L3 HA agent restart" [Undecided,Incomplete] 14:30:06 both slaweq and i have added comments, but could not reproduce 14:31:31 looks like submitter will continue to investigate 14:32:08 any other bugs? 14:32:43 or patches that need review? 14:33:19 not from me 14:33:23 #topic Open Discussion 14:33:24 no thanks 14:33:52 any other topics? or we call it early 14:34:31 +1 to finish earlier 14:35:13 ok, thanks for coming everyone, have a good week! 14:35:17 #endmeeting