14:16:48 <haleyb> #startmeeting networking 14:16:48 <opendevmeet> Meeting started Tue Nov 11 14:16:48 2025 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:16:48 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:16:48 <opendevmeet> The meeting name has been set to 'networking' 14:16:56 <haleyb> Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, slaweq, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, haleyb, ralonsoh 14:16:57 <mlavalle> \o 14:17:06 <mtomaska> o/ 14:17:08 <lajoskatona> o/ 14:17:09 <rubasov> o/ 14:17:25 <haleyb> #announcements 14:17:47 <haleyb> We are in week R-20 of Gazpacho 14:17:59 <haleyb> Which is the G-1 milestone 14:18:07 <haleyb> #link https://releases.openstack.org/gazpacho/schedule.html 14:19:07 <haleyb> obviously we had the PTG two weeks ago, i will send out the highlights to the ML this week, have been traveling 14:19:48 <ralonsoh> hello 14:20:04 <haleyb> there was a release patch pushed for neutron 14:20:19 <haleyb> #link https://review.opendev.org/c/openstack/releases/+/966456 14:20:33 <haleyb> but we are waiting for a couple more patches to land first 14:20:51 <haleyb> one is to update requirements to os-ken 4.0.0 14:21:02 <haleyb> #link https://review.opendev.org/c/openstack/neutron/+/966613 14:21:10 <haleyb> that needs another +2 14:21:35 <haleyb> we are still waiting for the requirements bump to happen though 14:21:44 <haleyb> #link https://review.opendev.org/c/openstack/requirements/+/966536 14:21:58 <haleyb> that was blocked on a nova issue i tracked down yesterday with sean 14:22:41 <haleyb> are there any other neutron changes we should get in the G-1 release? 14:23:30 <mlavalle> is it ok to approve https://review.opendev.org/c/openstack/neutron/+/966613? 14:23:47 <ralonsoh> it is, yes 14:23:48 <haleyb> mlavalle: it should be as it depends on the requirements bump 14:24:01 <mlavalle> ok, approved 14:24:15 <haleyb> ralonsoh: i guess you can abandon https://review.opendev.org/c/openstack/neutron/+/966663 then 14:25:07 <haleyb> we will update the release tag when that merges 14:25:48 <haleyb> there were also releases proposed for other things 14:25:55 <haleyb> eg ovsdbapp 14:26:00 <haleyb> #link https://review.opendev.org/c/openstack/releases/+/966559 14:26:20 <haleyb> don't know if there are any other changes that should merge there? 14:26:29 <haleyb> #link https://review.opendev.org/q/project:openstack/ovsdbapp 14:26:51 <ralonsoh> nothing urgent, we can proceed as is now 14:27:02 <ralonsoh> well, maybe https://review.opendev.org/c/openstack/ovsdbapp/+/952871 14:27:07 <ralonsoh> this one is important 14:27:18 <haleyb> ack, think it was you and jlibosva that had things proposed, and terry 14:28:15 <haleyb> i will look after meeting or tomorrow 14:28:40 <haleyb> although that relies on https://review.opendev.org/c/openstack/ovsdbapp/+/962732/ which has a -1 14:29:45 <haleyb> i also saw there was a release for neutronclient 14:29:54 <haleyb> #link https://review.opendev.org/c/openstack/releases/+/966575 14:30:25 <haleyb> i will have to check if that is the hash that has the tapaas revert 14:30:40 <lajoskatona> I will check these also 14:30:43 <ralonsoh> no 14:30:54 <ralonsoh> https://review.opendev.org/c/openstack/python-neutronclient/+/966356 14:30:59 <ralonsoh> still under review 14:31:25 <ralonsoh> but we didn't merge the OSC patch yet 14:31:37 <lajoskatona> yes there's a -1 on the osc patch 14:31:38 <haleyb> ah, ok 14:33:00 <haleyb> i guess that's all for release patches then 14:33:33 <haleyb> Reminder: If you have a topic for the drivers meeting on Friday, please add it to the wiki @ https://wiki.openstack.org/wiki/Meetings/NeutronDrivers 14:34:05 <haleyb> i cannot attend the meeting this friday, and not sure if there's anything new, so will propose canceling 14:34:20 <mlavalle> ack 14:35:20 <haleyb> And continue to use the priorities dashboard for important changes or things ready to merge 14:35:34 <haleyb> i have a few easy ones there that could use another +2 14:35:41 <haleyb> https://tinyurl.com/59z278km 14:35:59 <haleyb> i can't put a shortener in the wiki, but that's the dashboard 14:36:04 <lajoskatona> cool 14:36:44 <haleyb> that was all the neutron announcements i had, any others? 14:37:31 <haleyb> i will note that i am off today (US holiday) and Friday, hopefully my irc bot keeps my status this time 14:38:07 <haleyb> but i might do a few reviews later 14:38:33 <haleyb> ok, moving on 14:38:36 <haleyb> #topic bugs 14:39:02 <haleyb> without a meeting last week, people might have gotten out of sync 14:39:23 <haleyb> last week was supposed to be otherwiseguy, but i didn't see an email 14:39:44 <haleyb> i did see one from yatin for the week before that we can go through 14:40:05 <haleyb> and this week the deputy is mtomaska, next week bcafarel - hopefully that work for both 14:40:47 <haleyb> we can go through the unassigned ones 14:40:49 <mtomaska> ACK 14:41:06 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2130380 14:41:07 <haleyb> OOM called from keepalived, "neutron-tempest-plugin-openvswitch" job 14:41:50 <ralonsoh> I saw it several times last week 14:42:31 <ralonsoh> and everytime it was related to the keepalived process 14:43:30 <haleyb> ack, would the patch you proposed from the PTG topic about memory usage help here? 14:43:38 * haleyb doesn't have the link handy 14:43:56 <ralonsoh> haleyb, yes, I have this in the on-demand section 14:43:59 <ralonsoh> we can talk later 14:44:05 <haleyb> ack 14:44:26 <haleyb> the other unassigned from 2 weeks ago is 14:44:29 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2129959 14:44:35 <haleyb> Semantic violated with bulk creation of security groups 14:45:02 <ralonsoh> I will try to check it this week 14:45:16 <haleyb> ack, thanks 14:45:36 <haleyb> i can look quickly at new ones from last week 14:46:12 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2130644 14:46:18 <haleyb> [neutron-vpnaas] unit and functional tests stuck and job TIMEOUTs 14:46:24 <ralonsoh> there is a patch proposed 14:46:42 <ralonsoh> https://review.opendev.org/c/openstack/neutron-vpnaas/+/966090 14:46:55 <ralonsoh> I'll update the bug 14:48:05 <haleyb> right, the closes-bug number is a little off :) 14:48:29 <ralonsoh> ups... 14:48:33 <ralonsoh> I'll update the patch now 14:48:50 <opendevreview> Rodolfo Alonso proposed openstack/neutron-vpnaas master: [eventlet-removal] Fix unit and functional tests https://review.opendev.org/c/openstack/neutron-vpnaas/+/966090 14:49:00 <haleyb> next bug is 14:49:04 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2130885 14:49:13 <haleyb> neutron-dhcp-agent ip conflict in qdhcp namespace causing dns timeouts 14:50:26 <ralonsoh> I remember commenting this issue in IRC 14:50:38 <ralonsoh> the problem is I don't know how do you get into this state 14:51:04 <haleyb> that will need more investigation obviously, i don't know either 14:51:55 <haleyb> i think that was all the unassigned from last week 14:52:01 <haleyb> any other bugs to discuss? 14:52:30 <ralonsoh> yes 14:52:36 <ralonsoh> the n-d-r one 14:52:58 <ralonsoh> https://bugs.launchpad.net/neutron/+bug/2130631 14:53:17 <ralonsoh> this is because we removed the eventlet utils from Neutron 14:53:28 <ralonsoh> I'm waiting for the os-ken 4.0.0 release 14:53:36 <ralonsoh> maybe the dr agent will work find with it 14:53:50 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/956747 14:53:58 <haleyb> will this be fixed when we release neutron for G-1 ? 14:54:17 <ralonsoh> but is that is not the case, we'll need to make ndr non-voting 14:54:25 <ralonsoh> haleyb, no 14:54:33 <ralonsoh> it is installing neutron from branch 14:54:37 <ralonsoh> from master, I mean 14:55:15 <ralonsoh> the main issue are the tempest jobs 14:55:58 <ralonsoh> dr-agent log: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_0c8/openstack/0c8f026475774091b3918d2dbad87d61/controller/logs/screen-q-dr-agent.txt 14:56:03 <ralonsoh> (from this patch) 14:56:42 <ralonsoh> so I don't know if something needs to be fixed in os-ken, I dind't have too much time for htis 14:57:58 <haleyb> ack, and did using 4.0.0 help? maybe after the requirements bump we'll know? 14:58:22 <ralonsoh> no, is installing os-ken from master too 14:59:55 <haleyb> alright, so that will need an owner as well 15:00:15 <lajoskatona> I suppose os-ken is behind the issues of ovs-agent so anyway we have to dug into it 15:00:32 <haleyb> should we at least wait until we're past the import error to look into it? 15:00:47 <lajoskatona> Currently I am trying to fix the dhcp agent but after I hope I can check os-ken / ovs-agent also 15:01:26 <haleyb> ack, we can loop back on it next week 15:02:25 <haleyb> we are over time, i'll quickly go through the rest 15:02:29 <haleyb> #topic specs 15:02:36 <haleyb> #link https://review.opendev.org/q/project:openstack/neutron-specs+status:open 15:02:50 <haleyb> i'm sure there are updates, review if you have time 15:02:51 <opendevreview> Merged openstack/neutron master: Remove temporary directories created in policy tests https://review.opendev.org/c/openstack/neutron/+/936116 15:02:59 <haleyb> #topic community goals 15:03:24 <haleyb> i know there are neutronclient patches in flight 15:04:03 <haleyb> please review antying in flight 15:04:04 <lajoskatona> yes, but I had no time in last weeks to check them, I have to go back to nova for example 15:04:25 <lajoskatona> I will check them and if something is hanging on neutron reviews I will ping you :-) 15:04:36 <lajoskatona> I mean the team 15:04:49 <haleyb> and we are treating any eventlet issue as a new bug, but if there are any reviews needing eyes add a note here or add to priority dashboard 15:05:12 <haleyb> #topic on-demand 15:05:17 <haleyb> ralonsoh: you had added something 15:05:49 <ralonsoh> yes related to the keepalived state change process 15:06:01 <ralonsoh> https://bugs.launchpad.net/neutron/+bug/1970944 15:06:26 <ralonsoh> I reviewed when we moved from keepalived notify_scripts to this python script we are using now 15:06:42 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron/+/125384 15:07:09 <ralonsoh> The rationale: with quick transitions, the status report of the local instances could be wrong 15:07:26 <ralonsoh> but I would like to highlight that this problem was also happening some years before 15:07:34 <ralonsoh> sorry years after 15:07:53 <ralonsoh> this is why we merged https://review.opendev.org/q/I70037da9cdd0f8448e0af8dd96b4e3f5de5728ad 15:08:23 <haleyb> i wish there was someone with a focus on ml2/ovs that could take on this bug, like Liu 15:08:29 <ralonsoh> So my proposal: to implement (again) something similar to the notify_events 15:08:51 <ralonsoh> and have both implementations available, via config option 15:09:06 <ralonsoh> and switch to the notify_events implementation next release 15:09:30 <ralonsoh> if during the next cycles this code is stable, we can delete the current one (python script) 15:10:03 <ralonsoh> maybe I should propose this in the drivers meeting, to allow voting 15:10:12 <ralonsoh> in any case, we are out of time 15:10:21 <lajoskatona> +1 for drivers meeting 15:10:25 <haleyb> ralonsoh: if you want to keep working on it that's fine, or if there is someone that can take from you would be better. 15:10:36 <ralonsoh> haleyb, agree 15:10:54 <haleyb> ok, we are over time 15:10:55 <ralonsoh> I think someone interested in ML2/OVS will be better 15:11:17 <haleyb> have a good week, reminder i am out rest of day, and no driver's meeting friday (i'm out too) 15:11:23 <haleyb> #endmeeting