13:00:20 #startmeeting networking 13:00:20 Meeting started Tue Apr 29 13:00:20 2025 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:20 The meeting name has been set to 'networking' 13:00:28 Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, obondarev, slaweq, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, haleyb, ralonsoh 13:00:29 \o 13:00:34 o/ 13:00:39 o/ 13:00:53 o/ 13:00:58 hello 13:01:06 o/ 13:01:26 o/ 13:01:33 ok, we can get started 13:01:36 #announcements 13:01:46 \o 13:01:47 We are currently in Week R-22 of Flamingo 13:01:54 Final 2025.2 Flamingo release: October 3rd, 2025 13:02:01 #link https://releases.openstack.org/flamingo/schedule.html 13:02:21 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 13:02:30 o/ 13:03:11 i have no other announcements this week, anyone else have anything? 13:03:41 #topic bugs 13:03:55 ykarel was the deputy last week, his report is at 13:04:03 #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/UACIJRVC7LZ3ZVSEODQRVYNGIHIKAPCN/ 13:04:52 first bug 13:04:58 #link https://bugs.launchpad.net/neutron/+bug/2109592 13:05:05 [master][periodic] oslo master jobs broken with ImportError: cannot import name 'format_inspector' from 'oslo_utils.imageutils' 13:05:46 it looks like there was not a CI meeting yesterday so we can go over now, just don't see yatin here 13:05:49 shouldn't this be opened against nova? 13:06:20 yep ^ 13:06:44 File "/opt/stack/nova/nova/virt/images.py", line 28, in 13:06:44 from oslo_utils.imageutils import format_inspector 13:06:48 yes it does look like nova 13:08:24 i added nova 13:08:43 from the timing this might be related to new setuptools, too? 13:09:10 * haleyb shrugs 13:11:01 i will follow and ping nova after meeting in case they don't know 13:11:35 next bug also in periodic 13:11:39 #link https://bugs.launchpad.net/neutron/+bug/2109591 13:11:47 master periodic job running with centos 9-stream broken with py39 constraint drop https://review.opendev.org/c/openstack/requirements/+/948285 13:12:37 i remember leaving py39 in setup.cfg but guess other things did not 13:13:11 yes, keeping upper-constraints for py39 up to date was too much work and nobody wanted to help with it 13:13:54 so i guess onus is getting streams-10 job created and working? 13:14:28 so are we removing centos9 jobs? 13:14:33 stream-10 isn't going to work on rackspace hardware 13:14:42 that will remove the only one running FIPs 13:15:45 ralonsoh: i guess we could pin things to maybe get it working, this is only the periodic queue 13:16:12 so updating the requirements in the job? 13:16:56 that would work for a while maybe, until some service requires a new dependency. not sure if that would be worth the effort 13:17:36 at least we should try for a while 13:17:36 ralonsoh: i guess? like frickler said that would work for a while 13:17:48 we should keep running the fips job 13:19:07 is there someone that can take this? or we can wait for yatin? 13:19:26 I'll try this week but I have one day left only 13:20:18 ack, thanks 13:20:25 Ill be here this week ralonsoh. let me know if you need me to take over 13:20:44 mtomaska, most probably 13:20:49 ACK 13:21:26 thanks everyone. next high bug 13:21:33 #link https://bugs.launchpad.net/neutron/+bug/2109456 13:21:39 AttributeError occurs when querying ports immediately after starting neutron-server 13:21:54 Gwangseok took it, propsed change 13:22:10 #link https://review.opendev.org/c/openstack/neutron/+/948306 13:22:29 sorry, the issue is happening in Yoga with Linux Bridge 13:22:40 we don't even use the eventlet server anymore 13:22:48 so any patch for master could be useless 13:24:04 in any case, I'll try to reproduce it after this meeting 13:24:17 should be straightforward 13:24:22 ralonsoh: ack, thanks 13:24:25 o/ (late late) 13:24:47 next high bug 13:24:51 #link https://bugs.launchpad.net/neutron/+bug/2107925 13:24:59 [ovn] neutron_pg_drop deleted when using network log 13:25:14 elvira took ownership, thanks! 13:25:32 i see two patches 13:25:50 #link https://review.opendev.org/c/openstack/neutron/+/948053 13:26:07 #link https://review.opendev.org/c/openstack/neutron/+/948215 13:26:31 looks legit 13:28:28 i know i looked at one of the changes, don't think i had time for the other, will put it on my list 13:29:46 alright, and one more high priority 13:29:55 #link https://bugs.launchpad.net/neutron/+bug/2109326 13:30:02 test_release_for_ipv6_lease_no_dhcp_release6 unit tests fails randomly 13:30:39 I think yatin's recommendation is good 13:30:57 --> https://review.opendev.org/c/openstack/neutron/+/948180 13:31:34 yup. my only question was why it started happening now, but guess that's the randomness aspect 13:32:40 alright, we still have a couple more unassigned so will go through them 13:32:50 #link https://bugs.launchpad.net/neutron/+bug/2109354 13:32:58 Incorrect behavior when filtering networks by provider:network_type with a limit parameter 13:33:48 I'll check it, I've been playing recently with this code 13:34:04 ralonsoh: ack, thanks 13:34:40 next is actually rfe 13:34:44 #link https://bugs.launchpad.net/neutron/+bug/2108952 13:34:51 [RFE][ml2][ovs] distributed DHCP extends supports for DPU baremetal 13:35:17 i'll ask for more info from Liu, not much in the description to go on 13:35:51 I thought DPU ports were HW offloaded ports 13:36:01 that had the same behaviour as a kernel port 13:36:07 (but let's wait for more info) 13:37:13 and finally last one 13:37:16 #link https://bugs.launchpad.net/neutron/+bug/2108985 13:37:22 [BGP DVR] Incorrect /32 Fixed IP Route Advertisement Logic 13:37:50 this was on Zed, asked to reproduce on something that is maintained, or master 13:39:03 that was all the bugs in the list, any others to discuss? 13:39:45 this week otherwiseguy is the deputy, next week is mtomaska 13:40:09 i forgot to ping terry last week will do that today 13:40:17 ACK 13:40:57 alright, moving on 13:41:00 #topic community goals 13:41:21 ralonsoh: did you want to give an update on eventlet patches? 13:41:30 i don't think i reviewed much last week 13:41:32 not this week, sorru 13:41:38 I'm still working on this one 13:41:48 ack 13:41:58 https://review.opendev.org/c/openstack/neutron/+/938404 13:42:04 I reviewd some of ovs-agent patches, thanks sahid 13:42:33 https://review.opendev.org/q/topic:%22bug/2087939%22 13:42:58 lajoskatona: i think we both had the same question about one of the ovs patches 13:43:22 #link https://review.opendev.org/c/openstack/neutron/+/940983 ? 13:43:23 yes that is the first in the chain 13:43:35 exactly 13:44:11 I sent the link of the topic 13:44:14 all patches are there 13:44:30 ralonsoh: thanks, will try and get through a few more 13:44:36 ralonsoh: thanks 13:45:23 I was thinking to start playing with the oslo.services patch, but I had no time recently, and I am not sure if it really useful at themoment for neutron 13:45:38 ralonsoh: can you also look at the patch i linked above from sahid ? would be good to make sure we are asking for something we really need 13:45:48 haleyb, sure 13:46:04 lajoskatona, could be interesting, for example, testing with the dhcp agent 13:46:17 we had problems because of the backend used (only eventlet) 13:46:34 I try to allocate time to it, but for me this is also a short week, so no promise :-) 13:46:36 if they are providing the kernel threads one, we can use it 13:46:40 lajoskatona, thanks! 13:46:59 lajoskatona: oh, is there a new oslo.services patch or a release? i have been offline for a few days 13:47:25 no it is under review 13:47:37 #link https://review.opendev.org/c/openstack/oslo.service/+/945720 13:47:40 ack 13:48:09 we had out dhcp agent with all Neutron parts without eventlet 13:48:14 and gibi started to add fixes for it as I see 13:48:16 then we needed to revert some of them 13:48:30 we can propose again the patches reverted depending on this one 13:48:54 +1, that can be good testing in the CI also 13:49:00 we'll need a CI job that can install oslo.service from repository 13:49:17 (for example, *oslo-master) 13:50:58 agreed 13:51:34 lajoskatona: any new neutronclient changes? 13:51:48 yes 13:52:01 woo-hoo! 13:52:04 for fullstack: https://review.opendev.org/c/openstack/neutron/+/947851 13:52:35 and it is actually not progressing as with sdk not that easy to have no-auth client 13:52:45 at least I can't find the magic to have it :-) 13:52:57 still wip, right? 13:53:17 I had a discussion with gtema last week, some suggestion from him but still missing something to have the endpoints 13:53:45 malavalle: yes, it is failing just uploaded to have it out of my local machine before I lose it :-) 13:54:07 lajoskatona: ack, thanks for working on it 13:54:08 lol, I do the same 13:54:16 :-) 13:54:47 that's it for this topic 13:55:12 ok, can move on 13:55:20 #topic on-demand 13:55:33 there was actually one more bug, noticed a follow-on from yatin 13:55:39 #link https://bugs.launchpad.net/neutron/+bug/2109388 13:55:47 [OVN] Slow "router add interface" when MTU differs and ovn_emit_need_to_frag=true 13:56:01 he pinged me this morning 13:56:04 I'm checking this now 13:56:30 I'll assign it to me 13:56:46 ralonsoh: ack, thanks 13:57:29 any other topics to discuss? almost at top of hour 13:58:36 ok, thanks for attending everyone, have a good week! 13:58:39 #endmeeting