14:00:28 <haleyb> #startmeeting networking 14:00:28 <opendevmeet> Meeting started Tue Mar 4 14:00:28 2025 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:28 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:28 <opendevmeet> The meeting name has been set to 'networking' 14:00:30 <haleyb> Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, obondarev, slaweq, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, haleyb, ralonsoh 14:00:38 <slaweq> o/ 14:00:42 <obondarev> o/ 14:00:42 <mlavalle> \o 14:01:01 <rubasov> o/ 14:01:21 <ralonsoh> hello 14:01:24 <frickler> \o 14:01:36 <haleyb> #announcements 14:01:38 <haleyb> hi everyone 14:01:42 <jlibosva> o/ 14:01:52 <haleyb> We are currently in week R-4 of Epoxy 14:01:54 <cbuggy> o/ 14:02:03 <haleyb> 2025.1 Epoxy release requirements are now frozen 14:02:14 <haleyb> Focus should be on finding and fixing release-critical bugs 14:02:33 <haleyb> stable/2025.1 branches should be created soon for all not-already-branched libraries 14:02:38 <sahid> o/ 14:02:52 <lajoskatona> o/ 14:02:53 <haleyb> RC1 deadline: March 14th, 2025 (R-3 week) 14:03:33 <haleyb> I was just pinged for Epoxy cycle highlights, please send me anything you think is significant, or put it here 14:04:21 <elvira> bit late but o/ 14:04:23 <haleyb> in summary of all that, we should focus on getting important things merged, like the eventlet patches 14:05:12 <haleyb> and a reminder to check the neutron priority dashboard for things ready to merge, there are a few things there 14:05:31 <bcafarel> late o/ 14:05:49 <haleyb> the PTG sign-up is open 14:05:51 <haleyb> https://ptg.openinfra.dev/ 14:06:00 <haleyb> April 7-11 14:06:15 <haleyb> i will get an etherpad created and sign-up for timeslots 14:07:09 <haleyb> and finally, 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 14:07:56 <haleyb> i will not be able to chair this week as I'm out Friday, we can either postpone or someone else can take over 14:08:22 <haleyb> at this point, it almost seems like we can wait since we should be getting things done, and fixing any bugs 14:08:48 <ralonsoh> I can chair the drivers meeting 14:08:50 <ralonsoh> there is agenda 14:08:58 <slaweq> IMO we can wait a week as new RFEs will be for new cycle for sure 14:09:13 <ralonsoh> as you decide 14:10:15 <haleyb> let's wait a week and fix the bugz 14:10:39 <haleyb> any other announcements? 14:12:03 <haleyb> #topic bugs 14:12:11 <haleyb> i was the deputy last week 14:12:13 <haleyb> report is at 14:12:20 <haleyb> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/GD5O7VBJ4QMXELV54GQW5BG3U7JDR2TA/ 14:12:26 <haleyb> there were quite a few :( 14:13:04 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2099981 14:13:11 <haleyb> [ovn] Adding an IPv6 subnet to public network breaks all FIPs 14:13:45 <haleyb> i wasn't able to reproduce this in my stack, but noonedeadpunk has multiple times, so it's something i'm doing 14:14:20 <haleyb> i will try again to see as i wanted to see if it was neutron or OVN by trying different versions 14:15:08 <haleyb> he also filed two seemingly related issues 14:15:17 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2099982 14:15:25 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2100294 14:15:47 <haleyb> once i can trigger the first i will test the others, but likely related 14:15:57 <haleyb> if i'm remembering correctly 14:16:19 <ralonsoh> ping me if you need help on these ones 14:16:30 <ralonsoh> will you take them? 14:17:57 <haleyb> i'll ping you as i might not be able to take them all, and i am out friday and monday so am time-limited on getting fixes out 14:18:07 <ralonsoh> sure 14:18:38 <noonedeadpunk> feeel free to ping me if any extra info is needed :) 14:19:29 <haleyb> noonedeadpunk: ack, thanks 14:19:38 <haleyb> next bug is 14:19:42 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2099911 14:19:49 <haleyb> Castellan as config source fails with "greenlet.error: cannot switch to a different thread" 14:20:31 <ralonsoh> DHCP agent is no longer using eventlet 14:20:32 <haleyb> i asked submittor to test with latest master branch to see if it helps 14:20:59 <ralonsoh> the L3 agent is, but this error usually happens if we don't monkey patch on time 14:21:00 <haleyb> right, and i don't know which version this was 14:21:05 <ralonsoh> that means, at the beginning 14:21:43 <haleyb> they had proposed a patch but it was adding config.setup_logging(fix_eventlet=True) so i -1'd it 14:22:03 <ralonsoh> how is possible to reproduce this? 14:22:29 <ralonsoh> I don't know what is "Setup Castellan as config source" 14:23:28 <haleyb> i don't either, which is why i asked them to try later code, but will ask 14:24:11 <haleyb> and it's deployed via kolla with l3-ha 14:24:39 <ralonsoh> in any case, as soon as we remove eventlet from L3, we'll be safe here too 14:25:29 <haleyb> right, i'd rather finish that and have then retry 14:25:53 <haleyb> next bug 14:25:57 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2100505 14:26:03 <haleyb> Neutron OVN external ports creation change HA chassis group master 14:26:12 <haleyb> ralonsoh: you confirmed this was a bug 14:26:27 <ralonsoh> yes, I'll push a patch this week 14:26:33 <ralonsoh> and I'll ask this folk to test it 14:26:43 <haleyb> ack, thanks 14:27:43 <haleyb> #link Neutron OVN external ports creation change HA chassis group master 14:27:45 <haleyb> [eventlet-removal] Remove the usage of eventlet in the Metadata agent 14:27:53 <haleyb> ralonsoh: you picked that up as well, thanks 14:28:01 <ralonsoh> there is a patch 14:28:15 <ralonsoh> https://review.opendev.org/c/openstack/neutron/+/942916 14:28:27 <ralonsoh> CI is passing (on ml2/ovs jobs, using this agent) 14:28:34 <haleyb> perfect, will review after meeting 14:28:43 <ralonsoh> there is a tech debt 14:28:56 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2100585 ? 14:28:58 <ralonsoh> I opened another bug to unify the metadata code 14:29:05 <ralonsoh> exactly 14:29:10 <ralonsoh> but this can wait 14:29:14 <haleyb> ack, we can work on that in F 14:29:51 <haleyb> i'll see if i can pick up, i like those kinds of refactors 14:30:32 <haleyb> and the last bug i had... 14:30:38 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2100268 14:30:45 <haleyb> wishlist to remove the cli from os-ken 14:31:22 <slaweq> that one should be probably easy to do, just require time to delete things properly :) 14:31:46 <ralonsoh> (a way to avoid new bugs related to the CLI) 14:32:12 <haleyb> but let's wait until after RC1 to work on it 14:32:37 <haleyb> that's all the bugs i had, any others to discuss? 14:32:45 <sahid> haleyb: 14:32:59 <sahid> may be the one related ml2/ovs? 14:33:07 <sahid> removal evetnthread 14:33:34 <haleyb> link? did i miss it? 14:33:37 <sahid> https://review.opendev.org/q/topic:%22bug/2087939%22 14:34:20 <ralonsoh> that belongs to the community goals 14:34:23 <haleyb> oh, let's talk about eventlet removal in community topic 14:34:30 <sahid> oh sorry :-) 14:34:31 <ralonsoh> I have one 14:34:38 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/2100853 14:34:52 <ralonsoh> please check it after the meeting 14:35:05 <ralonsoh> just to know if the current behaviour is expected or not 14:35:32 <ralonsoh> because I don't see reasonable that rules of different directions should interfere 14:35:45 <rubasov> to me it looks like a proper bug 14:36:06 <slaweq> rubasov I agree 14:36:16 <haleyb> rubasov: i'll let you decide as you're bug deputy this week :) 14:36:24 <rubasov> slaweq: thanks :-) 14:36:29 <ralonsoh> perfect then, thanks for the confirmation! 14:37:03 <haleyb> and mlavalle is deputy next week, ok with that? 14:37:53 <haleyb> we did creep up a little on bug count, probably just due to past two weeks increase in bugs - currently 774 14:38:41 <haleyb> and as i mentioned last week, please check the bug deputy list on the wiki for dates as they moved one week 14:38:47 <haleyb> #link https://wiki.openstack.org/wiki/Network/Meetings 14:39:30 <haleyb> alright, think we can move on 14:39:36 <haleyb> #topic community goals 14:39:59 <haleyb> lajoskatona: any update on neutronclient/horizon changes? i saw one was close 14:40:40 <lajoskatona> slow reviews, had to ping them again :-) 14:41:17 <haleyb> ack, they are ready to merge at least 14:41:36 <haleyb> #link https://review.opendev.org/q/topic:%22bug/1999774%22+status:open is the dashboard 14:41:48 <haleyb> ok, onto the main eventlet :) 14:42:17 <ralonsoh> so we have sahid's patches for OVS agent 14:42:21 <mlavalle> haleyb: ack no problem 14:42:29 <ralonsoh> #link https://review.opendev.org/q/topic:%22bug/2087939%22 14:42:49 <ralonsoh> (I still need to review the latest ones) 14:42:55 <ralonsoh> sahid, something to add here? 14:43:07 <sahid> no no thank you 14:43:30 <ralonsoh> and we have the other active LP, the metadata agent 14:43:45 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron/+/942916 14:43:50 <ralonsoh> that's all for now 14:44:01 <lajoskatona> There 's a wip for sriov agent also: https://review.opendev.org/c/openstack/neutron/+/942884 14:44:01 <ralonsoh> next week I'll start (continue) with the L3 agent 14:44:04 <ralonsoh> ahh yes 14:44:15 <ralonsoh> the SR-IOV agent one 14:44:33 <ralonsoh> lajoskatona, did you try with the VM with SR-IOV ports? 14:45:27 <lajoskatona> I still have issues with my env, but I am on it 14:45:50 <ralonsoh> if not, I'll try to borrow a sriov capable env in my company 14:45:52 <lajoskatona> if anybody has an sriov capable environment close to hands, please try it 14:46:01 <lajoskatona> ralonsoh: thanks 14:47:04 <ralonsoh> that's all 14:47:45 <MengyangZhang[m]> Hi, I have a question about the neutron qos. 14:47:47 <haleyb> ralonsoh: so do you think we can claim eventlet removal as a cycle highlight? or maybe you have a sentence to use? 14:48:17 <ralonsoh> haleyb, this is more a community goal 14:48:23 <haleyb> i mean, i will claim it as a highlight anyways with all the work that is done 14:48:24 <ralonsoh> I don't think this is a feature 14:48:33 <haleyb> true, not a feature 14:48:59 <ralonsoh> open a gerrit with the highlights to allow anyone to comment there 14:49:14 <haleyb> yes, will do that this week 14:49:17 <lajoskatona> +1 14:49:19 <slaweq> cycle highlights are things for the marketing/adverts more 14:49:30 <slaweq> so eventlet removal I am not sure is good fit there 14:49:45 <slaweq> it should IMO be highlighted in the release notes but not in the highlights maybe 14:50:45 <haleyb> sure, i'll take a look through things that merged 14:51:07 <haleyb> #topic on-demand 14:51:18 <haleyb> MengyangZhang[m]: you had a question about qos? 14:51:23 <MengyangZhang[m]> yes, As I was reading the doc(https://docs.openstack.org/neutron/latest/admin/config-qos.html), I see this "Each project can have at most one default QoS policy, although it is not mandatory. If a default QoS policy is defined, all new networks created within this project will have this policy assigned, as long as no other QoS policy is explicitly attached during the creation process. If the default QoS policy is unset, no 14:51:23 <MengyangZhang[m]> change to existing networks will be made." 14:51:23 <MengyangZhang[m]> I was wondering if the qos policy is applied to all nova instances under the project or can only be applied to new networks created under that project. Basically, how would instances under the project receive the qos policy? Any help on this would be much appreciated. 14:52:12 <ralonsoh> MengyangZhang[m], we can talk about this after the meeting 14:52:22 <MengyangZhang[m]> sure 14:53:08 <haleyb> anything else to discuss? i don't see anything on the agenda 14:54:10 <haleyb> ok, have a good week everyone. remember to try and focus on getting eventlet changes merged, and add things to the priority dashboard (and check the dashboard :) 14:54:18 <haleyb> #endmeeting