14:00:59 <lajoskatona> #startmeeting networking 14:00:59 <opendevmeet> Meeting started Tue Aug 30 14:00:59 2022 UTC and is due to finish in 60 minutes. The chair is lajoskatona. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:59 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:59 <opendevmeet> The meeting name has been set to 'networking' 14:01:01 <mlavalle> o/ 14:01:02 <lajoskatona> o/ 14:01:08 <rubasov> o/ 14:01:08 <obondarev> hi 14:01:12 <ralonsoh> hi 14:01:18 <frickler> \o 14:01:45 <isabek> hi 14:01:59 <slaweq> o/ 14:02:47 <lajoskatona> Ok ,let's start 14:02:52 <lajoskatona> #topic Announcements 14:03:02 <lajoskatona> The usual Zed schedule: https://releases.openstack.org/zed/schedule.html 14:03:12 <lajoskatona> and the release countdown mail: 14:03:18 <lajoskatona> [release] Release countdown for week R-5, Aug 29 - Sep 02 (#link https://lists.openstack.org/pipermail/openstack-discuss/2022-August/030138.html ) 14:03:34 <lajoskatona> Some dates from it: 14:03:52 <lajoskatona> Zed-3 milestone (feature freeze): September 1st, 2022 (R-5 week) 14:03:59 <lajoskatona> that is Thursday this week 14:04:17 <lajoskatona> RC1 deadline: September 15th, 2022 (R-3 week) 14:04:17 <lajoskatona> Final RC deadline: September 29th, 2022 (R-1 week) 14:04:17 <lajoskatona> Final Zed release: October 5th, 2022 14:04:17 <lajoskatona> Next PTG: October 17-21, 2022 (virtual PTG!!!) 14:04:42 <lajoskatona> The etherpad for it: https://etherpad.opendev.org/p/neutron-antelope-ptg 14:05:01 <lajoskatona> I add a mental not to myself to ask around for the schedule 14:05:29 <lajoskatona> Cycle Highlights, please check it: https://review.opendev.org/c/openstack/releases/+/853813 14:05:52 <lajoskatona> My memory faded so thanks for the good comments on it already 14:07:04 <lajoskatona> Perhaps I add here a small news: there will be again edge session on the PTG: 14:07:09 <lajoskatona> https://lists.openstack.org/pipermail/openstack-discuss/2022-August/030128.html 14:07:19 <lajoskatona> and there is an etherpad for it: https://etherpad.opendev.org/p/ecg-ptg-october-2022 14:07:30 <lajoskatona> if you are interested check it 14:08:06 <lajoskatona> My personal point is that I would like to see exactly what and why we need at the edge sites 14:09:13 <lajoskatona> In Berlin I was on a session for edge, and I heard some bullshit (self driving cars, and things like that which are far from mass usage at the moment) and the only thing which sounded as real was gaming which needed some edge features 14:10:00 * slaweq don't wants to have self driving car with Openstack running on it :P 14:10:14 <ralonsoh> hehehe 14:10:19 <lajoskatona> :-) 14:10:30 <mlavalle> I wouldn't mind 14:11:05 <lajoskatona> there was even a slide as I remember with a fighter jet with edge sites inside it :-) 14:11:44 <lajoskatona> ok, that's it for the PTG from me 14:11:45 <haleyb> can Openstack drive a manual transmission? :) 14:12:23 <slaweq> haleyb: I don't think it will be fast enough :P 14:12:34 <slaweq> and rabbitmq can be stucked somewhere in the middle :D 14:13:14 <lajoskatona> :D 14:13:37 <obondarev> anything - but let it not be brakes please! 14:13:43 <slaweq> LOL 14:13:49 <slaweq> obondarev++ 14:14:19 <lajoskatona> Another thing from me for announcements: 14:14:25 <lajoskatona> Please recheck with reason 14:14:31 <lajoskatona> Please check the numbers from slaweq: https://etherpad.opendev.org/p/neutron-ci-meetings#L42 14:15:55 <lajoskatona> If there is no more comments or questions for announcements we can move on 14:16:12 <frickler> PTL nominations? 14:16:25 <lajoskatona> frickler: good comment, thanks 14:16:37 <ralonsoh> I'm on it, I'll send a mail/patch today, in the next hour 14:16:48 <lajoskatona> ralonsoh: thanks 14:16:57 <ralonsoh> (and before that, lajoskatona thanks a lot!) 14:16:59 <mlavalle> ralonsoh: ++ 14:18:00 <lajoskatona> Ok, next topic 14:18:04 <lajoskatona> #topic Bugs 14:18:13 <lajoskatona> Report from slaweq: https://lists.openstack.org/pipermail/openstack-discuss/2022-August/030168.html 14:18:27 <lajoskatona> bug numbers are going upper as summer period is over :-) 14:18:44 <lajoskatona> I saw some which needs more attention 14:18:51 <lajoskatona> https://bugs.launchpad.net/neutron/+bug/1987530 - Duplicate external_ip in NAT table lead to loss of N/S connectivity - needs attention 14:19:52 <ralonsoh> I'll take a look at this one, this should not happen of course 14:20:14 <lajoskatona> ralonsoh: thanks 14:20:30 <lajoskatona> Another one without owner: 14:20:33 <lajoskatona> https://bugs.launchpad.net/neutron/+bug/1987666 - Race condition when adding two subnet with same cidr to router - unassigned currently 14:20:45 <slaweq> I think that froyo is working on it currently 14:20:51 <slaweq> but maybe he didn't assign it to himself 14:21:08 <froyo> slaweq, lajoskatona yeah! 14:21:16 <lajoskatona> ahh, ok, thanks 14:21:17 <slaweq> froyo++ thx 14:21:28 <froyo> done! 14:22:12 <mlavalle> :-) 14:22:31 <lajoskatona> An old cloud-init bug: https://bugs.launchpad.net/cloud-init/+bug/1899487 - cloud-init hard codes MTU configuration at initial deploy time - bug which was reported for cloud-init first but now after long discussion it's on Neutron and we need to check it 14:22:49 <slaweq> yeah, that one I wanted to raise today too :) 14:23:08 <lajoskatona> :-) 14:26:49 <lajoskatona> If you have some time please check it, I also try to check the history of this issue 14:27:15 <lajoskatona> Do you have any bug which you would like to discuss? 14:27:35 <slaweq> nothing more to discuss 14:27:39 <frickler> there was the CI issue with designate yesterday 14:27:45 <lajoskatona> ok, thanks 14:27:47 <slaweq> but there is one https://bugs.launchpad.net/neutron/+bug/1988026 14:28:01 <slaweq> which looks like low-hanging-fruit for me 14:28:06 <frickler> not sure if there is really a bug hidden there for neutron in terms of deleting SGs 14:28:21 <slaweq> so maybe there would be someone who would like to take a look into it 14:28:24 <slaweq> that's all from me 14:28:30 <frickler> that bug came out of debugging things 14:28:45 <frickler> but it doesn't seem to be new 14:29:10 <frickler> the patch in tempest for what triggered the issue is https://review.opendev.org/c/openstack/tempest/+/854973 14:29:36 <lajoskatona> friskler: thanks for checking it 14:29:39 <frickler> but that has been in place for a long time and things in CI only started to fail around friday 14:29:45 <lajoskatona> frickler, sorry 14:30:09 <frickler> so maybe there is yet another issue. I didn't get round to reproducing locally yet 14:30:42 <lajoskatona> yes that is strange that it started to explode in designate CI from end of the last week 14:31:29 <slaweq> frickler: do you have opened bug somewhere? 14:31:43 <frickler> designate has a lot of tests that do not use network resources, which I think is why this was visible there most 14:31:59 <lajoskatona> slaweq: this one I think: https://bugs.launchpad.net/neutron/+bug/1988026 14:32:13 <frickler> slaweq: not yet, because I'm not sure if there actually is a bug. maybe neutron now is only more correct than earlier 14:32:17 <slaweq> lajoskatona: regarding that cloud-init bug, I was going to say that I will try to look into it this week but I wrote it in different window (internal irc) :D 14:32:46 <slaweq> lajoskatona: ahh, ok 14:32:52 <lajoskatona> slaweq: ack, the message arrived 14:32:56 <slaweq> so that is the one which I think may be low-handing-fruit in neutron 14:33:03 <lajoskatona> frickler: ok, thanks 14:33:30 <frickler> 1988026 is related but not what caused CI to fail now 14:34:45 <frickler> the underlying issue is fixed with the tempest patch, so also no CI impact any longer. at least none I'm aware of 14:34:55 <frickler> and I think that's it for now. ;) 14:35:41 <lajoskatona> ok, so let's keep our eyes open if similar issues appear 14:36:00 <lajoskatona> One more sentence for the bug deputy week: This week haleyb is the deputy and next week amotoki will be. 14:36:12 <amotoki> lajoskatona: ack 14:36:34 <haleyb> lajoskatona: ack 14:37:05 <lajoskatona> amotoki, haleyb: thanks 14:37:15 <lajoskatona> #topic On Demand Agenda 14:37:24 <lajoskatona> (slaweq) PTG - TC+Leaders Interaction Session - poll opened https://framadate.org/zsOqRxfVcmtjaPBC 14:37:43 <slaweq> ahh, right 14:38:10 <slaweq> I just wanted to mention that gmann opened poll to choose the best time slot for the TC+Leaders session during the PTG 14:38:24 <slaweq> so if anyone is interested in participating in it, please vote there 14:38:48 <slaweq> lajoskatona: ralonsoh looking at You especially ;) 14:38:56 <ralonsoh> I'm updating it now 14:39:02 <lajoskatona> slaweq: thanks, I voted on the time slots :-) 14:39:11 <slaweq> thx 14:39:15 <slaweq> that's all from me 14:40:05 <lajoskatona> I have another one, more a question: 14:40:10 <lajoskatona> (lajoskatona): OVS (&OVN) version in CI 14:40:38 <lajoskatona> I just greped quickly and saw that even in our CI we have different hashes for on and ovs branch 14:41:09 <lajoskatona> perhaps we should have a common version for all master jobs? 14:41:44 <slaweq> yeah, we should use "master" and "main" branch basically 14:41:53 <ralonsoh> right and we should update it anytime OVN master (or any other branch) update the dependency 14:41:55 <ralonsoh> not really 14:41:58 <ykarel> but those do not work correctly sometime 14:42:03 <ralonsoh> master requires a specific OIVS version 14:42:09 <ykarel> yeap ^ 14:42:35 <ralonsoh> e.g.: https://review.opendev.org/c/openstack/ovn-octavia-provider/+/854008 14:42:59 <ralonsoh> tomorrow I'll check all OVN master references 14:43:04 <lajoskatona> yeah I was looking for this one 14:43:06 <ralonsoh> and update the needed OVS version 14:43:14 <slaweq> ++ 14:43:20 <ykarel> may be we can looking into adding an option so ovs version is auto detected based on ovn source 14:43:34 <ykarel> atleast when ovn branch is main 14:43:46 <lajoskatona> perhaps only some guide in docs is enough if it is hard to automate 14:43:57 <ralonsoh> we can introduce an ansible role to read that yeah 14:44:24 <ralonsoh> first I'll update the OVS versions statically 14:44:31 <ralonsoh> then I'll open a LP for this improvement 14:44:35 <lajoskatona> ralonsoh: thanks 14:44:38 <lajoskatona> good idea 14:44:41 <ykarel> +1 14:46:04 <lajoskatona> That's it from me, do you have anything more to discuss? 14:46:18 <slaweq> nothing from me 14:46:19 <frickler> actually yes 14:46:38 <frickler> I added a topic to the PTG etherpad: neutron-legacy in devstack 14:47:01 <frickler> we planned for new neutron code to be dropped in A cycle after it was deprecated earlier 14:47:17 <frickler> maybe someone here is interested in actually doing this 14:47:34 <frickler> would save you a lot of duplication when adding new features 14:48:07 <frickler> and it would be good to do it early in the cycle, so preparing a patch could actually start soon 14:48:08 <lajoskatona> frickler: good idea to discuss it, and write down/discuss the basics before removing anything 14:48:58 <lajoskatona> and now I see that this could be a shared session with QA team 14:49:56 <lajoskatona> but we use now neutron-legacy or am I wrong? 14:50:02 <frickler> yes, though not sure if anyone other than me would join, but here's hoping 14:50:15 <frickler> well currently the usage is mixed I think 14:50:44 <frickler> would need to search for the deprecation message in logs 14:51:09 <frickler> but good idea to check that before removing things, too 14:51:16 <lajoskatona> +1 14:52:23 <lajoskatona> frickler: thanks for bringing it here 14:52:35 <lajoskatona> If nothing more we can close the meeting 14:53:00 <frickler> not from me, thx lajoskatona 14:53:20 <lajoskatona> We have the CI meeting after this one, and it will be video meeting today: https://meetpad.opendev.org/neutron-ci-meetings 14:53:31 <lajoskatona> see you there 14:53:31 <slaweq> ++ 14:53:35 <lajoskatona> #endmeeting