14:00:39 <haleyb> #startmeeting networking 14:00:39 <opendevmeet> Meeting started Tue Jun 25 14:00:39 2024 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:39 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:39 <opendevmeet> The meeting name has been set to 'networking' 14:00:41 <slaweq> ralonsoh I already +2 this patch 14:00:41 <haleyb> Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, obondarev, slaweq, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, amotoki, haleyb, ralonsoh 14:00:42 <mlavalle> \o 14:00:46 <elvira> o/ 14:00:47 <ralonsoh> slaweq, thanks! 14:00:49 <ralonsoh> hello 14:00:50 <obondarev> hi 14:00:55 <slaweq> o/ 14:00:59 <lajoskatona> o/ 14:01:17 <ihrachys> o/ 14:01:42 <rubasov> o/ 14:01:56 <haleyb> #topic announcements 14:02:15 <haleyb> We are now in Dalmatian release week (R - 13) 14:02:17 <haleyb> #link https://releases.openstack.org/dalmatian/schedule.html 14:02:28 <haleyb> The Dalmatian-2 milestone will happen next month, on July 4th, 2024 14:02:35 <haleyb> At milestone-2, the release team will propose releases for any library that has not been otherwise released since milestone-1 14:03:06 <haleyb> So land any *-lib patches in the next week to make sure they land 14:03:29 <ykarel> o/ 14:03:41 <slaweq> deadline is end of next week? 14:03:44 <slaweq> or when exactly? 14:04:04 <slaweq> ahh, sorry, you said already 14:04:11 <slaweq> 4th July 14:04:22 <haleyb> slaweq: yes, July did sneak up on us, it's next week 14:04:51 <haleyb> and like most in the US, i'll be offline 4th and 5th 14:05:00 <mlavalle> \o/ 14:05:03 <mlavalle> me too 14:05:08 <lajoskatona> do we have high prio items in n-lib to be merged quickly? 14:05:16 <ralonsoh> we need to +1 the release patches before this day then 14:05:28 <ralonsoh> lajoskatona, ^^ thanks! I would like to have it 14:06:12 <haleyb> #link https://review.opendev.org/q/project:openstack/neutron-lib 14:06:21 <lajoskatona> by the way do os-ken is in the same lib category as n-lib? Do we have to release that also next week? 14:06:28 <bcafarel> late o/ 14:06:41 <slaweq> lajoskatona I believe os-ken is lib also 14:06:50 <slaweq> so it has the same deadline 14:07:09 <haleyb> lajoskatona: you must be asking about https://review.opendev.org/c/openstack/os-ken/+/922622 ? :) 14:07:11 <ralonsoh> so we need the os-ken patch for the dead loop 14:07:28 <slaweq> regarding neutron-lib I would maybe like to include https://review.opendev.org/c/openstack/neutron-lib/+/921649 if possible 14:07:33 <haleyb> ^^ is on my list right after meeting 14:07:40 <lajoskatona> haleyb: that one :-) 14:07:51 <slaweq> haleyb thx 14:08:18 <haleyb> slaweq: oh, i was pointing at the os-ken one, but will look at that once it passes functional :) 14:08:39 <slaweq> haha, sure 14:09:33 <slaweq> I will fix that failed tests just after this meeting 14:10:09 <haleyb> ack, i didn't know if it was just the gate, that job had been flaky i thought 14:10:33 <slaweq> I think it may be related to my changes there 14:11:13 <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 14:11:31 <haleyb> i think there was one topic from mlavalle 14:11:57 <ralonsoh> there is one mine 14:12:07 <haleyb> ralonsoh: ack 14:12:20 <haleyb> the other one was regarding dns domain, https://review.opendev.org/c/openstack/neutron/+/920459 14:12:28 <slaweq> ++ 14:12:52 <haleyb> so there will be a meeting on Friday 14:13:29 <haleyb> mlavalle: thinking of 14:13:33 <haleyb> arg 14:13:57 <mlavalle> haleyb: thanks! 14:14:01 <haleyb> mlavalle: thinking of D-2 milestone, was there any changes for ironic that we needed? 14:14:13 <mlavalle> not that I know of 14:14:23 <mlavalle> but I can check with them 14:14:32 <mlavalle> and let you know later this week 14:15:01 <haleyb> mlavalle: ack, i left the note in our meeting wiki to remind myself but hadn't seen anything 14:15:20 <mlavalle> I have no problem checking with them 14:15:58 <haleyb> any other announcements? 14:16:48 <haleyb> alright, next topic 14:16:54 <haleyb> #topic bugs 14:17:38 <haleyb> isabek was supposed to be deputy last week, but he pinged me late saying he could no longer to it 14:17:59 <haleyb> instead of moving everyone a week i just did it, but we are now down one deputy 14:18:37 <haleyb> so everyone make sure to see when your next week is, and if you cannot do it please ping me early, or swap 14:19:16 <haleyb> 14 deputies is still a good number 14:19:35 <mlavalle> it is. still, it is bad to see another one go 14:19:36 <haleyb> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/GTCE24CAUSUABSSCM6SDPGLQ6VPNQV4R/ 14:20:17 <haleyb> that was my report, it was mildly busy 14:20:31 <haleyb> and i had a couple of bugs to talk about 14:20:49 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2069689 14:21:31 <haleyb> this was regarding breakind n-d-r and vpnaas with wsgi changes - do we still need the neutron workaround now that patches have landed? 14:21:50 <haleyb> i.e. https://review.opendev.org/c/openstack/neutron/+/922185 14:22:26 <ralonsoh> we need new releases of these projects, to avoid people form downloading the lastest version with the error 14:23:07 <ralonsoh> but no, we should not need the neutron patch 14:23:17 <ralonsoh> this is only in master (that is dev only) 14:23:53 <haleyb> ralonsoh: ack, i just wanted to make sure i didn't miss anything, figured there would have been pings if still needed 14:24:13 <ralonsoh> we didn't release any Neutron version with this error 14:24:30 <ralonsoh> and next one will be used by the updated n-d-r and vpnaas 14:24:44 <ralonsoh> ^ we need to update the requirements once the next neutron is released 14:24:50 <ykarel> we had recent failures in dynamic-routing weekly jobs https://zuul.openstack.org/buildset/a21e500158db4abfa9f2c3cae8a73d3e 14:25:25 <ykarel> after the linked ndr patch merged, is there anything up to fix ^? 14:26:28 <lajoskatona> I have to ping the exabgp folks about the fix for py312 & bagpipe 14:26:44 <haleyb> ykarel: i didn't think so, is artem around? i'm bad at remembering nicks 14:27:01 <slaweq> gtema (Artem Goncharov) 14:27:14 <haleyb> slaweq: thanks 14:27:33 <ralonsoh> the problem is that n-d-r is retrieving a stable neutron version neutron-24.0.0 14:27:47 <ralonsoh> this neutron version has the wsgi module in the older place 14:28:03 <ralonsoh> so what we need is, asap, to release a neutron version and update the requirements 14:28:11 <lajoskatona> yes usually we releae a rc for these issues 14:28:19 <lajoskatona> release ---^ 14:28:20 <ralonsoh> or make these projects robust enough to import from 2 places 14:28:39 <haleyb> right, with a try/except 14:28:52 <ykarel> +1 14:28:58 <ralonsoh> I prefer, for now, the try/except option 14:29:02 <ralonsoh> with a comment to remove it 14:29:08 <gtema> I think better is to release neutron 14:29:31 <gtema> this try/except is necessary only temporary and would need to be dropped later - too much additional work 14:29:50 <ralonsoh> gtema, additional work? one patch? 14:30:08 <ralonsoh> that's ok but with the release you also need to update the requirements of the projects 14:31:06 <gtema> under additional work I mean one patch introducing try/except and second dropping if afterwards (keeping it for long time is not cool) 14:31:33 <ralonsoh> haleyb, I'll propose them tomorrow morning and add all of you as reviewers 14:31:38 <ralonsoh> I think we can continue 14:31:42 <lajoskatona> I can push the try except if everybody agrees, but I can live with old school neutron rc release also 14:32:58 <haleyb> ralonsoh: ack, thanks 14:33:11 <haleyb> the other bug i wanted to mention was 14:33:16 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2069718 14:33:43 <haleyb> i see stefan did drop a comment he was working on a patch 14:33:51 <ralonsoh> I will comment on the LP bug 14:33:57 <ralonsoh> why northd is stopped? 14:34:13 <ralonsoh> this is a required service, if we stop it, how OVN is intended to work? 14:34:28 <haleyb> ralonsoh: thanks. i think to reproduce they stopped it, but in prodoction it's maybe just slow because it has a lot to process? 14:34:28 <ralonsoh> this is like in ML2/OVS stopping the OVS agent 14:34:47 <slaweq> haleyb that is also my understanding 14:35:21 <slaweq> stopping it kind of simulate of the slow down which they experience in the production 14:35:27 <haleyb> i.e. the NB DB was updated and live migration proceeded, but failed 14:35:41 <ralonsoh> but this is an scale problem 14:36:01 <ralonsoh> not a Neutron problem itself, they would probably need to improve the HW of the controllers 14:36:20 <haleyb> ralonsoh: right, i guess i saw the question as "is there another way for neutron to do this, like waiting for SB to update?" 14:36:45 <haleyb> the answer can be No, but wanted to see what others thought 14:36:54 <ralonsoh> maybe with Ihar's improvements with multiple port binding 14:37:02 <ralonsoh> (I don't have the patch here) 14:37:08 <slaweq> but should't neutron somehow restore VM's connectivity after live-migration at some point? 14:37:12 <ralonsoh> but this is not in Zed, for sure, but in newer versions 14:37:35 <slaweq> IIUC reproduce steps, in such case after migration is done connectivity is permanently broken 14:37:44 <slaweq> or maybe I am missing something there 14:38:16 <ralonsoh> that's the point: you can have a slow northd and regain connectivity 4-20 secs later (as state in the first line) 14:38:44 <slaweq> ahh, ok 14:38:55 <ihrachys> you can detect the completion of live migration - nova tells us - requested-chassis field is updated back to single value 14:39:09 <slaweq> from the reproduction steps (step 4 in particular) I understood something different 14:39:19 <ralonsoh> ihrachys, but in newer versions, not zed 14:40:04 <haleyb> slaweq: i focused on step 4 as well 14:40:55 <haleyb> please drop any comments in the bug, and we can wait for any patch to see if it's worth taking 14:40:55 <ralonsoh> ihrachys, no, my bad: https://review.opendev.org/c/openstack/neutron/+/828455 this is in Zed too 14:41:31 <ihrachys> nbctl commands have --wait=hv - these are supposed to wait for ovn-controller to get the memo 14:42:12 <ihrachys> (I will admit I haven't given the LP much thought, so just spitting random ideas - will check later with more attention) 14:42:59 <haleyb> ihrachys: thanks 14:43:23 <haleyb> i had no other bugs, but maybe someone else has one to discuss? 14:44:06 <haleyb> ok, moving on 14:44:09 <haleyb> #topic specs 14:44:15 <haleyb> #link https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open 14:45:03 <haleyb> just one and it has -1 so waiting on an update 14:45:33 <haleyb> #topic community-goals 14:45:49 <haleyb> slaweq: any update on S-RBAC? 14:45:54 <slaweq> no updates about S-RBAC from me still 14:45:56 <slaweq> sorry 14:46:10 <slaweq> I need to get into this finally but still haven't got time yet 14:46:21 <haleyb> np 14:46:29 <haleyb> #topic on-demand 14:46:43 <haleyb> any other topics to discuss? 14:46:54 <ralonsoh> quick one: can we add "eventlet-deprecation" to community goals? 14:47:14 <ralonsoh> to weekly share the open specs, patches and strategies 14:47:42 <ralonsoh> for now, there are two open specs, one in the TC and the other one for oslo.services 14:47:42 <haleyb> ralonsoh: yes, that's a good idea, i can add it for next week 14:47:49 <ralonsoh> perfect, thanks, that's all 14:47:52 <mlavalle> can I get someone to look at https://review.opendev.org/c/openstack/neutron/+/918151? 14:48:12 <haleyb> ralonsoh: and regarding your patches, do you need any reviews? 14:48:22 <ralonsoh> yes, if you don't mind 14:48:30 <ralonsoh> and the n-lib one I have open 14:48:40 <ralonsoh> mlavalle, ack, on my pile for tomorrow morning 14:48:51 <mlavalle> ralonsoh: thanks 14:48:53 <haleyb> do they share a topic? or i'll just look in review emails 14:49:27 <ralonsoh> for now I'm opening bugs, for example to add new services running with wsgi 14:49:39 <ralonsoh> the ovn implementation with wsgi is related to the deprecation 14:49:45 <ralonsoh> eventlet deprecation 14:49:58 <ralonsoh> https://review.opendev.org/q/topic:%22bug/2069581%22 14:50:00 <ralonsoh> and 14:50:14 <ralonsoh> https://review.opendev.org/q/topic:%22bug/1912359%22 14:50:16 <ralonsoh> that's all 14:50:25 <haleyb> ack, thanks will look later 14:50:49 <haleyb> CI meeting is in 10 minutes, in-person ykarel ? 14:51:05 <ykarel> yeap video today 14:51:08 <lajoskatona> +1 14:51:14 <slaweq> see You there 14:51:43 <haleyb> ack, i might have to miss as someone is here to fix something in a few minutes 14:52:00 <haleyb> thanks for attending, ping if you need any reviews 14:52:04 <haleyb> #endmeeting