14:00:26 <haleyb> #startmeeting networking
14:00:26 <opendevmeet> Meeting started Tue Apr 30 14:00:26 2024 UTC and is due to finish in 60 minutes.  The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:26 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:26 <opendevmeet> The meeting name has been set to 'networking'
14:00:29 <mlavalle> \o
14:00:29 <haleyb> Ping list: bcafarel, elvira, frickler, mlavalle, mtomaska, obondarev, slaweq, tobias-urdin, ykarel, lajoskatona, jlibosva, averdagu, amotoki, haleyb, ralonsoh
14:00:30 <ihrachys> o/
14:00:32 <lajoskatona> o/
14:00:35 <ykarel> o/
14:01:06 <slaweq> o/
14:01:36 <haleyb> #topic announcements
14:01:45 <haleyb> hopefully a short meeting
14:02:04 <haleyb> We are now in Dalmatian release week (R - 22)
14:02:25 <opendevreview> Merged openstack/ovn-octavia-provider master: Remove leftover OVN LB HM port upon deletion of a member  https://review.opendev.org/c/openstack/ovn-octavia-provider/+/916637
14:02:29 <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
14:02:33 <bcafarel> o/
14:02:41 <haleyb> i think there is one new RFE this week
14:03:05 <haleyb> Zed has transitioned to unmaintained
14:03:07 <haleyb> #link https://review.opendev.org/c/openstack/releases/+/916492
14:03:33 <haleyb> so you might see some reno and other small patches in all neutron and networking* repos
14:03:38 <haleyb> please approve them
14:04:45 <lajoskatona> one neutron-tempest-plugin ptch for it: https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/917696
14:05:28 <opendevreview> Christian Berendt proposed openstack/ovn-octavia-provider stable/2024.1: Remove leftover OVN LB HM port upon deletion of a member  https://review.opendev.org/c/openstack/ovn-octavia-provider/+/917724
14:05:29 <haleyb> lajoskatona: thanks, hopefully we can get that green
14:05:49 <lajoskatona> +1
14:06:25 <haleyb> for those that didn't attend last Friday's drivers meeting we had a good discussion with the ironic team
14:06:32 <haleyb> #link https://meetings.opendev.org/meetings/neutron_drivers/2024/neutron_drivers.2024-04-26-14.01.log.html
14:06:48 <opendevreview> Christian Berendt proposed openstack/ovn-octavia-provider stable/2023.2: Remove leftover OVN LB HM port upon deletion of a member  https://review.opendev.org/c/openstack/ovn-octavia-provider/+/917725
14:07:00 <haleyb> we promised not to break ML2 :)
14:07:24 <haleyb> and to keep an open dialogue if they discover any changes are necessary
14:08:02 <opendevreview> Christian Berendt proposed openstack/ovn-octavia-provider stable/2023.1: Remove leftover OVN LB HM port upon deletion of a member  https://review.opendev.org/c/openstack/ovn-octavia-provider/+/917726
14:08:52 <haleyb> one other thing that came up on the ML yesterday was regarding deprecating eventlet-based wsgi server
14:08:56 <haleyb> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/PY6NUNYFU63XOIZHDDBONJDHRXJMZF5J/
14:09:41 <lajoskatona> +1 for bringing this up
14:09:50 <haleyb> there were two neutron bugs listed
14:09:54 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2062009
14:10:05 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/1912359
14:10:25 <opendevreview> Takashi Kajinami proposed openstack/os-vif master: Remove old excludes  https://review.opendev.org/c/openstack/os-vif/+/917611
14:11:21 <haleyb> the first one has one patch merged, and another proposed
14:11:26 <haleyb> #link https://review.opendev.org/c/openstack/neutron/+/916112
14:11:45 <haleyb> but both are only marked as "partial fix"
14:12:51 <haleyb> i don't know Sebastian's nick, but need to figure out what else is required for that
14:13:29 <ihrachys> I *think* because there were two patches for same symptom in 2 different plugins?
14:14:12 <slaweq> but the second one is kind of a blocker which prevents us from going to uwsgi only IMO
14:14:26 <seba> heyjo, I'm seba in irc as in gerrit
14:14:34 <slaweq> and I don't think anyone is currently working on it
14:15:22 <seba> they are both marked as partial fix as only together they solve the problem described in the bug
14:15:30 <haleyb> seba: hi, regarding that bug and the patches (thanks for them)
14:15:41 <haleyb> oh, so after the second merges it will be fixed?
14:16:15 <lajoskatona> this bug also handled by that patch: https://bugs.launchpad.net/neutron/+bug/1912359 ?
14:16:25 <seba> yes, as far as I understood the problem only the two service plugins start rabbitmq rpc servers that use the same lock between native and eventlet threads
14:17:13 <haleyb> seba: ack. so since one has merged we should mark the final one as 'Closes-bug' for 2062009, otherwise we'll have to manually update the state of the bug
14:17:37 <seba> sure, I can update the review
14:17:57 <haleyb> it looks ready to merge, can probably get that in today
14:18:31 <haleyb> slaweq: and right, the OVN one doesn't have an owner
14:19:31 <slaweq> I remember lucasagomes was looking at this long time ago but I don't think he had anything done for it then
14:20:10 <opendevreview> Sebastian Lohff proposed openstack/neutron master: Start trunk plugin RPC via service framework  https://review.opendev.org/c/openstack/neutron/+/916112
14:20:37 <haleyb> So deploying with NEUTRON_DEPLOY_MOD_WSGI=True will show the failure? at least according to the bug
14:22:00 <lajoskatona> Do we have job with that setting? in periodic or something?
14:22:36 <slaweq> lajoskatona you mean ml2/ovn or ovs?
14:22:56 <lajoskatona> perhaps we need both?
14:23:06 <haleyb> i see something in zuul.d
14:23:21 <slaweq> IIRC NEUTRON_DEPLOY_MOD_WSGI=True is default in devstack
14:23:30 <slaweq> so ml2/ovs jobs are using that
14:23:38 <slaweq> but I may be wrong
14:23:44 <slaweq> I will need to check
14:24:45 <haleyb> looks like it defaults to False
14:24:52 <lajoskatona> https://codesearch.openstack.org/?q=NEUTRON_DEPLOY_MOD_WSGI&i=nope&literal=nope&files=&excludeFiles=&repos=
14:24:53 <haleyb> NEUTRON_DEPLOY_MOD_WSGI=$(trueorfalse False NEUTRON_DEPLOY_MOD_WSGI)
14:25:13 <lajoskatona> yes, in some of our jobs it is set to false anyway
14:25:38 <haleyb> i only see fullstack and functional setting it to True
14:26:48 <haleyb> so maybe there's two things we need here - 1) a non-voting or experimental job with this set to True; 2) owner for the bug
14:28:10 <lajoskatona> I can create the jobs
14:28:49 <haleyb> lajoskatona: ack, thanks, and i'm assuming/hoping they show a failure initially
14:29:48 <slaweq> so maybe I am wrong but I was almost sure we had job(s) like that already somewher
14:30:19 <haleyb> slaweq: i don't know? i didn't look in n-t-p repo
14:30:47 <lajoskatona> indeed: https://zuul.opendev.org/t/openstack/job/neutron-ovn-tempest-with-uwsgi
14:30:54 <haleyb> # TODO(lucasagomes): Re-enable MOD_WSGI after
14:30:54 <haleyb> # https://bugs.launchpad.net/neutron/+bug/1912359 is implemented
14:30:57 <lajoskatona> I don't know where it is defined
14:31:29 <haleyb> so the plugin-base job has it False due to that bug
14:31:36 <slaweq> maybe it was only funvtional and fullstack
14:31:39 <lucasagomes> haleyb, yup, we may need to work on that now that oslo is thinking about deprecating eventlet wsgi
14:32:20 <haleyb> lucasagomes: hey, thanks for joining
14:33:24 <opendevreview> Merged openstack/neutron master: reno: Update master for unmaintained/zed  https://review.opendev.org/c/openstack/neutron/+/917566
14:34:02 <haleyb> lucasagomes: do you think the work required is on the small/medium side for this?
14:34:15 <haleyb> i hadn't seen the bug until yesterday
14:34:52 <lucasagomes> haleyb, I would say medium at least
14:35:08 <lucasagomes> haleyb, we will have to move/rewrite parts of the post_fork method that we have in the driver
14:35:20 <lucasagomes> which is the bit that connects to the ovndbs, start maintenance thread etc...
14:36:01 <lucasagomes> so there's some work to do there
14:36:22 <haleyb> ok, so not trivial it seems
14:36:36 <lucasagomes> not trivial no
14:37:20 <lucasagomes> but I don't don't expect it to be too large tho, it's mostly moving code around and adapting the tests I would say (or hope for)
14:37:52 <haleyb> lucasagomes: ack
14:37:53 <opendevreview> Merged openstack/neutron-lib master: Remove old excludes  https://review.opendev.org/c/openstack/neutron-lib/+/917609
14:37:54 <opendevreview> Merged openstack/neutron-lib master: reno: Update master for unmaintained/zed  https://review.opendev.org/c/openstack/neutron-lib/+/917560
14:40:08 <haleyb> okay, well if anyone wants to take ownership of that please ping me, else i will have to start learning about it
14:40:29 * haleyb jinxed himself saying it would be a short meeting
14:40:47 <lajoskatona> :D
14:40:54 <mlavalle> LOL
14:41:27 <haleyb> lets move to bugs, thanks for the discussion here
14:41:34 <haleyb> #topic bugs
14:41:56 <haleyb> report from slaweq is here
14:42:00 <haleyb> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/P3X4LWI76V7IGDGA3BVCXD4G4G4OOZ7T/
14:42:13 <slaweq> it was easy week
14:43:08 <haleyb> i see two un-assigned bugs
14:43:15 <haleyb> #link https://bugs.launchpad.net/neutron/+bug/2063669
14:43:18 <slaweq> there are 2 unassigned, one related to ovn-octavia-provider
14:43:26 <haleyb> slaweq: go ahead
14:43:32 <slaweq> and second seems like low-hanging-fruit thing for me
14:43:55 <slaweq> related to the extensions reported in the API incorrectly
14:44:02 <slaweq> that's all
14:45:56 <haleyb> slaweq: so is that dns extension supported by OVN? or is that what needs to be checked?
14:46:57 <opendevreview> Merged openstack/networking-odl unmaintained/zed: Update .gitreview for unmaintained/zed  https://review.opendev.org/c/openstack/networking-odl/+/917549
14:47:23 <slaweq> this extension has nothing to do with ovn backend AFAIK
14:47:41 <slaweq> but it is listed as "supported" to not be filtered out when ovn backend is used
14:48:25 <opendevreview> Fernando Royo proposed openstack/ovn-octavia-provider stable/2024.1: Adding isolation to functional tests  https://review.opendev.org/c/openstack/ovn-octavia-provider/+/917719
14:48:43 <haleyb> oh, but not in extension_drivers, got it
14:49:09 <haleyb> again, if someone wants to take it and learn about extension driver framework... :)
14:50:20 <haleyb> froyo: can you take a look at the other one mentioned? https://bugs.launchpad.net/neutron/+bug/2063463 ? ovn-octavia-provider VIP issue
14:51:02 <froyo> haleyb: sure, should be during next week but I will take a look!
14:51:14 <haleyb> froyo: thanks!
14:51:24 <haleyb> any more bugs to discuss?
14:51:48 <haleyb> if not, i am deputy this week, amotoki is next week
14:52:02 <haleyb> count has been about the same
14:52:30 <haleyb> i need to find some cycles to go through our old list, link in wiki
14:52:54 <haleyb> #topic specs
14:53:18 <haleyb> just one remaining BGP spec if anyone has cycles
14:53:32 <haleyb> #link https://review.opendev.org/c/openstack/neutron-specs/+/899210
14:53:50 <haleyb> #topic community_goals
14:53:58 <haleyb> #link https://review.opendev.org/c/openstack/horizon/+/891205
14:54:13 <haleyb> i see there is a new version, thanks lajoskatona
14:54:16 <lajoskatona> haeyb: thanks for checking, and pointing to mistakes :-)
14:54:50 <haleyb> minor nits really, but i'll take a look at the update
14:55:11 <haleyb> let's get it over the finish line :)
14:55:36 <haleyb> #topic on_demand
14:55:51 <haleyb> nothing on the wiki, but does anyone have a topic to discuss?
14:56:29 <haleyb> alright, CI meeting in a few minutes, video this time i believe
14:56:38 <ykarel> yeap
14:56:45 <haleyb> thanks for attending everyone
14:56:45 <lajoskatona> +1
14:56:55 <haleyb> #endmeeting