14:00:50 <ralonsoh> #startmeeting networking
14:00:50 <opendevmeet> Meeting started Tue Nov 22 14:00:50 2022 UTC and is due to finish in 60 minutes.  The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:50 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:50 <opendevmeet> The meeting name has been set to 'networking'
14:00:51 <mlavalle> o/
14:00:52 <ralonsoh> hello all
14:00:57 <bcafarel> o/
14:01:04 <rubasov> o/
14:01:16 <ralonsoh> bcafarel, busy bug week, last one
14:01:27 <lajoskatona> o/
14:02:04 <ykarel> o/
14:02:04 <ralonsoh> ok, let's start
14:02:14 <bcafarel> ralonsoh: indeed, I drew the short straw :)
14:02:21 <ralonsoh> hehehe
14:02:24 <ralonsoh> #topic announcements
14:02:30 <ralonsoh> #link https://releases.openstack.org/antelope/schedule.html
14:02:40 <ralonsoh> we are in R-17
14:02:55 <ralonsoh> in 4 weeks we have the Zed Cycle-Trailing Release Deadline
14:02:57 <ralonsoh> https://releases.openstack.org/antelope/schedule.html#a-cycle-trail
14:03:30 <ralonsoh> and just as a heads-up
14:03:38 <ralonsoh> today we don't have the CI meeting after this one
14:03:54 <ralonsoh> CI seems to be stable (apart from the live migration issue we'll discuss)
14:04:16 <ralonsoh> something else you want to add?
14:04:28 <lajoskatona> nothing from me
14:04:47 <ralonsoh> ok, let's move then
14:04:50 <ralonsoh> #topic bugs
14:05:01 <ralonsoh> last report if from bcafarel
14:05:10 <ralonsoh> #link https://lists.openstack.org/pipermail/openstack-discuss/2022-November/031249.html
14:05:20 <ralonsoh> busy week last one, but all bugs are assigned
14:05:38 <ralonsoh> I would like to discuss some of them
14:05:41 <bcafarel> yes and patches for most of them
14:05:54 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1997185
14:06:02 <ralonsoh> and the patch https://review.opendev.org/c/openstack/neutron/+/865081
14:06:21 <lajoskatona> bcafarel: I hope you drained the source of bugs :-)
14:06:24 <ralonsoh> just to comment that we have the possibility of spawning a DHCP agent with OVN
14:06:26 <ralonsoh> hehehe
14:06:32 <bcafarel> :)
14:06:48 <ralonsoh> the next one is
14:06:51 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1996741
14:07:07 <ralonsoh> I have just one question: do we need to discuss it in the drivers meeting?
14:07:23 <ralonsoh> the scope is small I think and doesn't need spec
14:07:26 <ralonsoh> patch: https://review.opendev.org/c/openstack/neutron/+/861719
14:08:05 <mlavalle> probably we don't need to discuss iton Friday
14:08:17 <bcafarel> I wondered about it too, scope in patch did not look that large
14:08:28 <ralonsoh> so the discussion in the gerrit patch is enough
14:08:38 <mlavalle> I think so
14:08:42 <ralonsoh> perfect
14:08:51 <lajoskatona> agree, I actually asked if there is a bug or RFE, but not asked directly for RFE :-)
14:08:59 <mlavalle> just maybe make sure it comes with some sort of documentation
14:09:12 <lajoskatona> 1
14:09:14 <lajoskatona> +1
14:09:15 <ralonsoh> ^^ right, some docs are needed for sure (and a reno)
14:09:22 <mlavalle> which I don't see in the patch
14:09:34 <ralonsoh> please, leave a comment
14:09:39 <mlavalle> will do
14:09:44 <ralonsoh> thanks
14:09:45 <mlavalle> I'll review it today
14:09:57 <ralonsoh> next one is
14:09:59 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1996788
14:10:09 <ralonsoh> I did a small triage (or analysis)
14:10:22 <ralonsoh> but I would like you to check it too, if you have time
14:10:32 <ralonsoh> IMO, this is not an OVS agent bug
14:10:46 <ralonsoh> but maybe you can add new arguments (in favor or not)
14:11:35 <ralonsoh> and the last one
14:11:35 * mlavalle will look at the end of the meeting during the CI time slot
14:11:41 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/1997025
14:11:53 <ralonsoh> I'm working on this one, I think I've identified the issue
14:12:03 <ralonsoh> the subports are not set as active when migrated
14:12:27 <ralonsoh> because we (OVS agent) update the binding with the source host
14:12:34 <mlavalle> ahhh
14:12:50 <ralonsoh> this is because now we execute this action before the parent port have been updated (by Nova)
14:12:53 <mlavalle> I had gotten to the subports not becoming active, but not the cause
14:13:08 <ralonsoh> so I'm testing now in https://review.opendev.org/c/openstack/neutron/+/865039
14:13:30 <ralonsoh> of course, the last PS (that was supposed to solve this issue), didn't
14:13:53 * mlavalle will keep and I on it and can help if needed
14:14:02 <ralonsoh> should we disable the test during the testing?
14:14:14 <ralonsoh> just to allow the CI to merge new patches
14:14:15 <mlavalle> yeah, to give us time
14:14:26 <ralonsoh> the problem is that this test is in tempest
14:14:32 <ralonsoh> we don't directly control this repo
14:14:47 <mlavalle> I know, but we skip some other test cases anyway
14:15:00 <ralonsoh> I'll propose a patch today
14:15:03 <mlavalle> in those classes
14:15:12 <lajoskatona> I don;t think QA team will aceept the reasoning
14:15:25 <ralonsoh> lajoskatona, don't they?
14:15:32 <lajoskatona> I mean they will accept, sorry
14:15:37 <ralonsoh> ah ok
14:15:45 <mlavalle> yeah, it's for a few days
14:15:49 <ralonsoh> is there any other option? for example disabling this test via zuul?
14:15:50 <mlavalle> at most
14:15:52 <ralonsoh> is that possible?
14:16:41 <ralonsoh> tempest_test_regex
14:16:42 <bcafarel> with tempest_exclude_regex maybe?
14:16:46 <lajoskatona> the black_regex or what's the name of it now, perhaps
14:16:49 <ralonsoh> exactly
14:16:55 <bcafarel> (we use it in old stable branches to skip some unstable tests)
14:17:03 <ralonsoh> ok, I'll try first back_regex
14:17:19 <ralonsoh> this is just to keep things in Neutron only
14:17:30 <ralonsoh> if that doesn't work, I'll propose a patch for tempest
14:17:53 <lajoskatona> thanks ralonsoh
14:18:02 <ralonsoh> and last comment in this topic, the usual heads-up
14:18:04 <ralonsoh> this week lajoskatona is the deputy, next week will be slaweq
14:18:31 <ralonsoh> next topic
14:18:35 <ralonsoh> #topic os-ken
14:18:41 <ralonsoh> https://review.opendev.org/q/project:openstack%252Fos-ken+status:open
14:18:56 <ralonsoh> related to the ryu patches I talked about last week
14:19:09 <ralonsoh> if you have time, please check them
14:19:31 <lajoskatona> I will check them
14:19:32 <ralonsoh> ah, about https://review.opendev.org/c/openstack/os-ken/+/852821
14:19:40 <ralonsoh> the other patch was abandoned
14:19:48 <ralonsoh> so I think we can approve it
14:20:21 <ralonsoh> let's move to the next topic
14:20:24 <ralonsoh> #topic specs
14:20:37 <ralonsoh> first of all, THANK YOU ALL for taking care of the active specs
14:20:50 <ralonsoh> the 3 active specs were merged last week
14:20:58 <ralonsoh> #link https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open
14:21:00 <bcafarel> nice
14:21:17 <ralonsoh> there is only one, but still waiting to have some comments addressed
14:21:42 <ralonsoh> once the author continues the work, I'll add it to the Neutron meeting too
14:22:08 <ralonsoh> again, thank you all
14:22:12 <ralonsoh> next topic
14:22:26 <ralonsoh> #topic community-goals
14:22:33 <ralonsoh> the first one, S-RBAC
14:22:51 <ralonsoh> there is one pending patch to be merged
14:22:51 <ralonsoh> https://review.opendev.org/c/openstack/tempest/+/614484
14:23:35 <ralonsoh> and we have a new related S-RBAC patch: https://review.opendev.org/c/openstack/neutron/+/865040
14:24:00 <ralonsoh> but from Neutron point of view, we have completed the S-RBAC migration
14:24:25 <ralonsoh> once we have tempest using those new roles, we'll close the community goal
14:25:06 <ralonsoh> next goal is the migration to Ubuntu 22.04
14:25:13 <ralonsoh> the last patch to be merged is https://review.opendev.org/c/openstack/neutron/+/862492
14:25:18 <ralonsoh> all the dependencies are merged too
14:25:33 <ralonsoh> this one is failing because of the CI issue (trunk live migraiton)
14:25:38 <ralonsoh> but looks fine
14:25:54 <ralonsoh> so thank you all for working on this!
14:26:20 <ralonsoh> the next goal is the deprecation of neutron-legacy
14:26:22 <ralonsoh> #link https://bugs.launchpad.net/devstack/+bug/1996748
14:26:36 <ralonsoh> there are you patches submitted (links in the bug)
14:26:42 <ralonsoh> slaweq needs to check the CI
14:27:08 <ralonsoh> but please add these two patches to your review piles
14:27:24 <lajoskatona> cool, I missed that slaweq started to work on this
14:27:49 <ralonsoh> yeah and I think (slaweq will confirm) that those 2 patches are enough (at least in devstack)
14:28:34 <ralonsoh> last topic is the fix of zuul errors
14:28:42 <ralonsoh> I have more questions than solutions
14:28:43 <ralonsoh> first
14:28:47 <ralonsoh> https://review.opendev.org/q/project:openstack%252Fnetworking-midonet+status:open
14:28:54 <ralonsoh> I tried to fix the midonet issues
14:29:03 <ralonsoh> but the CI is 100% broken
14:29:10 <ralonsoh> what should I do there?
14:29:30 <mlavalle> reach out to whomever is responsible of maintaining it
14:29:49 <ralonsoh> yamamoto is not taking care of this project anymore
14:29:52 <ralonsoh> at least in U/S
14:29:58 <frickler> mark everything n-v and just fix config errors?
14:30:12 <ralonsoh> what is n-v?
14:30:25 <ralonsoh> ah
14:30:30 <ralonsoh> non voting
14:30:34 <ralonsoh> perfect, I'll do it
14:30:46 <lajoskatona> if nobody maintains midonet can't we retire it?
14:30:53 <mlavalle> also send a message to the ML
14:30:53 <ralonsoh> that's the second part
14:31:01 <ralonsoh> to retire midonet
14:31:09 <mlavalle> indicating that if nobody steps up, we will retire the project
14:31:25 <ralonsoh> I'll send a mail (1) asking for mantainers and (2) if not, anouncing the retirement
14:31:29 <ralonsoh> do you agree?
14:31:33 <mlavalle> I suggest giving a period of time for somebody to step up
14:31:37 <ralonsoh> for sure
14:31:41 <ralonsoh> that's mandatory
14:31:50 <mlavalle> nd if nobody does, retire the project
14:32:07 <mlavalle> and yes, I agree in potentisally retiring the project
14:32:25 <ralonsoh> perfect, I'll do it (providing reasonable time slots_
14:32:31 <mlavalle> +2
14:32:36 <lajoskatona> +1, thanks
14:32:43 <ralonsoh> the next question was related to taas, lajoskatona
14:32:53 <ralonsoh> just to be clear (related to the meeting last week)
14:33:27 <ralonsoh> should I propose a EOL patch in releases project for those versions pointing to x/taap-as-a-service?
14:34:00 <ralonsoh> horribly explained...
14:34:06 <ralonsoh> in other words
14:34:16 <lajoskatona> as I remember as there is no yaml file for those old branches, these branches should be deleted and tagged manually
14:34:19 <ralonsoh> taas doesn't have a EOL release patch in some version
14:34:34 <ralonsoh> how is that possible? how can we do this?
14:35:35 <lajoskatona> this was the discussion: https://meetings.opendev.org/meetings/releaseteam/2022/releaseteam.2022-11-18-14.01.log.html#l-111
14:35:52 <lajoskatona> release team and gerrit admins has right to do this I think
14:36:02 <frickler> iiuc the release managers agreed to do that task when requested
14:36:41 <lajoskatona> frickler: yeah, I understood also that way
14:36:42 <ralonsoh> ok, I wasn't sure about this
14:36:58 <ralonsoh> this is why I asked about it today
14:37:18 <lajoskatona> good idea, I forgot about it and better to have record of iit here also
14:37:27 <ralonsoh> perfect, so this issue won't bother us again
14:37:38 <frickler> so maybe prepare a mail with the hashes to be tagged and branches to be deleted
14:37:39 <lajoskatona> yes
14:37:50 <lajoskatona> frickler: ok, I send it
14:37:55 <ralonsoh> lajoskatona, thanks!
14:38:35 <ralonsoh> I'll continue checking the zuul errors once we have both projects (midonet and taas) fixed
14:39:06 <ralonsoh> those are responsible of most of the errors related to Neutron
14:39:33 <ralonsoh> and that's all I have in this topic, something else?
14:40:06 <ralonsoh> cool, let's move on
14:40:10 <ralonsoh> #topic on-demand
14:40:30 <ralonsoh> I have nothing in the agenda so please, if you need to add something, this is the best moment
14:40:42 <mlavalle> nothing from me
14:40:42 <opendevreview> Merged openstack/os-vif stable/xena: Move mtu update request into ovsdb transaction  https://review.opendev.org/c/openstack/os-vif/+/863995
14:40:58 <lajoskatona> nothing from me
14:41:09 <bcafarel> a quick note to highlight a recent issue updated: https://bugs.launchpad.net/neutron/+bug/1969615
14:41:33 <bcafarel> just a heads-up, some users may hit an OVS kernel bug depending on their versions https://www.mail-archive.com/ovs-discuss@openvswitch.org/msg08762.html
14:41:51 <bcafarel> (reported by mnaser who confirmed it is fixed with current versions)
14:42:09 <ralonsoh> are we executing any of the affected versions in our CI?
14:42:13 <bcafarel> but as we do not have a strict check on OVS version, communicating this in case someone else hits the issue
14:42:44 <bcafarel> from what I checked no, all good in CI, kernels have the fix
14:42:58 <ralonsoh> we are using 2.13, 2.15 and specific hashes
14:43:12 <ralonsoh> in any case I'll check if those hashes belong to affected versions
14:43:38 <ralonsoh> should we document this issue? or should we provide a check/upgrade script?
14:44:57 <bcafarel> hmm I wonder if this can be checked at runtime (may be dependant on Linux distro)
14:45:50 <lajoskatona> we have examples of such check for other 3rd party tools which we use in neutron, so theoreticslly
14:46:10 <lajoskatona> the question what to do if we detect the bad version of OVS? stop the agent?
14:46:26 <ralonsoh> no, just emit a warning
14:46:33 <ralonsoh> and refer to this bug
14:46:41 <lajoskatona> ok
14:47:06 <ralonsoh> I'll see how we can address this from Neutron, just to make the user aware of this problem
14:47:18 <frickler> bcafarel: regarding n-d-r+OVN which seems to be fixed by ralonsoh's patch, do you want to update https://review.opendev.org/c/openstack/neutron/+/738641 to become mergeable or should I do it?
14:47:31 <frickler> testing in https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/814055 went smoothly
14:48:35 <bcafarel> frickler: oh nice! feel free to take over the patch as "real" one (there is only neutron/common/ovn/extensions.py list additions anyway)
14:48:55 <frickler> bcafarel: o.k., will do
14:49:10 <bcafarel> thanks!
14:49:16 <ralonsoh> I'll try to merge https://review.opendev.org/c/openstack/neutron/+/864051 ASAP
14:49:55 <ralonsoh> thanks bcafarel from bringing the OVS bug
14:50:28 <ralonsoh> ok, I think we can close the meeting now
14:50:37 <ralonsoh> thank you for attending
14:50:42 <bcafarel> o/
14:50:43 <ralonsoh> #endmeeting