14:00:46 <ralonsoh> #startmeeting networking
14:00:46 <opendevmeet> Meeting started Tue May  9 14:00:46 2023 UTC and is due to finish in 60 minutes.  The chair is ralonsoh. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:46 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:46 <opendevmeet> The meeting name has been set to 'networking'
14:00:49 <ralonsoh> hello all
14:01:02 <slaweq> o/
14:01:06 <rubasov> o/
14:01:10 <ykarel> o/
14:01:14 <mtomaska> o/
14:01:32 <ralonsoh> let's start
14:01:39 <ralonsoh> #topic announcements
14:01:40 <elvira> o/
14:01:47 <ralonsoh> #link https://releases.openstack.org/bobcat/schedule.html
14:02:00 <ralonsoh> we are finally in bobcat milestone 1
14:02:09 <ralonsoh> #link https://releases.openstack.org/bobcat/schedule.html#b-1
14:02:35 <ralonsoh> there are some patches related
14:02:38 <ralonsoh> #link https://review.opendev.org/c/openstack/releases/+/882599
14:03:05 <ralonsoh> and at the same time I'm releasing n-lib 3.6.0
14:03:08 <ralonsoh> #link https://review.opendev.org/c/openstack/releases/+/881231
14:03:17 <ralonsoh> so please check if something is missing and comment in the patches
14:03:20 <ralonsoh> and ping me please
14:03:37 <ralonsoh> Vancouer: https://etherpad.opendev.org/p/neutron-vancouver-2023
14:04:00 <ralonsoh> the forum proposal presented by slaweq was accepted
14:04:07 <ralonsoh> but it will take only 30 mins
14:04:30 <ralonsoh> this is because of the packed agenda during the 3 days the meeting will take
14:04:38 <ralonsoh> slaweq, can you add something there?
14:04:47 <slaweq> not much actually
14:04:55 <ralonsoh> on how we'll collect info and then prepare the PTG
14:05:03 <slaweq> we will have "meet and greet" session during forum
14:05:20 <slaweq> and as ralonsoh said - it's 30 minutes, as all other project related forum sessions
14:05:50 <opendevreview> yatin proposed openstack/neutron-tempest-plugin master: Revert "Revert "Update nested-virt testing for the 2023.1 cycle""  https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/882719
14:06:10 <ralonsoh> I hope that will be enough to gather all the information from the users
14:06:23 <slaweq> it will be Tuesday morning so we will have time to prepare maybe some follow up agenda for PTG part which will be Wednesday and Thursday
14:06:24 <slaweq> that's our plan at least :)
14:07:24 <ralonsoh> ok, let's move to the next topic then
14:07:35 <ralonsoh> #topic bugs
14:07:48 <ralonsoh> the report from last week
14:07:51 <ralonsoh> #link https://lists.openstack.org/pipermail/openstack-discuss/2023-May/033596.html
14:08:04 <ralonsoh> there are some bugs not assigned
14:08:06 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/2018474
14:08:12 <ralonsoh> Unnecessary agent resource version
14:08:32 <ralonsoh> this could be an optimization in the agent RPC communication
14:08:56 <ralonsoh> anyone willing to dig into this code, please check the description
14:09:08 <ralonsoh> the next one is
14:09:10 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/2018599
14:09:14 <ralonsoh> Disable config option use_random_fully does not work
14:09:45 <ralonsoh> I still need to check that in the code
14:10:07 <ralonsoh> but if I'm not wrong, that was added in https://review.opendev.org/q/Idfe5e51007b9a3eaa48779cd01edbca2f586eee5
14:10:54 <ralonsoh> but I think the problem is more related to an upgrade/update procedure
14:11:06 <ralonsoh> where one of the nodes was not properly configured
14:11:26 <mlavalle> ralonsoh: I'll look at 2018474
14:11:26 <ralonsoh> I'm not in favor of disabling a config option just because an update went wrong due to a config issue
14:11:34 <ralonsoh> mlavalle, thanks
14:12:00 <ralonsoh> so I'll comment in LP#2018599 but please feel free to do the same
14:12:19 <ralonsoh> if there are clashing opinions, we can always bring this topic to the drivers meeting
14:12:54 <ralonsoh> and that's all I have
14:13:03 <ralonsoh> something else you want to bring here??
14:13:33 <ralonsoh> ok, let's move then
14:13:36 <ralonsoh> #topic specs
14:13:48 <ralonsoh> we have 2 active Neutron specs
14:13:51 <ralonsoh> #link https://review.opendev.org/q/project:openstack%252Fneutron-specs+status:open
14:13:58 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron-specs/+/882151
14:14:02 <ralonsoh> ERSPAN for tap-as-a-service
14:14:42 <ralonsoh> that was discussed a couple of weeks ago
14:14:53 <ralonsoh> so please spend some time reviewing it
14:15:04 <mlavalle> I'll review both this week
14:15:08 <ralonsoh> next one is
14:15:11 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron-specs/+/882272
14:15:15 <ralonsoh> Port extension to create hardware offloaded ports
14:15:35 <ralonsoh> trivial feature, commented during the PTG and approved recently
14:15:55 <ralonsoh> the last one, from Nova, is
14:15:58 <ralonsoh> #link https://review.opendev.org/c/openstack/nova-specs/+/859290
14:15:59 <mlavalle> the one about dvr on openflow surprises me a bit. did we discuss that one?
14:16:07 <ralonsoh> this is related to Napatech LinkVirt SmartNICs
14:16:19 <ralonsoh> mlavalle, no, this is not an active one
14:16:29 <mlavalle> ahh, ok, good to know
14:16:41 <mlavalle> thanks for the clarification
14:17:06 <ralonsoh> so about the Nova spec, we commented during the PTG that these guys want to support this NICs in Nova and Neutron
14:17:23 <ralonsoh> the Neutron code should be easy but the main concern is the testing part
14:17:52 <sahid> i think this one is good enough from neutron pov, nova are waiting for you to +1, no?
14:17:57 <ralonsoh> so please check the spec, focus on the Neutron implementation and let's keep an eye on the external CI these folks are implementing (if I'm not wrong)
14:18:23 <ralonsoh> sahid, to be honest, I didn't have a minute in the last 2 weeks to review it
14:18:43 <ralonsoh> so this is in my TODO list for tomorrow morning (and lajoskatona's one)
14:19:01 <sahid> ++
14:19:19 <ralonsoh> in any case, I don't provide any approval, that should became from Neutron cores
14:19:33 <ralonsoh> being PTL doesn't provide me any tech improvement
14:20:13 <ralonsoh> ok, I think that's all in this topic
14:20:22 <ralonsoh> something else you want to discuss here?
14:20:50 <ralonsoh> ok, let's move
14:21:06 <ralonsoh> #topic community_goals
14:21:12 <ralonsoh> 1) sRBAC
14:21:23 <ralonsoh> we have some patches to be reviewed
14:21:32 <ralonsoh> related to qos and fips
14:21:38 <slaweq> I still need to verify qos one
14:21:44 <slaweq> why one test is failing there
14:22:02 <ralonsoh> yeah, I commented in the patch
14:22:08 <slaweq> thx
14:22:18 <ralonsoh> the list of patches is
14:22:20 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron/+/882688
14:22:28 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron/+/882414
14:22:35 <slaweq> and I would also like to thanks ccamposr who is doing great job with testing of this RBACs in our d/s product
14:22:37 <ralonsoh> #link https://review.opendev.org/c/openstack/neutron/+/882691
14:22:47 <ralonsoh> for sure! good catch
14:22:49 <slaweq> and he's the one who is finding all those small issues :)
14:23:10 <ralonsoh> slaweq, apart from this bug, what about enabling by default sRBAC?
14:23:21 <slaweq> it's enabled
14:23:31 <slaweq> last week we merged https://review.opendev.org/c/openstack/neutron/+/879827
14:24:16 <slaweq> we also merged https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/879828 to change our ci jobs
14:24:17 <slaweq> but then gmann proposed revert of this change
14:24:18 <ralonsoh> Ihttps://review.opendev.org/c/openstack/neutron-tempest-plugin/+/882518
14:24:25 <ralonsoh> I'm talking about this patch
14:24:40 <slaweq> but main part is done - neutron by default is now using new policies
14:25:25 <ralonsoh> yes but why this revert?
14:25:39 <ralonsoh> just to make this public
14:26:15 <slaweq> so after https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/879828 was merged, I though that we will use new defaults in all jobs and old defaults in one specific job
14:26:42 <slaweq> but it seems that it's not true and devstack was setting enforce_new_defaults=False explicitly in all jobs
14:27:02 <ralonsoh> so when will update the devstack flag?
14:27:06 <slaweq> so gmann reverted it and we still have only one job which is testing new defaults
14:27:51 <slaweq> I think it will be in 2024.1 cycle
14:28:05 <ralonsoh> perfect, so next release
14:28:09 <slaweq> that's what gmann wrote in https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/882518
14:28:29 <ralonsoh> in any case, as you mentioned, we are already testing that in our CI
14:28:39 <slaweq> yes
14:28:46 <ralonsoh> that's for the time spent on this
14:29:05 <ralonsoh> ah, once merged (the 3 mentioned patches) we'll need to backport them
14:29:15 <slaweq> yes, up to stable/zed
14:29:20 <ralonsoh> thanks
14:29:21 <slaweq> I will do it for sure
14:29:54 <ralonsoh> lajos is not here today so we'll skip the neutron client migration
14:30:00 <ralonsoh> but I would like to add a new topic here
14:30:13 <ralonsoh> 2) Ubuntu Jammy migration
14:30:38 <ralonsoh> ykarel, do you have any update related to the vexxhost issues?
14:30:49 <ralonsoh> #link https://bugs.launchpad.net/neutron/+bug/2017992
14:30:58 <ykarel> ralonsoh, no last update was technical team looking into the issue
14:31:16 <ralonsoh> but you are pushing again https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/882719
14:31:29 <ralonsoh> (we have these nodes disabled, I know)
14:31:32 <ykarel> yes pushed today and that should be ok to move with
14:31:43 <ralonsoh> perfect then
14:32:06 <ralonsoh> so are we moving the py310 during this release?
14:32:12 <ralonsoh> or that will be postponed
14:32:26 <ralonsoh> moving to*
14:33:24 <ykarel> aren't we already doing that?
14:33:44 <ralonsoh> not for us, we are still running in focal (py38)
14:34:04 <ralonsoh> so one of the reasons to move to jammy is to use py310
14:34:05 <ykarel> ok ok i meant non tempest jobs
14:34:42 <ykarel> so yes we can say moving to jammy this release for rest jobs
14:34:47 <ralonsoh> yes, we do
14:34:53 <ralonsoh> only tempest is missing
14:35:39 <ralonsoh> ok so I'll keep an eye on https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/882719
14:35:50 <ralonsoh> ^^ and anyone else
14:36:03 <ykarel> there are some other patches to allow running jobs in non nested nodes https://review.opendev.org/q/topic:libvirt-tb-cache-size , which we can followup after ^
14:36:04 <ralonsoh> this is important, we should not delay this migration
14:37:20 <ralonsoh> but the performance is terrible, right?
14:37:44 <ykarel> yes if compared to nested-virt nodes
14:38:32 <ralonsoh> ok, let's check first the virt nested nodes and your n-t-p patch
14:38:35 <ralonsoh> ykarel, thanks!
14:38:59 <ralonsoh> and that's all in this topic
14:39:05 <ralonsoh> #topic on_demand
14:39:12 <ralonsoh> something you want to comment?
14:39:56 <ralonsoh> please remember the CI meeting is in 20 mins, in this channel
14:40:00 <ralonsoh> see you in a while
14:40:06 <ralonsoh> #endmeeting