14:02:19 <lajoskatona> #startmeeting networking
14:02:19 <opendevmeet> Meeting started Tue Oct 12 14:02:19 2021 UTC and is due to finish in 60 minutes.  The chair is lajoskatona. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:02:19 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:02:19 <opendevmeet> The meeting name has been set to 'networking'
14:02:27 <mlavalle> o/
14:02:31 <slaweq> o/
14:02:32 <bcafarel> o/
14:02:36 <haleyb> hi
14:02:39 <lajoskatona> ok, so I have to be sure to end the meeting always
14:02:41 <lajoskatona> Hi
14:02:53 <rubasov> hi
14:03:15 <lajoskatona> Let's start than
14:03:16 <isabek> hi!
14:03:22 <lajoskatona> #topic Announcements
14:03:52 <lajoskatona> Yoga cycle started, please check the cycle calnedar: https://releases.openstack.org/yoga/schedule.html
14:04:26 <mlavalle> so we will all be doing sun salutations and downward facing dos
14:04:40 <mlavalle> dogs
14:04:58 <lajoskatona> we can if you can train us :P
14:05:50 <bcafarel> 6 months of exercise :)
14:05:53 <lajoskatona> Please prepare your RFE and spec ideas and add them to https://etherpad.opendev.org/p/tc-yoga-ptg and next week we can discuss them on the PTG
14:06:41 <rubasov> wrong link?
14:06:56 <lajoskatona> https://etherpad.opendev.org/p/neutron-yoga-ptg
14:07:03 <lajoskatona> sorry that was the tc etherpad....
14:07:25 <lajoskatona> actually I have no more announcements for this week
14:07:43 <lajoskatona> Do you have anything to announce or comments / questions ?
14:07:47 <slaweq> lajoskatona: when is the deadline to add topics to the etherpad?
14:08:43 <lajoskatona> I plan to finalize the schedule Thursday
14:09:03 <lajoskatona> so before that :-)
14:09:10 <slaweq> ok
14:09:45 <lajoskatona> Ok if no more announcements than
14:09:51 <lajoskatona> #topic Bugs
14:10:01 <lajoskatona> ralonsoh was bug deputy last week: http://lists.openstack.org/pipermail/openstack-discuss/2021-October/025267.html
14:10:37 <lajoskatona> when last checked these bugs were unassigned: https://bugs.launchpad.net/neutron/+bug/1946456
14:10:44 <lajoskatona> https://bugs.launchpad.net/neutron/+bug/1946187
14:11:38 <lajoskatona> ralonsoh: do you have anything regarding the bugs from last week?
14:11:41 <slaweq> I will check https://bugs.launchpad.net/neutron/+bug/1946187
14:11:56 <slaweq> ralonsoh is off today, they have some public holiday in Spain
14:12:28 <lajoskatona> slaweq: thanks
14:12:34 <mlavalle> Columbus Day
14:12:44 <lajoskatona> slaweq: ok, thanks,
14:13:14 <lajoskatona> I saw that in google calendar, good to know that it is celebrated in Spain as well
14:14:59 <lajoskatona> the other bug  (#1946456) is for OVN, perhaps I try to fetch somebody with better ovn knowledge to check it
14:15:45 <slaweq> maybe ksambor can take a look at it? :)
14:17:27 <lajoskatona> Let's find a "volunteer" after the meeting :-)
14:18:03 <slaweq> ++
14:18:03 <lajoskatona> This week lucasgomes is the deputy, and next week jlibosva will be.
14:18:03 <lajoskatona> I sent mail to lucasgomes, so He started the guard
14:18:25 <lajoskatona> jlibosva is offline, but will ask him if he can do next week
14:18:58 <lajoskatona> and we are at the end of the list so I have to shuffle for the next period
14:19:14 <lajoskatona> Do you have question or noted for the bugs?
14:19:41 <lajoskatona> Ok, next topic
14:19:45 <lajoskatona> #topic L3 subteam
14:20:14 <lajoskatona> liuyulong: Do you have anything to discuss today?
14:21:05 <liuyulong> Hi
14:21:28 <lajoskatona> Hi liuyulong
14:21:54 <liuyulong> One small update
14:21:56 <liuyulong> #link https://review.opendev.org/c/openstack/neutron-specs/+/802854/4/specs/yoga/distributed-metadata-data-path.rst
14:22:08 <liuyulong> I added the IPv6 metadata details in this spec.
14:22:34 <lajoskatona> thanks, I will check it
14:23:08 <slaweq> liuyulong: I commented on it today
14:23:29 <liuyulong> slaweq, yes, I saw that.
14:23:34 <liuyulong> I'm testing the ovn now.
14:23:51 <liuyulong> Because I do not have too much information about how ovn metadata works...
14:24:21 <liuyulong> For the spec now, the "br-meta" is a traditional topology for neutron ml2-ovs, do specific work on specific bridge, something like tunnel bridge and various provider network bridges.
14:24:26 <slaweq> liuyulong: basically it creates ovnmeta-XXX namespace per neutron network on node
14:24:35 <liuyulong> And more things we do not want to mix the new flows with those complicated list on br-int.
14:24:52 <slaweq> and there is "distributed" port created in neutron which is used in that namespace
14:26:06 <liuyulong> So, this spec looks have some advanced progress than ovn. We will have one haproxy process for one host's VMs.
14:26:17 <liuyulong> Higher availability
14:26:44 <liuyulong> And no namespace.
14:27:07 <slaweq> but more bridges/OF rules and some restricted CIDRs used on the nodes
14:27:17 <slaweq> anyway, let's continue discussion in the spec's review
14:27:19 <liuyulong> About the port, tap-meta, only one as well.
14:27:38 <lajoskatona> slaweq: ++
14:28:42 <lajoskatona> liuyulong: do you have any L3 related bugs to discuss?
14:29:53 <liuyulong> No, I don't
14:29:57 <lajoskatona> thanks for the update regarding distribued metadata
14:30:03 <lajoskatona> #topic ryu os-ken
14:30:57 <opendevreview> Przemyslaw Szczerbik proposed openstack/neutron-lib master: Make update_qos_minbw_allocation() more generic  https://review.opendev.org/c/openstack/neutron-lib/+/813660
14:31:02 <lajoskatona> we discussed last time to check the commits in ryu which we can cherry-pick to os-ken, and slaweq and ralonsoh already added a few
14:31:09 <lajoskatona> I created and etherpad for it:
14:31:18 <lajoskatona> https://etherpad.opendev.org/p/make_os-ken_and_ryu_sync
14:32:12 <slaweq> ++
14:32:14 <slaweq> thx
14:32:40 <lajoskatona> My thought is to refresh it periodically like monthly and check if we have to act on any of the patches
14:33:11 <slaweq> we should somehow mark there patches which we checked and don't need/want to backport
14:33:20 <slaweq> I don't see that in Your "legend"
14:33:42 <lajoskatona> I started to use a primitive mark system, see the legend
14:33:55 <slaweq> yes, I see
14:34:01 <lajoskatona> sorry it is really not perfect now, but can imrpove it :-)
14:34:26 <slaweq> but You should add something to mark patches like "reviewed, we don't want them"
14:34:28 <slaweq> wdyt?
14:34:50 <lajoskatona> yeah we should
14:35:07 <lajoskatona> I filtered out patches for documentation, CI, requirements
14:36:11 <lajoskatona> so I think these can be considered to include to os-ken, but I marked with "???" the ones which I don't see the need, but would be good to have more votes on it
14:36:34 <slaweq> thx, I will check that list and will start proposing patches to os-ken
14:36:49 <lajoskatona> thanks
14:37:03 <mlavalle> maybe we should have a "deputy" in charge of this monthly update
14:37:15 <lajoskatona> perhaps I will move it to an ethercalc to have a kind of mapping to gerrit
14:37:40 <lajoskatona> mlavalle: good idea
14:37:40 <slaweq> ++
14:38:02 <lajoskatona> I can be sure to add it to the agenda and prepare it monthly
14:38:03 <mlavalle> if nobody is responsible, pretty soon it will fall over the side
14:38:15 <slaweq> mlavalle: true :)
14:39:03 <lajoskatona> it's true, I will prepare than it (and ask google calendar to warn me before :-)))
14:39:53 <mlavalle> when I became PTL, I was advided to delegate as much as possible
14:40:01 <mlavalle> advised
14:40:07 <amotoki> is ryu still maintained well? I am not sure which is better to maintain a fork or ryu itself. just my question
14:40:09 <mlavalle> I pass the advise to you
14:40:33 <slaweq> amotoki: yeah, lajoskatona added that topic to the PTG IIRC :)
14:40:38 <lajoskatona> mlavalle: thanks :-)
14:40:40 <mlavalle> maybe send a message to the ML and ask for a volunteer
14:41:28 <lajoskatona> amotoki: it is now maintained "again", and I sent mail to one of the more active developers
14:42:05 <amotoki> slaweq: lajoskatona: thanks for the info. the situation tends to change from time to time.
14:42:33 <lajoskatona> mlavalle: I can send a mail, anyway I plan to extend the list of lieutenants (https://docs.openstack.org/neutron/latest/contributor/policies/neutron-teams.html#neutron-lieutenants ) with ovsdb and os-ken
14:42:51 <mlavalle> they just pretended that they wouldn't maintain anymore to kick us out and then came back to it... LOL
14:42:53 <lajoskatona> just to be sure we know who we can ping
14:43:50 <lajoskatona> malavalle: possible
14:44:02 <mlavalle> send the message to the ML. if nobody is intrested, I'll do it
14:44:14 <lajoskatona> it seems that the opensource world is small for us
14:44:25 <slaweq> IIRC ralonsoh was voluntering to do that too :)
14:44:37 <lajoskatona> mlavalle: thanks
14:44:40 <mlavalle> in that case let ralonsoh do it
14:44:53 <lajoskatona> we have a lot of enthusiastic people
14:44:58 <slaweq> ++
14:45:25 <lajoskatona> ok, I have no more to add to the ryu-os-ken topic
14:46:02 <lajoskatona> #topic On Demand Agenda
14:46:28 <lajoskatona> On the wiki there's no topic for today, do you have anything to discuss?
14:46:39 <slaweq> I have one qq
14:46:56 <lajoskatona> let's hear it
14:46:59 <lajoskatona> read it
14:47:04 <slaweq> as we just released Xena, should we again do new releases of other stable branches?
14:47:52 <mlavalle> I think so
14:48:26 <lajoskatona> there was a mail from elodilles: http://lists.openstack.org/pipermail/openstack-discuss/2021-October/025276.html
14:48:26 <bcafarel> IIRC yes we planned to have stable point releases for these milestones
14:48:54 <lajoskatona> it's for ussuri em trainsition, but I suppose it's open to release things on other branches
14:49:06 <lajoskatona> bcafarel: thanks
14:49:20 <slaweq> lajoskatona: if we are going to EM with ussuri, we should do last one release for it before EM
14:49:27 <bcafarel> definitely
14:49:34 <lajoskatona> exactly
14:50:18 <bcafarel> previous stable releases were on Sep 9 (because of security issues), we can wait a bit and do a round for all (and prepare ussuri EM at the same time)?
14:50:28 <lajoskatona> so we can prepare the release and check what we can merge before (and keep things green)
14:51:31 <lajoskatona> bcafarel: don't know if there's an order of doing it
14:51:58 <isabek> i also have qq, when neutron-lib will be released?
14:52:21 <slaweq> isabek: we can release neutron-lib when we need it
14:52:57 <bcafarel> lajoskatona: no strict order I think, but we can check with Elod - and have a separate EM-transition release if preferred
14:53:19 <lajoskatona> bcafarel: ok
14:53:28 <isabek> will be released near future?
14:53:41 <slaweq> if there are some patches merged there which needs to be release to unblock some other development, I think we can do it even this week
14:54:25 <amotoki> isabek: do you have any waiting feature? it is usually released based on needs.
14:55:22 <isabek> yes, local ip feature
14:55:46 <slaweq> I see that Local IP API def is merged but not released yet
14:55:59 <slaweq> and some constants and exceptions for that BP
14:56:16 <slaweq> so IMO we can do new release to unblock You, wdyt lajoskatona and amotoki?
14:56:32 <amotoki> slaweq: sounds good to release it. let's do it.
14:56:54 <isabek> Thanks!
14:57:36 <lajoskatona> we have quite a few merged things so yeah we can do a release even this week
14:58:17 <lajoskatona> ok if nothing else, I try to end the meeting this time :P
14:58:24 <lajoskatona> #endmeeting