14:00:55 <mlavalle> #startmeeting networking
14:00:56 <openstack> Meeting started Tue Nov  6 14:00:55 2018 UTC and is due to finish in 60 minutes.  The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:57 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:01:00 <openstack> The meeting name has been set to 'networking'
14:01:07 <annp_> hi
14:01:08 <lajoskatona> o/
14:01:11 <slaweq> o/
14:01:14 <bcafarel> hello
14:01:16 <haleyb> hi
14:01:26 <mlavalle> hello everybody
14:01:35 <rubasov> hello
14:01:36 <amotoki> hi
14:01:49 <kailun_> hi
14:02:18 <mlavalle> #topic Annoucements
14:02:35 <mlavalle> We are in the Stein-2 part of the cycle
14:03:03 <mlavalle> which will end the week of January 7th to 11th
14:03:23 <mlavalle> Next week is the Summit in Berlin
14:03:29 <mlavalle> I'll be there
14:04:49 <mlavalle> since next week the meeting is on Monday, I think I can run it from Berlin
14:05:18 <slaweq> but it will be late in night for You then :)
14:05:28 <bcafarel> jetlag may help for once :)
14:05:48 <mlavalle> amotoki: let's talk about stable branches at the end of the meeting, ok?
14:05:55 <amotoki> mlavalle: sure
14:06:07 <mlavalle> any other announcements?
14:06:10 <amotoki> pike, queens and rocky stable updates have been released.
14:06:28 <mlavalle> amotoki: a that's great
14:06:30 <mlavalle> Thanks!
14:06:55 <amotoki> ocata is in EM phase, so we need to check the release team. I will care it.
14:07:17 <mlavalle> bcafarel: did we merge all we wanted to merge in Ocata?
14:08:03 <bcafarel> mlavalle: was checking that
14:08:26 <bcafarel> https://review.openstack.org/#/c/594559/ now has Z+1, and is the only one left open
14:08:39 <amotoki> Apart from tagging of a release we can still merge anything needed in ocata.
14:09:02 <amotoki> this is what EM (extended maintenance) means.
14:09:15 <mlavalle> amotoki: ack
14:09:42 <mlavalle> lajoskatona: would you reconsider your vote in https://review.openstack.org/#/c/594559?
14:09:52 <mlavalle> we are trying to wrap up Ocata
14:10:03 <mlavalle> for EM
14:10:09 <lajoskatona> mlavalle: yes, I just checked and removed -1, as now all the newer branches has this patch as I see
14:10:22 <mlavalle> lajoskatona: thanks :-)
14:10:36 <mlavalle> ok, I think we can move on then
14:10:54 <mlavalle> #topic Blueprints
14:10:58 <lajoskatona> mlavalle: no problem
14:11:40 <mlavalle> These are the blueprints we are targeting currently: https://launchpad.net/neutron/+milestone/stein-1
14:11:52 * mlavalle will roll them over to Stein-2 today
14:12:07 <mlavalle> any updates on blueprints?
14:12:34 <kailun_> Hi team, I'd like to raise your attention for https://review.openstack.org/599980/
14:13:21 <kailun_> This is the updated spec for StarlingX proposal - network segment range management
14:13:48 <mlavalle> kailun_: thanks. I see you addressed the comments I left a couple of weeks ago
14:13:59 <mlavalle> I will go over it today again
14:14:16 <mlavalle> thanks for reminding us
14:14:51 <slaweq> kailun_: I will back to it also (probably tomorrow)
14:15:06 <kailun_> I also addressed some comments from slawek, nate and hongbin
14:15:25 <mlavalle> yeah I can see that
14:15:25 <kailun_> thank you all for your comments
14:16:05 <kailun_> BTW, just one question for bp
14:16:34 <kailun_> do we have a ddl for feature/spec freeze, for stein for instance?
14:16:49 <mlavalle> no, we don't do that in Neutron
14:17:00 <mlavalle> it is practice in other projects, but not here
14:17:19 <mlavalle> we continue reviewing specs throughout the cycle
14:17:30 <kailun_> got it, thanks miguel :)
14:17:41 <mlavalle> of course, if you want the corresponding code to merge in Stein
14:17:53 <mlavalle> the sooner you get your spec approved, the better
14:18:16 <kailun_> sure, understood :)
14:18:31 <mlavalle> there is always a rush of reviews at the end of the cycle
14:18:55 <mlavalle> and your core team members also like to sleep at the end of the cycle ;-)
14:19:14 <kailun_> ;)
14:19:25 <mlavalle> I want to remind people of this blueprint: https://blueprints.launchpad.net/neutron/+spec/port-mirroring-sriov-vf
14:20:22 <mlavalle> The corresponding spec is here: https://review.openstack.org/#/c/574477/
14:20:45 <mlavalle> and the code is here: https://review.openstack.org/#/c/603501/
14:21:08 <mlavalle> please take a look if you can help revieiwing this code
14:22:09 <mlavalle> I don't see munish around
14:22:24 <mlavalle> but I'll help him getting this code revieiwed
14:22:44 <mlavalle> any other blueprints we should discuss?
14:23:37 <mlavalle> ok, let's move on
14:23:45 <mlavalle> #topic Community goals
14:23:55 <amotoki> hi Regarding policy-in-code, I have a question I would like to ask the team.
14:24:08 <amotoki> We have "context_is_admin":  "role:admin or user_name:neutron" in policy.json in DevStack environment.
14:24:15 <amotoki> This is specific to DevStack.  https://github.com/openstack-dev/devstack/blob/master/lib/neutron-legacy#L705
14:24:24 <amotoki> most of you are not aware of it though.
14:24:40 <amotoki> This was added in https://review.openstack.org/#/c/108167/
14:24:41 <amotoki> "user_name:neutron" allows advanced services to behave as neutron admin, but what I am not sure is whether we still need it.
14:24:54 <amotoki> is_advsvc should do this and hopefully we can drop it.
14:25:01 <mlavalle> long time ago
14:25:14 <amotoki> if someone knows the current situation some advise would be appreciated.
14:26:04 <mlavalle> mhhh, I don't know
14:26:10 <amotoki> my current plan is to modify devstack code rather than dropping it immediately.
14:26:10 <mlavalle> can we test it?
14:26:14 <slaweq> there was even some fix recently for advsvc: https://review.openstack.org/#/c/609633/
14:26:30 <amotoki> I also would like to drop it as a separate patch
14:26:48 <slaweq> I think we should at least try to use this instead of base on user name :)
14:27:18 <amotoki> slaweq: yeah, that helps us. I am not sure this is enough...
14:27:40 <amotoki> this is specific to DevStack. we can test it locally
14:27:46 <slaweq> amotoki: so IMO it's good moment to test this advsvc and fix it if necessary :)
14:27:54 <amotoki> totally agree
14:27:57 <mlavalle> ++
14:28:27 <amotoki> thanks for the advise
14:28:52 <slaweq> about upgrade check, base patch is merged already so we have "neutron-status upgrade check" cli already
14:29:25 <slaweq> recently I pushed second patch https://review.openstack.org/#/c/615196/ to add support for 3rd party and stadium projects checks based on entry_points
14:29:46 <slaweq> I need to fix docs there but except that, please review it and tell me what You think about it
14:31:19 <amotoki> have we agreed that neutron-status upgrade command will cover all stadium projects and 3rd party projects?
14:32:24 <slaweq> amotoki: I was talking about that with mriedem for sure and we think that it's good idea to allow sub projects to add own checks, especially that such projects may have own stuff in neutron db
14:32:34 <lajoskatona> amotoki: as I remember the decision was to use stevedore for plugin detection
14:33:07 <slaweq> for now I used  pkg_resources because it's the same way how it's done in alembic migrations
14:33:14 <amotoki> slaweq: thanks for discussing it. I think it is nice that neutron-status cover all related projects.
14:33:17 <slaweq> but later I can switch both places to use stevedore
14:33:34 <lajoskatona> slaweq: ok
14:34:14 <amotoki> from operator perspective, it is too much to run "status" command for all projects, so the current propsoed way looks fine to me
14:34:43 <slaweq> amotoki: thx
14:34:51 <amotoki> I just wanted to check the direction, thanks
14:35:53 <mlavalle> anything else on community goals
14:35:55 <mlavalle> ?
14:36:13 <bcafarel> on py3, I updated https://review.openstack.org/#/c/577383/ to switch functional to python3
14:36:14 <slaweq> amotoki: it was also mentioned by mriedem in his recent update: http://lists.openstack.org/pipermail/openstack-dev/2018-November/136281.html
14:36:47 <bcafarel> we still have that subunit.parser_StringException popping in when running all tests (individually they work fine)
14:36:56 <bcafarel> http://logs.openstack.org/83/577383/10/check/neutron-functional/461e8ad/logs/testr_results.html.gz
14:37:34 <mlavalle> thanks for the update bcafarel
14:37:46 <mlavalle> anything else on community goals?
14:38:09 <mlavalle> ok
14:38:38 <mlavalle> before moving on, liuyulong reminded me that the code for this blueprint is ready for reviews: https://blueprints.launchpad.net/neutron/+spec/router-gateway-ip-qos
14:39:04 <mlavalle> This is the server side: https://review.openstack.org/#/c/424468/
14:39:18 <liuyulong> My thoughts is to finish it at stein-1
14:39:24 <mlavalle> and this is the agent side: https://review.openstack.org/#/c/568526/
14:39:36 <mlavalle> so let's take a look at those patches
14:40:05 <liuyulong> Do not give other reviews much work/pressure for next stein cycles. : )
14:40:06 <mlavalle> next time, liuyulong, please feel free to bring your blueprints up during the Blueprints topic :-)
14:40:23 <liuyulong> s/reviews/reviewers
14:40:24 <mlavalle> ok, moving on
14:40:33 <mlavalle> #topic Bugs
14:40:34 <liuyulong> OK, thanks
14:41:23 <mlavalle> Our bugs deputy last week was Zhaobo
14:41:34 <mlavalle> he sent this bugs report: http://lists.openstack.org/pipermail/openstack-dev/2018-November/136265.html
14:42:13 <mlavalle> The First one is already being addressed by Federico Ressi
14:42:54 <amotoki> the ML archive is not easy to read :(   the rendered version is here http://paste.openstack.org/show/734256/
14:43:13 <slaweq> thx amotoki :)
14:44:18 <hle2> mlavalle, sorry to check, for #https://bugs.launchpad.net/neutron/+bug/1800599, do you prefer it as RFE or bug?
14:44:18 <openstack> Launchpad bug 1800599 in neutron "[RFE]Neutron API server: unexpected behavior with multiple long live clients" [Undecided,New]
14:45:12 <mlavalle> hle2: don't worry about clasifying it
14:45:26 <mlavalle> I'll take a look and re-classify if necessary
14:45:49 <amotoki> hle2: does it happen only when eventlet server is used?
14:45:55 <hle2> mlavalle, Thanks much!
14:46:11 <hle2> Yes, eventlet server will cause such issue
14:46:11 <amotoki> does it happen when WSGI apache integration is used too?
14:46:19 <amotoki> i think it depends on our choice.
14:46:54 <amotoki> because other projects are being migrated to apache WSGI integration and neutron has not migrated to it.
14:47:22 <hle2> do you mean a new web server for neutron?
14:47:27 <amotoki> hle2: yes
14:47:57 <hle2> Is there any timeline or plan that neutron will mograte to it?
14:48:24 <amotoki> hle2: actually WSGI support has been implemented in Rocky release and there is devstack support
14:48:42 <amotoki> we just haven't switched the default configuration in devstack yet.
14:48:42 <mlavalle> yes, we did that in Rocky
14:48:59 <mlavalle> and this bug could be the trigger for the siwtch
14:49:03 <mlavalle> switch
14:49:13 <annp_> +1
14:49:37 <slaweq> but even if we switch default, the non-default option is still there and is buggy, right?
14:49:45 <slaweq> do we have plans to drop support for it completly?
14:50:06 <mlavalle> ok, for the sake of time, let's continue the discussion in the bug
14:50:10 <amotoki> slaweq: good point. we need to check how other projets like nova handle it.
14:50:31 <hle2> it will be great to use the new web server, and we can verify it
14:51:02 <mlavalle> Our bug deputy this week is manjeets
14:51:14 <mlavalle> #topic neutron-lib
14:51:22 <mlavalle> boden: any updates?
14:51:30 <boden> yes quick b/c I have to run
14:51:47 <boden> we have a new release of neutron-lib I guess: https://review.openstack.org/#/c/615862/
14:52:31 <boden> also 2 lib related patches if folks get a min: https://review.openstack.org/#/c/615339/  and https://review.openstack.org/#/c/596959/
14:52:53 <liuyulong> boden, does the new release include https://review.openstack.org/#/c/613562/
14:53:08 <boden> liuyulong not sure off hand; you can check
14:53:18 <amotoki> yes, it is included
14:53:21 <boden> that's all I have... I have to run now
14:53:32 <mlavalle> boden: thanks
14:53:56 <mlavalle> I don't see hongbin today on-line
14:54:03 <mlavalle> so let's skip os-ken topic
14:54:20 <mlavalle> #topic On demand agenda
14:54:32 <amotoki> re CLI, I would like to raise bug 1801360 in neutronclient. https://review.openstack.org/#/c/615184/ is proposed.
14:54:32 <openstack> bug 1801360 in python-neutronclient "Recent openstacksdk release broke the client" [Undecided,In progress] https://launchpad.net/bugs/1801360 - Assigned to Lucian Petrut (petrutlucian94)
14:54:32 <mlavalle> amotoki: no need to discuss releases anymore
14:54:33 <amotoki> According to review comments it affects kuryr. I am checking lower-constaraints after the meeting.
14:54:59 <amotoki> mlavalle: i have nothing regarding stable releases. I will check ocata release on the relmgt team
14:55:28 <mlavalle> amotoki: yes, that's what I meant. I think we covered everything before
14:55:33 <slaweq> if I may, I would like to ask people to review https://review.openstack.org/#/c/597567/
14:55:51 <slaweq> it's very long standing patch which I hope is ready now )
14:55:53 <slaweq> :)
14:56:21 <mlavalle> in my pile for today
14:56:22 <slaweq> thx liuyulong, swami and haleyb for help with this :)
14:56:25 <slaweq> thx mlavalle
14:56:56 <mlavalle> anything else for today?
14:57:15 <mlavalle> ok team. Thanks for attending
14:57:19 <mlavalle> have a great week
14:57:22 <slaweq> thx
14:57:23 <mlavalle> #endmeeting