14:00:24 <jlibosva> #startmeeting networking
14:00:25 <openstack> Meeting started Tue Jun 20 14:00:24 2017 UTC and is due to finish in 60 minutes.  The chair is jlibosva. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:29 <openstack> The meeting name has been set to 'networking'
14:00:31 <jlibosva> hi everyone
14:00:34 <mlavalle> o/
14:00:37 <haleyb> hi
14:00:37 <hichihara> hi
14:00:38 <hoangcx_> o/
14:00:40 <amotoki> hi
14:00:52 <bcafarel> hello
14:00:56 <jlibosva> #topic Announcements
14:01:04 <jlibosva> I don't have any announcements to make :)
14:01:12 <jlibosva> does anybody have anything to announce?
14:01:25 <mlavalle> Just remind people that Pike-3 is around the corner
14:01:36 <jlibosva> it's 27 July, right?
14:01:57 <mlavalle> #link https://releases.openstack.org/pike/schedule.html
14:02:14 <mlavalle> yeap, Jul 24 - 28
14:02:37 <amotoki> note that library freeze is one week earlier
14:03:04 <mlavalle> good point
14:03:07 <jlibosva> ok, thanks guys for bringing this up
14:03:21 <amotoki> one more thing, (partially related) neutron-fwaas/vpnaas-dashboard repo have been created today.
14:03:46 <amotoki> i am preparing various things around this. fwaas-dashboard will be part of neutron.
14:03:47 <annp> hi, sorry for late
14:04:08 <amotoki> that's all
14:04:10 <hichihara> amotoki: Core member is you only?
14:04:14 <hoangcx_> amotoki: +1
14:04:44 <amotoki> hichihara: the initial core will be neutron-fwaas-core + horizon-core. it will change gradually
14:04:46 <hoangcx_> amotoki: will you announce a link to the thread you raised?
14:05:04 <hichihara> amotoki: I got it
14:05:14 <amotoki> hoangcx_: no announce thread. i will announce it once the setup is done
14:05:32 <hoangcx_> amotoki: Yep.
14:06:09 <amotoki> let's move on
14:06:12 <jlibosva> we can move one
14:06:30 <jlibosva> oh, one thing
14:06:38 <jlibosva> there won't be a neutron ci meeting today
14:06:49 <jlibosva> it was announced on ML
14:06:52 <jlibosva> #topic Blueprints
14:07:18 <jlibosva> as mentioned before
14:07:22 <jlibosva> we're in pike-3 now
14:07:23 <jlibosva> https://launchpad.net/neutron/+milestone/pike-3
14:07:43 <jlibosva> it seems some blueprints haven't been flipped yet
14:08:31 <amotoki> perhaps we need to re-target uncompleted blueprint from pike-2 to pike-3
14:08:35 <amotoki> https://launchpad.net/neutron/+milestone/pike-2
14:08:36 <jlibosva> yep
14:08:41 <jlibosva> there are a few
14:08:57 <jlibosva> is anybody blocked on his work and need to raise a flag here?
14:09:31 <tmorin> o/
14:09:35 <mlavalle> jlibosva: a couple of the ones that show up in the pike-3 panel are docs
14:10:43 <jlibosva> I think we need to go through bugs and blueprints and flip them to p3
14:11:30 <jlibosva> so nobody is blocked?
14:11:51 <mlavalle> nobody present in this meeting, it seems
14:12:09 <jlibosva> ok, let's move on then :)
14:12:11 <jlibosva> #topic Bugs and gate failures
14:12:12 <mlavalle> tmorin: did you mean your are blocked with a blueprint?
14:12:40 <tmorin> mlavalle: no, I was just arriving late and saying "hi"
14:12:45 <jlibosva> :)
14:12:56 <jlibosva> who was bug deputy for last week?
14:13:10 <hichihara> maybe kevin
14:13:12 <mlavalle> Kevinbenton took the role last week
14:13:33 <jlibosva> I guess kevinbenton is sleeping at this hour
14:13:49 <hichihara> I think so
14:14:16 <jlibosva> is anybody else aware of anything critical in the gate?
14:15:03 <amotoki> one thing related to gate failure. last week we broke heat dsvm py35 job (though it was already solved)
14:15:04 <mlavalle> well we have a couple of critical bugs with no owners
14:15:22 <jlibosva> https://bugs.launchpad.net/neutron/+bug/1697533 ?
14:15:23 <openstack> Launchpad bug 1697533 in neutron "test_install_flood_to_tun failed with: 'tun_id=0x378' not in u' unchanged'" [Critical,Confirmed]
14:15:26 <amotoki> I wonder we need to enable some py35 dsvm job from the experience
14:16:12 <amotoki> mlavalle: jlibosva: cover critical bugs first
14:16:22 <jlibosva> amotoki: we're working on functional job with python35 flavor, we had quite a progress there
14:16:29 <hichihara> I thought Ihar works it for py35
14:16:29 <jlibosva> so the bug I posted looks familiar
14:16:54 <jlibosva> we used to have a workaround for it, I remember that required some new openvswitch package
14:17:00 <amotoki> my point is that some other projects already enable py35 tempest job and we potentially breaks their gate.
14:18:04 <amotoki> so I wonder we need to enable minimum set which cover py35 dsvm in addition to the func job. thought?
14:18:27 <jlibosva> ok, let's talk about py35 first :)
14:18:44 <jlibosva> amotoki: we track the py35 goal in Neutron CI meetings
14:19:12 <jlibosva> amotoki: currently we aim to make functional jobs voting so we have some coverage of python3 functionality. Then we want to add tempest
14:19:24 <jlibosva> amotoki: do you have a link to that failure handy?
14:19:30 <jlibosva> I wonder whether functional job would cover that issue
14:19:42 <amotoki> bug 1698355
14:19:44 <openstack> bug 1698355 in oslo.serialization "py35 dsvm job failing with RemoteDisconnected error" [Critical,Confirmed] https://launchpad.net/bugs/1698355 - Assigned to Ihar Hrachyshka (ihar-hrachyshka)
14:20:44 <amotoki> it turned out an serialization issue and we hit MemoryError in py35. Good finding by Ihar
14:21:18 <amotoki> It happend inside db plugin module of SG.
14:21:36 <amotoki> so I am not sure we can cover it by func job
14:23:17 <jlibosva> that return value should be list sounds to me like UT should be able to catch that
14:24:32 <amotoki> yeah, it might be possible. it is a corner case around db_obj vs ovo with netaddr.IPNetwork handing in oslo.serialization.
14:24:33 <jlibosva> but yeah, definitely we should move towards having tempest python3 voting
14:25:15 <jlibosva> though we still have issues in functional tests that fail on real scenarios - like ip_lib commands
14:25:27 <jlibosva> which would make agents failing
14:26:03 <jlibosva> amotoki: so do you agree that functional tests should be first step?
14:26:12 <amotoki> jlibosva: agree.
14:26:18 <jlibosva> and also everyone else :)
14:26:25 <amotoki> otherwise, we will have more issues in tempest jobs
14:26:31 <jlibosva> yep
14:26:35 <jlibosva> ok, let's move on
14:26:48 <jlibosva> I'll assing https://bugs.launchpad.net/neutron/+bug/1697533 to myself as I made that workaround there
14:26:49 <openstack> Launchpad bug 1697533 in neutron "test_install_flood_to_tun failed with: 'tun_id=0x378' not in u' unchanged'" [Critical,Confirmed]
14:26:57 <jlibosva> it could be that ovs was updated or something
14:27:08 <jlibosva> but I don't see functional job busted, so it probably doesn't happen always
14:27:31 <jlibosva> another critical bug is https://bugs.launchpad.net/neutron/+bug/1696376
14:27:31 <openstack> Launchpad bug 1696376 in neutron "logging resource UT failures on networking-midonet gate jobs" [Critical,In progress] - Assigned to YAMAMOTO Takashi (yamamoto)
14:27:48 <jlibosva> which has a fix proposed https://review.openstack.org/#/c/471699/
14:28:23 <jlibosva> it seems it requires some eyes from fwaas cores
14:28:53 <mlavalle> jlibosva: I will take a look
14:29:02 <jlibosva> mlavalle: thanks!
14:29:25 <reedip_> mlavalle : it would be great if you can
14:29:42 <jlibosva> I don't see any other critical bugs
14:29:42 <reedip_> we are discussing the same on the fwaas meeting
14:30:06 <jlibosva> reedip_: is it running right now?
14:30:35 <amotoki> fwaas uses the same slot at #-meetng-4
14:30:37 <reedip_> jlibosva : yes
14:30:40 <reedip_> :)
14:30:53 <jlibosva> :-/
14:31:11 <jlibosva> so unless there is any other critical bug, I'd move to my favourite part of meeting
14:31:25 <jlibosva> and that's getting a volunteer for bug deputy role, yay!
14:31:30 <jlibosva> for this week
14:32:02 <jlibosva> if there is no volunteer then I'll take it
14:32:30 <jlibosva> ok, sold to me, I'll update wiki after the meeting
14:32:36 <amotoki> I will be a volunteer for the NEXT week
14:32:52 <jlibosva> amotoki: ok, thanks. I'll write you to the wiki page too
14:33:08 <jlibosva> let's move on then
14:33:12 <jlibosva> #topic Docs
14:33:15 <jlibosva> mlavalle: hi :)
14:33:17 <mlavalle> hi
14:33:47 <mlavalle> I want to highlight that there is a review spelling out the plan to move all the docs to the project trees
14:34:34 <mlavalle> If you don't have much time but still are interested on what is going on, there is a section here that summarizes what is going on:
14:34:48 <mlavalle> #link https://review.openstack.org/#/c/472275/8/specs/pike/os-manuals-migration.rst@230
14:34:48 <amotoki> the docs-spec is here https://review.openstack.org/#/c/472275/ if you are interrested in what is happening
14:35:29 <mlavalle> amotoki has already pushed patchsets covering some of thos action items, both for Neutron and the neutron client
14:35:53 <mlavalle> and I will start moving soon the networking guide to the Neutron tree
14:36:32 <mlavalle> If anyone is interested in helping with this process, pelase ping amotoki or me, so we can coordinate
14:36:46 <mlavalle> that's all from me
14:36:51 <jlibosva> mlavalle: thanks for updates
14:37:01 <mlavalle> amotoki: do you have anything else you would like to highlight?
14:37:10 <amotoki> nothing from me
14:37:17 <mlavalle> I do
14:37:33 <mlavalle> I want to highlight the great job amotoki is doing in this area :-)
14:37:44 <mlavalle> that's all
14:37:45 <jlibosva> and all other areas :)
14:37:49 <hichihara> ++
14:37:56 <reedip_> ++ amotoki :)
14:38:44 <jlibosva> we can move on then
14:38:46 <jlibosva> #topic Transition to OSC
14:39:13 <amotoki> on OSC side, the remaining few commands are in slow progress
14:39:18 <jlibosva> amotoki: hi :) another yours area you are great at
14:39:49 <amotoki> the remaining things are l3-agent and tags. I am the owner of these patches now.
14:40:10 <amotoki> good news is that SDK patch on network tag support has been merged and released.
14:40:23 <amotoki> so we can focus on OSC patches now.
14:40:23 <hichihara> cool
14:40:52 <amotoki> that's all from me
14:41:34 <jlibosva> amotoki: you mentioned earlier today that libraries have freeze week before p3, is neutronclient considered a library? or it doesn't have a freeze?
14:41:57 <amotoki> jlibosva: this is non-client library freeze
14:42:16 <amotoki> neutronclent has the same feature freeze as neutron
14:42:23 <jlibosva> amotoki: ok, thanks for clarification
14:42:30 <jlibosva> and updates :)
14:42:45 <jlibosva> I think we can move on
14:42:47 <jlibosva> #topic Neutron-lib and planned neutron refactoring
14:43:00 <jlibosva> boden couldn't make it to the meeting but he updated me beforehand
14:43:30 <jlibosva> he wanted to highlight this spec https://review.openstack.org/#/c/473531/
14:44:13 <jlibosva> and also this new extension attribute - https://review.openstack.org/#/c/475577
14:44:15 <mlavalle> he needs reviewing eyes, I guess
14:45:24 <jlibosva> yep :)
14:45:30 <jlibosva> and raise awareness
14:45:45 <jlibosva> and that's all from boden via me
14:45:47 <mlavalle> added to my pile
14:46:12 <jlibosva> does anybody have anything related to neutron-lib
14:46:39 <jlibosva> I take it as no
14:46:45 <jlibosva> #topic open discussion
14:47:12 <jlibosva> has anybody a topic to discuss here?
14:47:20 <ralonsoh> May I?
14:47:27 <jlibosva> ralonsoh: sure, go ahead! :)
14:47:32 <ralonsoh> https://review.openstack.org/#/c/474248/
14:47:59 <ralonsoh> There is an small bug in os-vif bug/1632372
14:48:08 <ralonsoh> related to Neutron, Nova and Neutron-lib
14:48:30 <ralonsoh> Because the possible change in the API, I need your reviews
14:49:04 <ralonsoh> All the information is in PS1 comment
14:49:31 <tmorin> sorry, doing two things at a time
14:49:34 <ralonsoh> And that's all, thank you in advance
14:49:39 <tmorin> there one thing I have about neutron-lib
14:50:17 <tmorin> https://review.openstack.org/#/c/467277/ has now had a few reviews, and would benefit from reviews from cores
14:51:13 <ralonsoh> hichihara: I will appreciate some reviews on this set of bugs
14:51:39 <jlibosva> ralonsoh: hichihara so what would be the next steps? maybe raising this in a ML?
14:52:26 <jlibosva> I think that would be the best way to get feedback from both nova and neutron folks
14:52:39 <ralonsoh> jlibosva: thanks!
14:53:14 <hichihara> ralonsoh: You're welcome. I agree witch jlibosva because the issue is cross project
14:53:33 <amotoki> regarding os-vif one, I am not sure it is an API change or not. we need to check where this new variable is consumed carefully. let's cross check
14:53:35 <ralonsoh> hichihara: I'll do this. Thanks
14:54:39 <jlibosva> tmorin: thanks for raising the review here
14:55:00 <jlibosva> hichihara: maybe you want to have a look at tmorin 's review :) https://review.openstack.org/#/c/467277/
14:55:23 <jlibosva> we have 5 mins left, any other thing to raise here?
14:55:27 <ebbex> https://bugs.launchpad.net/neutron/+bug/1687896 , suppose I'll just file a new one for the errors we're seeing (that we didn't with another patch) now a "fix" has been released for neutron-rpc-server?
14:55:29 <openstack> Launchpad bug 1687896 in neutron "neutron-rpc-server fails to start on configuration that works under neutron-server" [Undecided,Fix released] - Assigned to Sean McCully (sean-mccully)
14:56:32 <hichihara> tmorin jlibosva: I'll review again. But we'd like to see bgpvpn members +1 before merging.
14:56:41 <ebbex> Since technically neutron-rpc-server is actually running, but not very useful :)
14:56:56 <jlibosva> hichihara: thanks
14:57:05 <tmorin> hichihara: doude provided a +1, and pcarver as well on a later PS
14:57:49 <amotoki> ebbex: thanks for raising this. neutron-rpc-server needs more users to make it mature.
14:57:53 <tmorin> hichihara: I'm not sure matrohon will have time to review, and not sure who else is active enough
14:58:02 <jlibosva> ebbex: maybe it's worth ivestigating why we don't see the failure on the gate
14:58:20 <jlibosva> and possibly create a fullstack test for it
14:58:55 <amotoki> jlibosva: I think we have not enabled it in the gate
14:58:57 <hichihara> tmorin: I'm OK.
14:58:57 <ebbex> jlibosva: Yeah, cause I don't think anyone has been running it as two separace uwsgi+rpc before.
14:58:59 <jlibosva> ebbex: but yeah, if there are other new errors, please file a bug
14:59:41 <ebbex> jlibosva: thanks :)
14:59:42 <tmorin> hichihara: I'll see if matrohon can have a look, but please don't wait too long for him
14:59:47 <jlibosva> amotoki: I guess little fullstack test would be a good candidate :) but that might be just me :)
14:59:54 <jlibosva> ok, we're at the top of the hour
14:59:58 <hichihara> tmorin: OK. Thanks.
15:00:02 <jlibosva> thanks all for coming and all the discussions :)
15:00:10 <mlavalle> o/
15:00:10 <jlibosva> #endmeeting