21:00:20 <mlavalle> #startmeeting networking 21:00:22 <openstack> Meeting started Mon Dec 10 21:00:20 2018 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:25 <openstack> The meeting name has been set to 'networking' 21:00:27 <njohnston> o/ 21:00:42 <slaweq> o/ 21:00:48 <rubasov> o/ 21:01:20 <haleyb> hi 21:01:23 <mlavalle> Today's agenda is here: https://wiki.openstack.org/wiki/Network/Meetings 21:01:24 <amotoki> hi 21:01:41 <bcafarel> o/ 21:01:45 <mlavalle> #topic Announcements 21:02:44 <mlavalle> First of all, the next milestone Stein-2 is fast approaching. With the holidays season between now and then, the milestone is really around the corner 21:03:34 <mlavalle> Today, after the customary one week wait period, I added njohnston and hongbin to the Neutron core team 21:03:49 <mlavalle> #link http://lists.openstack.org/pipermail/openstack-discuss/2018-December/000780.html 21:03:53 <njohnston> \o/ 21:03:54 <hongbin> o/ 21:03:58 <bcafarel> congrats, both of you :) 21:04:08 <mlavalle> #link http://lists.openstack.org/pipermail/openstack-discuss/2018-December/000783.html 21:04:28 <slaweq> congrats :) 21:04:38 <mlavalle> Congratulations to both of you! Thanks for all your hard work and use your superpowers wisely 21:04:52 <haleyb> yes, congrats! 21:05:09 <hongbin> thank you. it is a pleasure to work in neutron team 21:05:12 <njohnston> Special thanks to mlavalle who has been teaching me the trade 21:05:34 <njohnston> and I echo hongbin - this is a wonderful team, and I am honored to be a part 21:05:40 <hongbin> i would like to thank slaweq for his mentorship to me 21:05:56 <mlavalle> Any other announcements? 21:06:35 <mlavalle> ok, let's move on 21:06:54 <mlavalle> #topic Blueprints 21:07:34 <mlavalle> These are the blueprints we are working on during this cycle: https://launchpad.net/neutron/+milestone/stein-2 21:08:09 <mlavalle> njohnston: please note that I updated the assignee to you of https://blueprints.launchpad.net/neutron/+spec/enginefacade-switch 21:08:09 <njohnston> I am going to try to push hard to get the remaining patch for bulk port upgrades ready this week, as I am on PTO from 12/15 to EOY - so as you mentioned before, not much time before S-2. 21:08:33 <njohnston> mlavalle: Got it! 21:08:42 <mlavalle> njohnston: ok, will keep an eye on those patches 21:09:10 <njohnston> thanks mlavalle 21:09:23 <mlavalle> I also updated the assignee to tidwellr of https://blueprints.launchpad.net/neutron/+spec/subnet-onboard 21:10:22 <mlavalle> Please note that I added to the Stein-2 dashboard https://blueprints.launchpad.net/neutron/+spec/network-segment-range-management 21:11:01 <mlavalle> We recently approved the spec: https://review.openstack.org/#/c/599980/ 21:11:47 <mlavalle> and Kailun Qin has indicated that he will push code in bite size patches this week 21:11:54 <mlavalle> so please keep an eye on those 21:12:50 <mlavalle> Finally, this sr-iov mirroring spec seems to be close to merge: https://review.openstack.org/#/c/574477/ 21:13:19 <mlavalle> yamamoto left a question in the patch, that I would like answered before merging 21:13:43 <mlavalle> earlier today I sent an email to munish. Hopefully he will respond soon 21:14:39 <mlavalle> Yes, I do have a question 21:14:54 <mlavalle> This one is done as far as I know: https://blueprints.launchpad.net/neutron/+spec/router-gateway-ip-qos 21:15:21 <mlavalle> we even merged the update to the Networking Guide, right slaweq? 21:15:31 <slaweq> mlavalle: I think so 21:15:41 <mlavalle> ok, I'll mark it complete 21:15:42 <slaweq> I will check it and update BP tomorow morning 21:16:37 <mlavalle> ok.... 21:16:41 <mlavalle> let's move on 21:16:56 <mlavalle> #topic Community Goals 21:17:32 <amotoki> I have an update on policy-in-code 21:17:39 <mlavalle> Great! 21:17:44 <amotoki> First, FYI, I changed the blueprint URL to https://blueprints.launchpad.net/neutron/+spec/neutron-policy-in-code (from get-policy-from-neutron-lib) because the old slug is no longer correct. 21:17:56 <mlavalle> ack 21:18:07 <amotoki> I succeeded to identify the root cause of unit test and func test failures 21:18:13 <amotoki> It is because policy-in-code default rules inside the code are loaded before API extensions. 21:18:21 <amotoki> As a result, convertor functions used in FieldCheck are not initialized properly. 21:18:43 <amotoki> I am testing the latest one in https://review.openstack.org/#/c/619898/5 and waiting for the whole test results 21:19:02 <amotoki> I will merge thsi into the main patch and address slaweq's comments then. 21:19:23 <mlavalle> ok 21:19:24 <amotoki> I hope all things are addressed in neutron side by this. 21:19:42 <slaweq> thx amotoki 21:19:47 <mlavalle> do you have the pointer to the main patch handy? 21:20:07 <amotoki> the main patch is https://review.openstack.org/#/c/585037/ 21:20:13 <mlavalle> Thanks! 21:20:41 <mlavalle> any other updates on community goals? 21:22:10 <mlavalle> ok, let's move on then 21:22:16 <mlavalle> #topic Bugs 21:22:30 <mlavalle> Last week, rubasov was our deputy 21:22:52 <mlavalle> Here's the report: http://lists.openstack.org/pipermail/openstack-discuss/2018-December/000770.html 21:23:13 <rubasov> you beat me to pasting it :-) 21:23:32 <bcafarel> he has some training at it ;) 21:24:13 <mlavalle> I see that the critical bug already has a fix proposed 21:24:33 <mlavalle> #link https://review.openstack.org/#/c/623275/ 21:25:00 <mlavalle> Reviwers, please take a look ^^^^ 21:26:19 <mlavalle> I see that the for the two undecided bugs, haleyb has volunteered: https://bugs.launchpad.net/neutron/+bug/1807153 and https://bugs.launchpad.net/neutron/+bug/1807157 21:26:20 <openstack> Launchpad bug 1807153 in neutron "Race condition in metering agent when creating iptable managers for router namespaces" [Undecided,New] 21:26:21 <openstack> Launchpad bug 1807157 in neutron "Metering doesn't work for DVR routers on compute nodes" [Undecided,New] 21:26:27 <mlavalle> Thanks! 21:26:44 <haleyb> yes, i need to take a look at those 21:27:33 <mlavalle> and we are looking for a volunteer to take over this patch: https://review.openstack.org/#/c/581360/ 21:27:34 <rubasov> haleyb: thank you, I did not feel qualified for those 21:27:54 <mlavalle> which addresses https://bugs.launchpad.net/neutron/+bug/1807396 21:27:55 <openstack> Launchpad bug 1807396 in neutron "With many VMs on the same tenant, the L3 ip neigh add is too slow" [Medium,In progress] - Assigned to Thomas Goirand (thomas-goirand) 21:28:37 <haleyb> this has been an issue in the past 21:28:48 <rubasov> the patch is about optimizing it by batching it up 21:29:35 <mlavalle> any takers? 21:30:07 <haleyb> mlavalle: i had found a previous patch from years ago, i think ralansoh was looking into it 21:30:12 <haleyb> just can't find the tab now 21:30:30 <mlavalle> haleyb: I can ping him tomorrow morning 21:30:39 <haleyb> https://bugs.launchpad.net/neutron/+bug/1511134 21:30:40 <openstack> Launchpad bug 1511134 in neutron "Batch DVR ARP updates" [Undecided,New] 21:31:00 <haleyb> that was the old original bug, we punted thinking pyroute2 would solve the problem 21:31:37 <haleyb> https://review.openstack.org/#/c/431197/ was the change 21:31:49 <mlavalle> haleyb: so the bugs are duplicated? 21:32:49 <haleyb> yes, seem to be, even updated the same code 21:33:18 <mlavalle> ok, I'll update one of them 21:33:20 <haleyb> but the proposed change isn't great since we've moved away from running /sbin/ip 21:33:55 <slaweq> IMO we should go with this proposed solution and we should move to pyroute2 here also 21:33:59 <slaweq> haleyb++ 21:34:17 <mlavalle> I'll check with ralonsoh and see if he is taking it 21:34:48 <haleyb> slaweq: yes, we might be able to localize the batching to a few places, instead of in the ip_lib code 21:35:28 <mlavalle> any other bugs we should discuss today? 21:36:13 <mlavalle> ok, let's move on 21:36:16 <amotoki> On bug deputy, I am assigned to the week of Dec 31 but this week is a holiday week in Japan. Can anyone swap weeks? 21:36:43 <slaweq> amotoki: I can take it 21:37:04 <amotoki> slaweq: thanks. 21:37:19 <mlavalle> amotoki: so you'' be on duty the week of February 4th 21:37:19 <slaweq> amotoki: sure, np 21:37:36 <mlavalle> slaweq: did you ping pasuder? 21:37:48 <amotoki> yeah, I will care the week of Feb 4 21:37:49 <slaweq> mlavalle: yes, and he confirmed me that he will do it 21:38:03 <mlavalle> slaweq: great! 21:38:38 <mlavalle> hongbin: just to make sure, uou are scheduled for the week of December 24th. Is that ok with you? 21:38:58 <hongbin> mlavalle: i am fine with that 21:39:03 <mlavalle> cool! 21:39:12 <slaweq> hongbin: that should be "easy" week :) 21:39:12 <bcafarel> should be a quiet week 21:39:13 <mlavalle> #topic neutron-lib 21:40:01 <mlavalle> boden: any updates this week? 21:40:11 <boden> hi, just a few quick things 21:40:52 <boden> a reminder that we released neutron-lib last week... and the neutron requirements were recently bumped to use it.. other projects should consider using the latest as they get time/need 21:41:52 <boden> also a follow-up from last week's discussion on work items for neutron-lib.. I sent a note to the ML, but to recap some items are now listed on https://etherpad.openstack.org/p/neutron-lib-volunteers-and-punch-list 21:42:44 <mlavalle> boden: ahhh nice. Last week I had a conversation with Sandhya form Cisco 21:43:06 <mlavalle> she is interested in helping with items in this list 21:43:21 <mlavalle> Sandhya, you around? 21:43:28 <sadasu> yes, I am here 21:43:35 <mlavalle> perfect 21:43:55 <mlavalle> so take a look at the items in the list 21:44:00 <boden> mlavalle ok great... I'd suggest one at a time (some require some research/design), but if someone feels optimistic grab a more than one :) 21:44:22 <mlavalle> sadasu: please take a look at the list 21:44:33 <sadasu> I did take a look at the list earlier. can I take something that is unassigned? 21:44:37 <mlavalle> even if you are not completely sure, pick one 21:44:40 <boden> sadasu and feel free to ping/email me with questions 21:44:53 <sadasu> boden: thank you! will do 21:45:21 <boden> if we are done with that topic, just 1 last thing 21:45:33 <mlavalle> go ahead 21:46:18 <boden> I asked for some input/buy-in on https://review.openstack.org/#/c/621000/ njohnston found time, but was hoping at least 1 more core could have a looksee before I spend the time completing the patch 21:46:45 <mlavalle> boden: my bad, I should have taken the time. I'll will tomorrow 21:47:01 <boden> mlavalle great thanks.. 21:47:10 <boden> and that's all I have for neutron-lib 21:47:21 * mlavalle seems to have inveneted a new form of the future tense 21:48:15 <njohnston> innovative! ;-) 21:48:25 <mlavalle> ok, let's move on 21:48:33 <mlavalle> #topic os-ken 21:48:42 <hongbin> hi 21:48:53 <hongbin> a few things about os-ken 21:49:05 <hongbin> 1. os-ken 0.3.0 is release: https://review.openstack.org/#/c/623025/ 21:49:30 <hongbin> 2. i am working on the migration from ryu to os-ken in neutron : https://review.openstack.org/#/c/607008/ 21:49:52 <hongbin> i am waiting for the CI result and see what is the next step 21:50:06 <mlavalle> doesn't seem too bad 21:50:15 <hongbin> 3. there is a similar effort in other project as well: https://review.openstack.org/#/c/608357/ 21:50:15 <mlavalle> as far as the amount of change 21:50:42 <hongbin> yes, see if it will pass the CI or not 21:50:50 <hongbin> mlavalle: that is all from my side 21:51:24 <tidwellr> hongbin: https://review.openstack.org/#/c/608357/ needs to be rebased due to come recent changes, but it's ready to go as soon as we want to pull the trigger 21:51:44 <mlavalle> hongbin: mhhhh, should we send a message to the ML with a heads up indicating that we are changing in Neutron and dynamic-routing, in case other projects need to do the same? 21:52:06 <hongbin> tidwellr: ack 21:52:24 <hongbin> mlavalle: sure 21:52:40 <mlavalle> hongbin: would you take care of it? 21:52:41 <hongbin> mlavalle: i will take care of the ML 21:52:50 <mlavalle> hongbin: thanks! 21:53:05 <mlavalle> let's move on then 21:53:15 <mlavalle> #topic on demand agenda 21:53:26 <mlavalle> amotoki: anyhting on CLI? 21:53:34 <hongbin> tidwellr: one thing though, you might want to bump to os-ken 0.3.0 on your patch 21:53:43 <amotoki> mlavalle: nothing from me on CLI 21:53:51 <tidwellr> hongbin: will do 21:54:05 <mlavalle> any other topics we should discuss today? 21:54:13 <tidwellr> mlavalle: sorry I was late, but I had a question about subnet onboard 21:54:24 <mlavalle> shoot, tidwellr 21:54:38 <tidwellr> I can take it to the mailing list or discuss in the channel 21:54:59 <mlavalle> if you want, send a message to the ML 21:55:38 <tidwellr> I'm scratching my head at why the API was defined a certain way, it's poorly designed if you ask me and I'm wondering what flexibility we have to change the defininition 21:56:03 <mlavalle> ahhh, those are interesting discussions.... 21:56:11 <mlavalle> let's take it to the ML 21:56:14 <tidwellr> yes, they consume more time than we have here :) 21:56:20 <bcafarel> :) 21:56:43 <bcafarel> but this one is worth it (at least for feature usability) 21:57:24 <mlavalle> ok team, thanks for attending 21:57:30 <mlavalle> #endmeeting