14:00:18 #startmeeting networking 14:00:19 Meeting started Tue Dec 17 14:00:18 2019 UTC and is due to finish in 60 minutes. The chair is slaweq. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:22 The meeting name has been set to 'networking' 14:00:22 welcome :) 14:00:27 o/ 14:00:31 o/ 14:00:38 \o 14:00:42 o/ 14:00:58 hi 14:01:15 ok, lets start the meeting 14:01:24 #topic Announcements 14:01:48 Reminder that we are still looking for "new daddies" for neutron-vpnaas and neutron-fwaas 14:01:58 I will send also reminder message to the ML this week 14:02:19 next one 14:02:46 Ussuri-1 was last week: https://releases.openstack.org/ussuri/schedule.html 14:03:07 now we are going to Ussuri-2 milestone which will be in February 14:03:34 and last announcement for today from me: 14:04:11 It's our last meeting in 2019 - for next weekes I will be on PTO and I think that this will be similar to many others also 14:05:03 so we will probably resume our meetings on 6th or 13th January 14:05:22 yes attendance next 2 weeks would probably be low 14:05:26 I will send email about it also 14:05:26 +1 14:06:04 and that's all from me for announcements 14:06:06 o/ 14:06:16 anything else You want to share with the team? 14:07:26 ok, so lets move on 14:07:28 #topic Blueprints 14:07:41 list updated for Ussuri-2 is here https://launchpad.net/neutron/+milestone/ussuri-2 14:08:01 I think we should now focus on implementation of them :) 14:08:39 for 3 of them we recently released neutron-lib with needed api definitions so I hope that neutron's side will be also done soon 14:09:12 I would like ask all of You to review once again spec for https://blueprints.launchpad.net/neutron/+spec/metadata-over-ipv6 14:09:32 it's on https://review.opendev.org/#/c/315604/ 14:09:46 I know haleyb reviewed it - thx haleyb :) 14:10:00 :) 14:10:27 haleyb: but I'm not sure if we really need to somehow "reserve" this link-local IPv6 address 14:10:46 as it's link-local class, isn't it safe to use it? 14:11:00 169.254/16 is also link-local class in IPv4 world 14:11:46 I think it's as safe as 169.254/16 in ipv4 14:11:48 perhaps we will have similar questions when we try to support it in cloud-init. 14:11:53 do we need more than that? 14:12:16 slaweq: that's just the IETF part of my brain saying that, technically we shouldn't, but we are giving it a special meaning 14:12:32 rubasov: that's exactly my understanding 14:12:47 haleyb: do You have any contacts there, which You can ask? 14:13:04 or should I try to find someone? 14:14:00 not any more, there is a ML for IPv6, i'll find it offline 14:14:36 haleyb: ok, thx 14:14:50 and I'll cleanup our spec 14:15:03 any other updates about blueprints? 14:15:15 thx rubasov :) 14:15:40 slaweq: unless slaweq already did it :-) 14:15:46 slaweq: sorry I'm a bit behind 14:16:03 rubasov: no problem, I just addressed some haleyb's comments there 14:16:12 but You can take care of it now if You want 14:16:20 I just want to move forward with this :) 14:16:28 slaweq: you're right 14:18:09 if there is no any other updates about BPs, lets move on 14:18:17 #topic Community goals 14:18:58 according to Support IPv6-Only Deployments I recently revived slowly work gmann's on neutron-tempest-plugin job for stadium projects 14:19:09 and I found out that bagpipe isn't ready for IPv6 yet 14:19:26 so we have to do this job without bagpipe for now 14:19:50 and we have some connectivity issues in this job which I will need to figure out how to fix now 14:20:04 njohnston: any updates about dropping py27 support? 14:20:31 etherpad: https://etherpad.openstack.org/p/neutron-train-zuulv3-py27drop 14:20:59 not really, I think things are slow as noone wants to risk introducing a breaking change right before so many people go on PTO 14:21:04 I think things will resume in January 14:22:08 there are few stadium changes https://review.opendev.org/#/q/status:open+topic:drop-py27-support (thanks amotoki for the reviews) 14:22:30 important to note that we are now in stage 2 14:22:42 so all projects are enabled to drop their py27 testing 14:23:03 finally :) 14:23:25 neutron already dropped py27 support, so py27 testing in stadium projects has less value 14:23:30 merging code that is incompatible with py27 is still to be done carefully and with an eye to cross-project ripple effects, and IMHO not to be done gratuitously 14:23:52 at least until stage 3 14:23:57 njohnston: yes, I agree with that 14:24:15 especially for things like neutron-lib we shouldn't drop all compatibility code yet 14:24:21 +100 14:24:52 thx for update njohnston 14:25:05 I hope it will start moving faster in January :) 14:25:11 I am sure of it 14:25:15 :) 14:26:22 ok, lets move on 14:26:26 #topic Bugs 14:26:35 yamamoto was on bug deputy last week 14:26:47 his report is here: http://lists.openstack.org/pipermail/openstack-discuss/2019-December/011587.html 14:27:04 fortunatelly almost all of the bugs are at least in progress 14:27:20 we only need volunteer for https://bugs.launchpad.net/neutron/+bug/1855912 14:27:20 Launchpad bug 1855912 in neutron "MariaDB 10.1 fails during alembic migration" [High,Confirmed] 14:27:38 we should switch our CI job to use MariaDB 10.3 at least 14:27:42 yes there is patch 14:27:43 and update documentation 14:28:00 #link https://review.opendev.org/#/c/698980/ 14:28:02 ralonsoh: really? I didn't saw any :) 14:28:10 if ubuntu bionic and mariadb 14:28:12 ralonsoh: thx 14:28:23 then I try to use the newer repos in Ubuntu 14:28:34 10.4 is working, manually tested 14:29:09 ralonsoh: thx 14:29:13 yw 14:29:16 so we will only need docs update 14:29:20 yes 14:29:29 can You update this LP with link to this patch also? 14:29:33 sure 14:29:40 thx 14:30:01 that's all about bugs from last week from me 14:30:08 any other bugs You want to talk about? 14:31:02 not a bug, but the version bump for hacking caused some issues that haleyb has a fix out for: https://review.opendev.org/699021 14:31:37 yeah, we really need this patch :) 14:31:38 hopefully that merges soon, recheck, recheck 14:31:42 yeah :/ 14:31:49 it had been some times since I saw so many +2 14:31:54 (and rechecks) 14:31:58 part of that is pinning the hacking version to an older range; I figure that in January we can revisit what it will take to get the newer version working 14:32:08 how about stable branches? 14:33:03 change is not too large so I was thinking of backporting it by itself when it finally merges 14:33:10 (we had similar strategy on a previous breakage) 14:33:29 we should pin if we see it, just don't know if they're affected do we? 14:33:31 yeah. we can also avoid pep8 failures by capping hacking<2. we don't usually cap python modules in stable branches but hacking is in blacklist in g-r. 14:33:58 I think we will need it in Train at least 14:33:59 haleyb: train and stein are for sure 14:34:02 and maybe Stein 14:34:10 bcafarel++ thx for confirmation :) 14:34:44 hacking is not in u-c, so bakcport or capping is required for old branches (stein/train at least, I am not sure for rocky) 14:35:24 bcafarel: we might just want to pin hacking, the other part shouldn't matter, but it's pretty minor stuff 14:35:49 looks like it's going to fail again in master :( 14:36:15 yes :( 14:36:20 ugh 14:36:27 :/ 14:37:25 ok, lets continue 14:37:32 our bug deputy this week is rubasov 14:37:43 ack 14:37:48 rubasov: thx 14:37:55 and next week we are starting new round 14:38:13 we will start with mlavalle who switched with bcafarel 14:38:47 * bcafarel will remember to pay him a $beverage next time for that 14:38:50 :) 14:38:55 ok, next topic 14:38:59 #topic Networking OVN and ML2+OVS+DVR Convergence 14:39:18 ralonsoh: haleyb: any updates? 14:39:26 what is the progress of it? 14:39:26 a but stuck, mainly because of the CI 14:39:35 last patch was merged on friday 14:39:44 and still waiting for the CI to be fixed 14:39:58 maybe in a couple of days we'll be able to make a better update 14:40:03 FTs are working 14:40:32 yes, some patches waiting for CI, the octavia driver is pending 14:40:33 and we are still testing the deployment of OVN with Neutron only, without the subproject networking-ovn 14:41:01 ok, are those all missing patches https://review.opendev.org/#/q/status:open+project:openstack/neutron+branch:master+topic:bp/neutron-ovn-merge or we will need anything else too? 14:41:05 https://review.opendev.org/#/c/697095/ - i asked about just dropping the stable policy flag if that's an issue in merging it 14:42:17 slaweq, hmm I don't see the FTs there 14:42:20 I'll check that 14:42:33 sorry yes https://review.opendev.org/#/c/697440/ 14:42:42 haleyb: About that, are you planning on pulling the rocky, stein, train, versions of the plugin into the repo? 14:43:20 haleyb: Or are you going to say "old releases use networking-ovn" so there is no need to hold the old stable branches? 14:43:22 njohnston: I think that for stable branches this driver will still be in networking-ovn repo, in same way like it's with mech driver 14:43:31 njohnston: i hadn't thought about doing that, as the other OVN stable branches will live on 14:44:01 haleyb: If you aren't pulling in the old stable branches into the new repo then I think that nullifies mugsie 14:44:04 's objection 14:44:40 njohnston: it does 14:44:52 \o/ 14:45:05 mugsie: thanks! 14:45:13 mugsie: thx 14:45:17 haleyb: so I would suggest clarifying that and then things can proceed 14:45:25 njohnston: ack, thanks! 14:45:54 ok, lets move on 14:45:58 next topic 14:46:00 #topic neutron-lib 14:46:11 I just wanted to mention that we released 1.30 last week 14:46:37 is it too soon to ask for 1.31.1?? 14:47:15 ralonsoh: You need it to have some new api-def released, right? 14:47:20 yes 14:47:33 but low priority 14:48:23 for me it would be ok, but I would like also https://review.opendev.org/#/c/695205/ to be merged 14:48:40 yes 14:48:49 this patch should be a must for next version 14:49:09 (and the dependant one https://review.opendev.org/#/c/698261/) 14:49:28 this one is for neutron 14:49:30 :) 14:49:33 I know 14:49:42 but it's important: once we have a new n-lib version 14:49:48 yes, I know 14:49:49 we need to merge the neutron patch 14:49:50 :) 14:50:11 ok, I will check this patch and when it will be merged, I will propose new version of neutron-lib again 14:50:19 +1 14:50:53 amotoki: will You be ok with this plan? 14:51:00 sounds fine 14:51:04 thx 14:51:25 ok, so we should be good with this one I hope 14:51:31 lets move to the last topic for today 14:51:33 #topic On Demand Agenda 14:51:38 cmurphy: Your turn :) 14:52:32 hi o/ 14:52:49 i wanted to stop by to check in before the holidays now that the policy popup team is official 14:53:03 #link https://governance.openstack.org/tc/reference/popup-teams.html#secure-default-policies 14:53:25 i mostly wanted to discuss how this team would like to organize and track the work for this effort 14:53:35 #link https://wiki.openstack.org/wiki/Consistent_and_Secure_Default_Policies_Popup_Team 14:54:07 we have a "progress" section in that wiki page that i'm hoping will link to each project's tracking 14:54:17 could be lp bugs, storyboard stories, an etherpad... 14:55:22 cmurphy: I don't think mlavalle started any work on this already 14:55:30 cmurphy: but I will talk with him about this 14:55:53 and will ask him to update this progress section as soon as he will have anything created 14:56:02 sure, it's still very early :) 14:56:25 thanks :) and also if you have design documents like a spec or blueprint or a wiki page like the barbican team has, please update that section as well 14:56:37 I can also talk with him if he want to have item on agenda of our meetings to track progress here too weekly 14:56:48 cmurphy: sure 14:56:52 ++ 14:57:19 let me know if you have any questions, i'll be lurking in your channel or you can ping me in #openstack-dev 14:57:34 ok, thx a lot 14:57:39 np 14:58:17 that's all i had on that 14:58:43 thx for comming to our meeting and for update 14:58:50 ok, I think we are done now 14:59:17 so as last thing I want to wish all of You great holidays and see You all in 2020 :) 14:59:19 o/ 14:59:20 enjoy the time off for all that take some 14:59:24 #endmeeting