21:01:54 <kevinbenton> #startmeeting networking 21:01:55 <openstack> Meeting started Mon Mar 6 21:01:54 2017 UTC and is due to finish in 60 minutes. The chair is kevinbenton. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:01:56 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:01:59 <openstack> The meeting name has been set to 'networking' 21:02:37 <dasanind> hi 21:02:54 <johnsom> o/ 21:03:23 <kevinbenton> #topic announcements 21:03:46 <kevinbenton> It doesn't look like anything new was added for announcements this week to the wiki 21:04:05 <dasm> o/ 21:04:15 <kevinbenton> The API job is currently broken and we are waiting for a fix to merge into devstack-gate 21:04:32 <kevinbenton> https://review.openstack.org/#/c/442123/ 21:04:35 <ihrachys> yeah, please refrain from W+1 if possible 21:05:17 <kevinbenton> ihrachys: is the concern that the devstack-gate will be ejected because of a failure? 21:05:38 <ihrachys> yeah, we don't want neutron patches to end up on top of it 21:05:49 <ihrachys> infra were kind enough to promote the fix 21:05:54 <ihrachys> so now it's in top of the gate 21:06:05 <ihrachys> hopefully it won't take too long 21:06:37 <kevinbenton> ack 21:06:54 <kevinbenton> does anyone else have any announcements to make? 21:07:21 <kevinbenton> armax: when are we merging your big context switch patch? 21:08:14 <ihrachys> kevinbenton: it's in gate 21:08:36 <kevinbenton> ihrachys: ah, do you have the link to it handy? 21:08:39 <boden> https://review.openstack.org/#/c/388157/ 21:08:40 <boden> ? 21:08:44 <ihrachys> sure, https://review.openstack.org/#/c/388157/ 21:08:59 <armax> kevinbenton: I don’t have an exact ETA 21:09:15 <armax> it’s been approved a few times, but it keeps getting bumped 21:09:32 <kevinbenton> #info neutron context module being removed in https://review.openstack.org/#/c/388157/ so imports will break 21:09:42 <kevinbenton> hopefully it will make it today 21:09:59 <ihrachys> I assume stadium fixes are all lined to merge right after 21:10:10 <armax> kevinbenton: as soon as it lands, I’ll check for breakages across the board and fix what I can 21:10:16 <ihrachys> I tried to find a definite list of those, but failed. so I pushed what I spotted. 21:10:28 <armax> ihrachys: we might have missed one or two, but the fixes should be easy to put together 21:10:53 <ihrachys> aye 21:11:03 <kevinbenton> ok 21:11:40 <kevinbenton> #topic bugs 21:12:42 <kevinbenton> nothing really critical bug-wise last week 21:13:05 <kevinbenton> we have one that sounds problematic for HA+ipv6 here: https://bugs.launchpad.net/neutron/+bug/1669765 21:13:05 <openstack> Launchpad bug 1669765 in neutron "RA is not disabled on backup HA routers" [High,Triaged] 21:13:38 <ihrachys> does dalvarez work on it? or just reported? 21:13:54 <kevinbenton> looks like just reported, which is why i brought it up 21:14:20 <ihrachys> I remember there was discussion about how to fix it, and I felt like Daniel was going to take it 21:14:31 <ihrachys> that's why I am asking, so that others don't repeat the same work 21:14:41 <haleyb> that's what i thought, i hadn't looked for that reason 21:14:46 <kevinbenton> ok, i'll assign to him 21:15:56 <kevinbenton> most of the rest of the stuff was RFEs or lower priority that we don't need to bring up here 21:17:05 <kevinbenton> #topic infra liaison 21:17:25 <kevinbenton> Nate had to step down being our liaison 21:17:27 <kevinbenton> http://lists.openstack.org/pipermail/openstack-dev/2017-March/113416.html 21:18:05 <ihrachys> that sucks 21:18:07 <kevinbenton> so we will need a volunteer to take his place 21:18:08 <njohnston> with great sadness 21:18:20 <dasm> What does infra liaison do? 21:18:49 <dasm> the same what oslo liaison? 21:19:11 <ihrachys> I think it's mostly approving any changes relevant to neutron gate in project-config and similar repos (also irc bots, pypi, etc.) 21:19:16 <armax> dasm: review infra patches, gate triage, be contact for infra related questions 21:19:26 <dasm> ihrachys: armax ack 21:19:34 <armax> it’s the bridge between the project and the infra team 21:19:39 <ihrachys> https://wiki.openstack.org/wiki/CrossProjectLiaisons#Infra 21:19:52 <armax> dealing with our CI setup etc 21:19:55 <dasm> if no one will step up, i could take it over 21:19:57 <kevinbenton> njohnston: thanks for your time so far. come back soon ;) 21:20:09 <ihrachys> I can do infra part too 21:20:16 <ihrachys> I am involved in it lately 21:20:27 <armax> ihrachys, dasm: appreciated, this is not a single man job 21:20:35 <armax> I am on the list too, but I can’t pull it off on my own 21:20:36 <dasanind> i would like to do it 21:20:46 <dasanind> but I will need some help initially 21:21:20 <kevinbenton> ok 21:21:23 <ihrachys> kevinbenton: seems like you should close the gate, or you will be flooded with volunteers 21:21:41 <dasm> ihrachys: i think it's better to have more volunteers, than lack of them :) 21:22:22 <ihrachys> I am happy to give a hand to anyone stepping in instead of taking the role 21:22:34 <kevinbenton> #info dasm, dasanind, armax and ihrachys to be infra liaisons 21:22:54 <kevinbenton> can one of you update the wiki? 21:23:15 <armax> kevinbenton: I am already an infra liasion!? 21:23:15 <ihrachys> I will do 21:23:26 <kevinbenton> armax: yes, apparently so :) 21:23:34 <armax> have I just beem demoted and promoted again? 21:23:47 <armax> there’s no respect for the elderly 21:23:50 <kevinbenton> armax: no, it's just a PUT operation 21:23:58 <kevinbenton> i have to respecify or you would be dropped 21:24:03 <kevinbenton> :) 21:24:14 <dasm> armax: "elderly"? really?! 21:24:25 <armax> btw 21:24:26 <armax> https://docs.openstack.org/developer/neutron/policies/neutron-teams.html 21:24:34 <armax> https://docs.openstack.org/developer/neutron/policies/neutron-teams.html#core-review-hierarchy 21:24:46 <armax> there’s a bit of dup here and there 21:24:57 <armax> it’s time to nuke dougwig 21:25:13 <dasm> poor dougwig :/ 21:25:27 <kevinbenton> ihrachys: can you catch that as well ^^ 21:25:36 <ihrachys> yes my lord 21:25:54 <kevinbenton> thanks :) 21:25:58 <kevinbenton> ok, moving on 21:26:12 <kevinbenton> #topic OSC 21:26:57 <kevinbenton> Abhishek reached out to me to let me know he couldn't make it to this meeting 21:27:17 <kevinbenton> so unless someone has any other updates for OSC we'll skip for this week 21:27:47 <kevinbenton> ok 21:27:50 <armax> ankur-gupta-f1: 21:27:54 <armax> he might have something to add 21:28:25 <dasm> i don't see ankur-gupta-f1 on his desk. 21:28:42 <armax> dasm: wow, I wish I had your x-ray vision 21:28:48 <kevinbenton> :) 21:28:51 <kevinbenton> ok 21:28:57 <dasm> armax: he's usually sitting next to me :) but not now 21:29:09 <armax> dasm: :) 21:29:15 <kevinbenton> i presume john-davidge is out partying at this hour 21:29:21 <kevinbenton> so no docs update this week 21:29:36 <ankur-gupta-f4> Hi 21:29:43 <ankur-gupta-f4> Phone irc 21:29:55 <kevinbenton> ankur-gupta-f4: hey, any updates for OSC you want to share? 21:30:13 <ankur-gupta-f4> Osc transition is nearing completion. Few core commands left 21:30:23 <ankur-gupta-f4> And waiting on the plugin advanced commands 21:31:10 <ankur-gupta-f4> https://review.openstack.org/#/c/439148/ 21:31:24 <ankur-gupta-f4> Has a complete info 21:31:32 <ankur-gupta-f4> Thats it 21:31:45 <kevinbenton> ankur-gupta-f4: thanks 21:31:56 <ankur-gupta-f4> Np 21:32:33 <kevinbenton> #topic neutron-lib 21:32:46 <kevinbenton> boden: any big changes after armax's context comes in? 21:33:00 <boden> hi 21:33:26 <boden> so we have one bigger one comming up that will replace neutron.callbacks with those in neutron-lib 21:33:29 <boden> #link https://review.openstack.org/#/c/439146/ 21:33:39 <boden> but it’ll have to wait for a new release of lib first 21:33:47 <armax> woot! 21:33:53 <ihrachys> +208, -1386? 21:33:57 <ihrachys> can I +100 it? 21:34:26 <ihrachys> boden: have we requested a new release already? 21:34:28 <dasm> boden: new release? we can think about that :) 21:34:33 <ihrachys> because I think kevinbenton wanted to speed releases up 21:34:37 <dasm> ihrachys: i didn't see anything 21:34:38 <boden> we need to wait for https://review.openstack.org/#/c/439610/ 21:35:15 <armax> boden: done 21:35:29 <kevinbenton> armax: HA! 21:35:30 <boden> once we have a release with ^, then we can proceed to start the fun 21:35:33 <dasm> boden: when this change will be merged, i'll cut new release 21:35:41 <kevinbenton> armax: i knew i would get you to review my gross decorator eventuallhttp://lists.openstack.org/pipermail/openstack-dev/2017-March/113416.html 21:35:50 <kevinbenton> how embarrassing 21:35:56 <kevinbenton> i just emptied my clipboard on accident 21:35:59 <armax> but with kevinbenton I am not sure if you won’t be in constant catch up mode 21:36:03 <armax> he likes to churn a lot of code 21:36:04 <ihrachys> at least it's not a password 21:36:22 <ihrachys> armax: just -2 all his patches 21:36:24 <armax> kevinbenton: damn it, you’re right 21:36:25 <kevinbenton> ihrachys: that's what you think :) 21:36:32 <armax> I intentionally left it uncommented 21:36:56 <armax> there, I left a comment 21:36:59 <kevinbenton> so after that goes in, a neutron-lib release would be nice 21:37:32 <boden> sure. it’ll free up the callback work anyway; and that one will be nice to get under our belt early in teh release 21:37:35 <kevinbenton> boden: so we're not leaving behind shims to keep existing imports working 21:37:44 <boden> no shims 21:38:35 <boden> there are a few others in the queue; I sent notes about them to the dev ML last week.. you can find them in the lib impact query in on the wiki 21:39:10 <kevinbenton> is there a reason we don't want the shims? 21:39:47 <ihrachys> kevinbenton: that's soft power ;) 21:39:50 <kevinbenton> i'm fine with it in this case since we have several months for people to adapt, just curious if there is a technical reason 21:40:11 <armax> kevinbenton: shims slow us down 21:40:21 <armax> and we ain’t got time for that 21:41:06 <boden> kevinbenton: we doc’d that appraoch somewhere.. I’m looking for it 21:42:03 <boden> kevinbenton: do you see some special need for shims? 21:42:42 <kevinbenton> boden: no, not in this case since we have such a large transition period 21:42:55 <kevinbenton> boden: i just wanted to make sure there wasn't some issue preventing us from using them in general 21:43:05 <kevinbenton> in case we get rid of something close to the end of this cycle 21:44:03 <boden> kevinbenton any more discussion on the callback approach? 21:44:06 <kevinbenton> #info dasm to make a neutron-lib release after https://review.openstack.org/#/c/439610/ merges 21:44:13 <kevinbenton> boden: nope 21:44:21 <boden> ok 1 last thing 21:45:00 <boden> armax and I were chatting last week… I started a curated list of neutron core “rehoming” patches — this would be a good list of core/driver folks to focus on reviewing if they can 21:45:16 <boden> #link https://etherpad.openstack.org/p/neutron-lib-core-review-queue 21:45:42 <boden> just want to make sure we get some experienced neutron core eyes on these patches :) 21:46:12 <boden> nothing else from me 21:46:12 <dasm> boden: ok, i won't look at this. noted 21:46:22 <boden> lol 21:46:26 <ihrachys> :) 21:46:47 <armax> boden: it’s on my bookmarks 21:47:14 <kevinbenton> ack 21:47:20 <kevinbenton> #topic open discussion 21:47:35 <kevinbenton> anyone else have anything to discuss? 21:48:21 <asingh_> https://review.openstack.org/#/q/topic:bug/1563069 21:49:01 <asingh_> there are 4 patches for centralizing config options remaining 21:49:28 <electrocucaracha> wsgi stuff? 21:49:30 <kevinbenton> https://review.openstack.org/#/q/topic:bug/1563069+is:open 21:50:10 <ihrachys> electrocucaracha: do we have devstack patch that makes use of your wsgi script? 21:50:30 <electrocucaracha> ihrachys: https://review.openstack.org/#/c/439191/ 21:50:56 <asingh_> kevinbenton as discussed on these patches ,merging them would affect stadium projects, what's the fate of these patches ? 21:51:02 <ihrachys> electrocucaracha: I still see neutron-server logs in logdir 21:51:17 <electrocucaracha> ihrachys: It's having a weird behavior but at least it deploys it 21:51:28 <ihrachys> well it's not used in actual tests? 21:51:44 <electrocucaracha> ihrachys: yeah, its' necessary to make neutron-api as default option 21:51:45 <kevinbenton> asingh_: it's fine to merge them now. i think if you send an email to the mailing list that would be good as a heads up 21:52:16 <asingh_> kevinbenton ok, thanks, will do 21:52:25 <electrocucaracha> ihrachys: I saw that you have been working on that, but not sure if I've to rebase my patch on all of yours 21:52:29 <ihrachys> kevinbenton: yes, I believe the only thing to be cautious about is to make sure we land them with giving stadium projects a chance to fix without squashing 21:52:38 <ihrachys> one breaking change in a week or smth 21:52:59 <kevinbenton> ihrachys: well for config import moving that seems like overkill 21:53:01 <ihrachys> electrocucaracha: no I abandoned all my patches for wsgi 21:53:12 <kevinbenton> ihrachys: it's something that makes sense logically to do in one change 21:53:19 <ihrachys> kevinbenton: well you can land all of them in one go, but do it in one go indeed 21:53:38 <ihrachys> I mean don't interleave with other breaking changes 21:53:44 <electrocucaracha> ihrachys: I mean the patches for deprecated neutron legacy in devstack 21:53:45 <ihrachys> like callback registry and stuff 21:53:53 <sindhu> kevinbenton: ihrachys : needed some direction for this: https://review.openstack.org/#/c/433234/ 21:53:57 <ihrachys> electrocucaracha: those are not relevant, not directly at least 21:54:41 <kevinbenton> sindhu: ok. i will add it to my review list 21:54:42 <ihrachys> ok if it was mentioned already, a heads-up for the team that I am working on switching whole gate to lib/neutron. Reviews welcome at: https://review.openstack.org/#/q/topic:new-neutron-devstack-in-gate 21:54:57 <sindhu> kevinbenton: thanks :) 21:55:01 <kevinbenton> electrocucaracha: this is running the RPC server inside of apache 21:55:13 <kevinbenton> electrocucaracha: is that the way we want to proceed? 21:55:30 <electrocucaracha> kevinbenton: afaik, it's only running api 21:56:03 <kevinbenton> electrocucaracha: where is it skipping the RPC listeners stuff? 21:56:04 <electrocucaracha> kevinbenton: https://review.openstack.org/#/c/409351/4/neutron/server/wsgi.py@29 21:56:09 <ihrachys> electrocucaracha: it runs whole core plugin, so it also runs RPC 21:57:28 <kevinbenton> ihrachys: we do have a separate rpc workers logic, we may want to move all RPC listens to there if they aren't already 21:57:41 <kevinbenton> we can continue this discussion offline 21:57:44 <electrocucaracha> ihrachys: regarding the devref, I'm gonna remove it once that it info is placed somewhere esle 21:58:00 <kevinbenton> anyone else have things to announce in the last 2 mins? 21:58:48 <kevinbenton> ok 21:58:52 <kevinbenton> that's all then 21:58:55 <kevinbenton> thanks everyone! 21:59:02 <dasm> thanks 21:59:04 <kevinbenton> #endmeeting