21:00:23 #startmeeting networking 21:00:24 Meeting started Mon Dec 7 21:00:23 2015 UTC and is due to finish in 60 minutes. The chair is armax. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:25 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:26 hi 21:00:27 hey all 21:00:28 The meeting name has been set to 'networking' 21:00:40 o/ 21:00:41 hello 21:00:41 hello all! 21:00:45 o/ 21:00:48 hello 21:00:56 #link https://wiki.openstack.org/wiki/Network/Meetings 21:01:00 yo 21:01:02 hi folks 21:01:05 Hello 21:01:06 emagana: back to your original nickname :-) 21:01:08 hello 21:01:10 * mestery looks for gary's doctor's note 21:01:17 * regXboi wanders in late 21:01:19 mlavalle: yeap.. laptop all receoved 21:01:23 recovered* 21:01:35 mestery: he’ll have to supply one after he’s fully recovered, I suppose 21:01:42 armax: Damn straight :) 21:01:42 mestery: garyk is celebrating Hanukah 21:01:46 Mitaka 1 is past us 21:02:03 o/ 21:02:13 o/ 21:02:19 o/ 21:02:29 armax : congratulations for M1 21:02:30 I suggested that we take this week’s meeting to talk mostly about blueprints 21:02:30 http://lists.openstack.org/pipermail/openstack-dev/2015-December/081468.html 21:02:34 #link http://lists.openstack.org/pipermail/openstack-dev/2015-December/081468.html 21:02:58 o/ 21:03:06 so I’ll go over the next section really quick 21:03:14 #topic Announcements 21:03:21 Mitaka-1 is out 21:03:23 #link https://launchpad.net/neutron/+milestone/mitaka-1 21:03:53 now the release schedule is captured here: 21:03:55 #link http://docs.openstack.org/releases/schedules/mitaka.html 21:04:01 o/ 21:04:17 o/ 21:04:43 another news worth sharing a global pep8 validation https://review.openstack.org/#/c/253676 21:05:01 armax: that's good news. 21:05:16 this means that the execution of your local pep8 job may pass 21:05:22 but the global one may fail 21:05:32 armax: how is it? why? 21:05:36 are there any differences? 21:05:44 dasm_: yes, look at that link 21:05:45 armax: how so? cached sdist? 21:06:00 dougwig: because the extra check is done in the gate 21:06:08 job that calls the pep8 21:06:12 tox target 21:06:32 we could detect the gate, and run the same check locally. 21:07:03 dougwig: we could, but we painted a while when the patch was posted 21:07:15 dougwig: you’re welcome to continue to paint if you like, but not now :) 21:07:21 we are artists 21:07:24 armax: indeed. 21:07:24 armax: what did you paint? 21:07:27 a bikeshed? 21:07:30 salv-orlando: yes 21:08:04 #topic Bugs 21:08:08 dougwig: anything worth sharing? 21:08:34 for the first time we reversed the trend 21:08:36 not a crazy week. lots of rfe's, one gate meltdown due to pylint requirements, since fixed, and one intermittent failure in the lbaas gate. 21:08:49 one item of news, linuxbridge is now voting and in our gate. 21:08:58 woohoo 21:08:58 yay for linuxbridge 21:09:00 * sc68cal hides 21:09:01 we went from 350+ bugs active to 250+ 21:09:02 Nice work sc68cal and dougwig :) 21:09:08 it's all sc68cal 21:09:14 boo for dougwig 21:09:15 ;) 21:09:17 sc68cal: get back here 21:09:18 I guess I should say that I believe oslo-utils is rather sick right now 21:09:29 regXboi: but the fix is already merged 21:09:37 dougwig: was the decision already made? 21:09:40 amuller: cool - I missed that 21:09:56 regXboi: or approved rather, anyway nothing we can do but wait 21:09:58 armax: the review merged in about 3 nanoseconds; i had wanted to discuss it here first. 21:10:00 A bunch of bugs expired recently. I think Cedric can get the kudos for that, from what he did 2 months ago. 21:10:06 amuller: ack - I have a recheck queued and waiting :) 21:10:18 dougwig: I noticed the on demand agenda item, yes 21:10:34 armax: we can still discuss that, and revert based on the result. but as of *right now*, it's voting/gating. 21:10:58 I'd like to talk about making multinode-full voting again when we get there :) 21:11:07 dougwig: ok 21:11:10 regXboi: i'd want to retire single node if so. 21:11:18 dougwig++ 21:11:24 dougwig: we can’t retire the single node, no 21:11:31 dougwig: ++ 21:11:48 armax: what's your conditions for retiring singlenode-full for multinode-full? 21:11:53 if other integrated projects still rely on it 21:11:55 too bad we cant use containers to emulate multi-node on a single vm 21:12:19 Sam-I-Am: that's .... sort of what functional does 21:12:45 Sam-I-Am: I think just net namespaces though 21:12:48 dougwig: anything else for bugs? 21:12:55 We could use Kolla eventually to deploy and run tests on that 21:12:56 armax: no, that's it. 21:13:04 actually, one bit. 21:13:07 amuller: or openstack-ansible :) 21:13:13 Sam-I-Am: or that yeah 21:13:14 lol 21:13:16 bug deputy was incredibly educational. i recommend that everyone try it. 21:13:35 either way we can have much more thorough testing using containers 21:13:42 I'd really like to explore that multinode/singlenode topic more thoroughly later this week 21:13:47 dougwig: kudos to you for stepping up 21:13:49 dougwig: maybe someday i'll grow up enough for it :) 21:13:56 regXboi: sure, let’s take this offline 21:14:04 armax: ack 21:14:10 #topic Blueprints 21:14:22 we are now in M-2 21:14:25 #link https://launchpad.net/neutron/+milestone/mitaka-2 21:14:32 which ends the week of Jan-16 21:15:02 we have a huge backlog 21:15:22 of 20+ blueprints 21:15:50 and a number of RFE’s 21:16:10 in other words, they won't all land, so can we nuke some, to focus review resources? 21:16:25 that’s not necessarily true, if we stay focussed 21:16:33 O 21:16:46 looking at assignments 21:16:47 #link https://blueprints.launchpad.net/neutron/mitaka/+assignments 21:16:58 we are relatively balanced right now 21:17:24 but there’s always more that meets in the eye 21:17:29 armax: who is the approver for keystone v3? 21:17:30 focus = psychological conditioning a la a clockwork orange? 21:17:32 do we have criteria for BPs that we think aren't going to make mitaka and can be moved out? 21:17:41 regXboi: there is no one 21:17:48 dasm__: that was my point 21:17:51 regXboi: good question 21:17:57 amrax: I have a question about RFE #link https://bugs.launchpad.net/neutron/+bug/1511574 - Have we reached enough of a consensus that I can go ahaead with a WIP patch, or would you like me to write a spec first? 21:17:57 Launchpad bug 1511574 in neutron "[RFE] Support cleanup of all resources associated with a given tenant with a single API call" [Wishlist,Triaged] - Assigned to John Davidge (john-davidge) 21:17:58 regXboi: let’s go in order thoug 21:18:00 though 21:18:04 armax: ^ 21:18:05 I think some keystone v3 patches landed over the weekend, right? 21:18:05 armax: ok 21:18:14 Maybe some small initial ones 21:18:15 Sorry, back on track 21:18:32 #link https://blueprints.launchpad.net/neutron/+spec/l2-api-extensions 21:18:39 mystery it broke devstack... 21:18:41 rossella and ihar are not here 21:18:43 mestery 21:18:45 so that’s easy 21:18:51 ajo: you there? 21:18:56 What broke devstack xgerman? 21:19:03 xgerman: The v3 changes? 21:19:14 I’ll have to reach out to them offline 21:19:22 yep, but I think it’s fluid 21:19:40 kevinbenton: ping 21:19:41 xgerman: nice 21:19:42 john-davidge: Sorry I didn't respond on that bug. Did you consider the idea of doing this with a client side script so that we wouldn't have to commit to an API, provided I push the cross project spec? 21:19:47 #link https://blueprints.launchpad.net/neutron/+spec/port-scheduler 21:19:59 john-davidge: or we can take this offline after the meeting 21:20:02 carl_baldwin: ping 21:20:06 armax: here 21:20:10 john-davidge, amuller: I thought there was a blueprint to do something similar as part of mistral 21:20:10 armax: hi 21:20:38 it looks like this is missing code, specs, etc 21:20:53 regXboi: I'll check that out thanks 21:20:57 what’s the latest? 21:21:05 amuller: ack 21:21:08 amuller: I’d prefer to expose it via the API from the beginning, but I’m open to discusion. We can take it offline 21:21:17 Yes, it won't really go anywhere until we get the routed networks figured out I think 21:21:32 kevinbenton: so there’s a strict depedency between the two? 21:21:47 regXboi: Do you have a link to the BP? 21:21:55 I think that's how RAX was solving their issue - segment aware scheduler 21:21:57 armax: I think the port scheduler is a piece that routed networks needs. I'm not sure about the other way around. 21:21:59 armax: my original hope was that it would be useful for get me a network too 21:22:08 john-davidge: not handy, 'll have to look offline 21:22:13 er *I'll* 21:22:18 regXboi: Ok, thanks 21:22:26 armax: but that is taking a different route than I envisioned 21:22:27 since work for l3 networks is still in progress 21:22:37 ok. Sounds good if you guys put something in writing to shed some light on it? 21:22:48 armax: I can do that 21:22:53 armax: I'll help. 21:22:55 i need to split a bit early today... there's no docs stuff anyway 21:23:01 even something in the whiteboard would suffice to start with 21:23:08 armax: There is some stuff that should be vetted with Nova. Having it in writing would help. 21:23:35 #link https://blueprints.launchpad.net/neutron/+spec/multi-l3-backends 21:23:47 kevinbenton: you’re tagged again 21:24:08 Working on the spec for that now 21:24:14 kevinbenton: ok cool 21:24:14 Hoping to get it up this week 21:24:32 kevinbenton you’re not oversubscribed are you? 21:24:59 kevinbenton: currently you have 4 bp’s you’re playing a major role on 21:25:17 Well several are blocked pending other stuff 21:25:50 The agent notification refactor can't move forward until objects are done 21:26:02 kevinbenton: ok 21:26:13 #link https://blueprints.launchpad.net/neutron/+spec/get-me-a-network 21:26:19 HenryG, haleyb: ping 21:26:27 armax: yo 21:26:37 any news on this one? 21:26:43 did we move an inch? 21:27:11 I am playing with the patch, but still need to sync up with haleyb 21:27:13 i continue to work on the code, didn't talk to henry last week 21:27:32 haleyb: we'll talk tomorrow 21:27:54 HenryG: ack 21:27:56 once you chat, can you guys share your notes on the ML? 21:28:05 to keep the team in the loop? 21:28:09 armax: will do 21:28:34 HenryG: thanks! 21:28:37 #link https://blueprints.launchpad.net/neutron/+spec/external-dns-resolution 21:28:42 carl_baldwin, mlavalle: ping 21:28:48 armax: hi 21:28:58 this must be pretty close, no? 21:29:02 what’s missing? 21:29:16 functionality is complete. Last night I finished debugging all the code 21:29:39 right now I am cleaning up the Jenkins tests and after that I think it is very close 21:29:53 expect Jenkins tests to be clean tomorrow night 21:29:53 mlavalle: so we’re feature complete, but we miss tests and docs? 21:30:03 feature complete both server and client side? 21:30:23 armax: feature complete both sides 21:30:40 I am adding doc in dervref as I am pushing patchsets 21:30:48 I added a doc to devref last night 21:30:48 ok, so tests and docs missing then? 21:31:02 mlavalle: I'm on tap for reviews. :) 21:31:11 and I have to add some unit tests 21:31:14 carl_baldwin: the hose is better? 21:31:16 : 21:31:17 :) 21:31:29 lol 21:31:44 #link https://blueprints.launchpad.net/neutron/+spec/modular-l2-agent 21:31:52 also thanks fto garyk... he's been providing instantaneous reviews over the past 3 weeks, providing excellent guidance 21:32:05 garyk kudos 21:32:11 ihar is not here 21:32:13 armax: I think I can comment on this spec 21:32:19 sc68cal: go for it 21:32:20 I push in the evening and by the following morning I have a complete review from him 21:32:29 sc68cal: link please? 21:33:15 sc68cal: ? 21:33:22 armax: for the spec or code? I think it is just an rfe - https://bugs.launchpad.net/neutron/+bug/1468803 21:33:22 Launchpad bug 1468803 in neutron "[RFE] Create a modular L2 agent framework for linuxbridge, sriov and macvtap agents" [Wishlist,In progress] - Assigned to Andreas Scheuring (andreas-scheuring) 21:33:50 sc68cal: on anything you were going to comment on :) 21:34:03 I think Andreas is making good progress, he has a couple patches up that breaks the LB agent into pieces so that the main agent can be re-used 21:34:25 the main target is having a macvtap piece that re-uses like 90% of the LB agent code 21:34:31 sc68cal: some stuff got caught in the cleaunup tool that abandons patches 21:34:35 btw 21:34:48 armax: yeah I think those patches were mine and they were correct to be gc'd 21:34:56 sc68cal: ok cool 21:34:59 since we ended up going in another direction. Anyway progress is good. 21:35:06 * sc68cal yields his time 21:35:12 sc68cal: tahnks! 21:35:39 #link https://blueprints.launchpad.net/neutron/+spec/restructure-l2-agent 21:35:45 yamamoto, rossella_s: ping 21:36:26 ok, I’ll take this offline 21:36:33 If rossella_s isn't here, I know she was close to done 21:36:44 kevinbenton: there was a setback 21:36:51 where we reverted one of her change 21:36:56 Just testing IIRC at this point to make sure the patch doesn't cause a regression again 21:36:58 I am not sure if that got resubmitted though 21:37:06 I know, I mean close to that being fixed 21:37:13 I’ll touch base with her anyhow, thanks for the update 21:37:17 #link https://blueprints.launchpad.net/neutron/+spec/neutron-lib 21:37:18 kevinbenton, armax: it's public holiday in rossella_s town, so I think she's not online today 21:37:34 salv-orlando: ok 21:37:49 dougwig, mestery: how’s it this going? 21:38:21 armax: in progress, i need to write the current plan down, and next big hurdle is coming up with a plan of attack for parts of neutron.db. no way will it be finished in m-2. 21:38:40 Some small things are merging at least, but yeah, the bigger stuff not yet 21:39:04 ok cool, let’s keep at chewing at this one 21:39:10 small progress is better than no progress 21:39:20 kudos 21:39:22 #link https://blueprints.launchpad.net/neutron/+spec/routed-networks 21:39:30 carl_baldwin, kevinbenton: you’re up 21:39:31 it'll go quick once the db stuff is sorted, but that'll be a big rat to swallow. 21:39:48 dougwig: ack 21:39:59 carl_baldwin, kevinbenton: it looks like we’re still going over the spec phase 21:40:08 armax: still iterating the spec. 21:40:24 Yes, the current state is something that will work but will require a ton of work 21:40:28 Some code is in progress as poc 21:40:43 carl_baldwin: do you need anyone/anything to get to a final consensus? 21:41:31 armax: Let me check the state of the spec since the weekend and see where it is at. 21:41:43 carl_baldwin: ok 21:42:06 #link https://blueprints.launchpad.net/neutron/+spec/add-tags-to-core-resources 21:42:13 gsagie isn’t ehre 21:42:21 and neither is gary 21:42:25 I’ll take this offline 21:42:38 #link 21:42:43 #link https://blueprints.launchpad.net/neutron/+spec/dhcp-options-per-subnet 21:42:52 sambetts-afk, amotoki: ping? 21:43:16 I’ll take this offline 21:43:29 #link https://blueprints.launchpad.net/neutron/+spec/fwaas-api-2.0 21:43:37 sc68cal, xgerman: ping 21:43:42 armax: pong 21:43:44 where are at herE? 21:43:50 still iterating on the spec? 21:44:05 yes, but we also divvied up the coding work 21:44:25 We are close, there is a couple small API details to work out but strong consensus is present 21:44:33 sc68cal: that’s nice 21:44:52 so I expect us to start work on the non controversial stuff in the meantime 21:44:59 xgerman: cool 21:45:01 likely will miss M2 21:45:03 xgerman: +1 21:45:05 though 21:45:14 not for the API 21:45:29 ok, that’s your piece — just being cautious ;-) 21:45:46 I mean the API spec finished probably end of this week 21:45:55 xgerman: so long as we make small tangible progress that doesn’t matter 21:46:07 k, will have API in M 21:46:24 #link https://blueprints.launchpad.net/neutron/+spec/improve-dvr-l3-agent-binding 21:46:29 carl_baldwin, obondarev: ping 21:46:42 * carl_baldwin will defer to obondarev 21:46:53 carl_baldwin: this doesn’t have anything on the whiteboard and no RFE bug attached 21:47:07 so it looks like there’s nothing tracking this effort? 21:47:25 armax: Yikes. I can create something. How would you prefer it? 21:47:36 Over easy 21:47:45 carl_baldwin: if you know of bugs, patches that are related to it 21:47:56 carl_baldwin: it’d be nice to capture this information on the whiteboard 21:48:14 carl_baldwin: even if this meant chaning the patche’s commit messages to reference the bleprint 21:48:18 carl_baldwin: thanks 21:48:19 armax: ok 21:48:25 obondarev: ^? 21:48:43 #link https://blueprints.launchpad.net/neutron/+spec/rename-tenant-to-project 21:49:03 dasm_: ping 21:49:05 here 21:49:20 mordred started with few patches 21:49:24 it looks like we’re making small progress here 21:49:30 one tweak left and then it's done 21:49:34 indeed 21:49:37 HenryG: link? 21:49:47 done, is a bold statement 21:49:49 :) 21:49:53 we need an approver here 21:49:55 any volnteers? 21:50:10 it looks like we want to make some headways in this releae 21:50:11 release 21:50:13 well, the api part 21:50:19 https://review.openstack.org/253782 21:50:21 HenryG: oh right 21:50:33 woot 21:50:39 armax: i can review. 21:50:41 armax: what sort of headways? 21:50:43 782 21:50:47 HenryG: wanna be the approver? 21:51:02 dougwig: no, I want you to be fully focused on neutronlibs 21:51:02 sure, or amotoki 21:51:11 njohnston: what’s 782? 21:51:14 * dougwig crawls back into his lib hole. 21:51:15 ok 21:51:23 dougwig: :) 21:51:23 i think that was the ending of the patchset 21:51:25 apologies, some line noise with my IRC client 21:51:34 njohnston: no problemo 21:51:45 do we need extra spec for this or ML discussion is enough? http://lists.openstack.org/pipermail/openstack-dev/2015-December/081304.html 21:51:46 dougwig: don’t worry I have something else for you  21:52:03 * regXboi shouts "run, dougwig, run!" 21:52:15 dasm_: if we want to be polished we’d take the ML conversation and put into spec format 21:52:24 armax: ack 21:52:44 #link https://blueprints.launchpad.net/neutron/+spec/performance-counters 21:52:49 regXboi, mestery: ping 21:52:50 yo 21:52:55 yo 21:53:12 part I of devref merged, collector PoC is open for reviews, starting to work on how to store 21:53:16 (part II of devref) 21:53:30 I'm almost ready to change status to good progress 21:53:34 but hadn't yet 21:54:06 ok 21:54:11 that's all I have 21:54:18 there’s a part III too is there not? 21:54:23 there will be 21:54:37 ok 21:54:38 but I'm planning that for after the PoC that codes part II 21:54:48 M-2 deadline is up in the air 21:55:12 regXboi: as I mentioned 21:55:20 there’s really no milestone planning 21:55:30 everything targets the release and we roll over stuff as it goes 21:55:46 that was just my guessing progress vs time 21:55:46 #link https://blueprints.launchpad.net/neutron/+spec/vm-without-l3-address 21:55:57 ajo: ping? 21:56:21 I’ll take this offline 21:56:31 #link https://blueprints.launchpad.net/neutron/+spec/security-group-logging 21:56:34 a little late for ajo 21:56:45 sc68cal, yushiro: ping 21:57:17 hi. i am behaft of yushiro 21:57:45 we re still on progres of making spec 21:58:02 s/re/are 21:58:09 ok 21:58:19 #link https://blueprints.launchpad.net/neutron/+spec/autogen-neutron-conf-file 21:58:21 mhickey: ping 21:58:27 pong 21:58:29 mhickey: what’s left for this to be complete? 21:58:41 bulk of the code merged, hasn’t it? 21:58:53 AFIK just the *aas parts; will need to sync with ihrachys then 21:58:57 ok 21:59:07 yes, the core parts; even the file removals! :) 21:59:24 small call out.. 21:59:25 timecheck 21:59:26 mhickey: *-aas people please help with the reviews 21:59:32 1 mins 21:59:40 actually 22 seconds 21:59:41 we managed to do only half 21:59:42 mhickey: just +2'd the fwaas one 21:59:44 yes;read my mind! :) thanks armax 21:59:51 you guys want to continue this next week? 21:59:59 or shall try and continue offline? 22:00:04 sc68cal: thanks! :) 22:00:06 #endmeeting