14:01:18 #startmeeting networking 14:01:19 Meeting started Tue Feb 16 14:01:18 2016 UTC and is due to finish in 60 minutes. The chair is ihrachys. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:23 The meeting name has been set to 'networking' 14:01:36 hi all 14:01:46 I expect the meeting today to be quite short 14:01:50 let's see... 14:01:56 Hi 14:02:11 #topic Announcements 14:02:11 hi 14:02:29 first thing first, we approach Mitaka-3 milestone 14:02:35 which is the last milestone in Mitaka, if you wonder :) 14:02:45 hii 14:02:48 the schedule can be found at 14:02:49 #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/086362.html 14:03:12 hi 14:03:13 I advice everyone to become aware of it 14:03:34 and to review a lot :P 14:03:41 one thing to note is that this time release folks want to be more strict about exceptions for libraries 14:03:51 and that means neutronclient too 14:04:10 so, please don't leave stuff for the last day, especially if you have client pieces to merge 14:04:14 yeah, there's a loophole there 14:04:17 or you may miss the release 14:04:24 some client pieces require the server side.. 14:04:35 client deadline is Mar 2 ;) 14:04:41 so dependencies are exchanged in time 14:04:45 I would say, pretty much all featureful pieces require it 14:04:56 I'm not sure it works 14:05:00 but ok 14:05:02 let's run 14:05:13 we'll see 14:05:22 exceptions are always there, if you ask me 14:05:41 ok, next thing to note is: voting for summit sessions is running for awhile already 14:05:47 and it will end tomorrow, Feb 17 14:05:55 o/ 14:05:59 so in case you haven't voted yet, it's time to do it 14:06:14 ok, let's move to upcoming events we may care about 14:06:16 O/ 14:06:29 as you all know, we have a midcycle in Minnesota 14:06:34 if you plan to go, please register at https://www.eventbrite.com/e/neutron-mitaka-mid-cycle-coding-sprint-tickets-21634783219 14:06:59 there are also some bug squashing events around the globe for Mitaka 14:06:59 one in Nuremberg 14:07:01 #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/086394.html 14:07:06 and one in Sillicon Valley 14:07:13 #link https://etherpad.openstack.org/p/OpenStack-Bug-Smash-Mitaka-BayArea 14:07:22 join them if you are around :) 14:07:33 no more announcements from anyone? 14:07:55 I guess nope. moving. 14:07:58 #topic Stable branches 14:08:07 * carl_baldwin has no idea there were eventbrite tickets 14:08:07 just wanted to give a clue about stable branches state 14:08:29 carl_baldwin: mestery sent the invite a couple of days ago 14:08:45 There are 14:09:00 ok, so stable branches 14:09:09 some of you probably noticed that kilo is broken 14:09:18 it's due to https://bugs.launchpad.net/python-fixtures/+bug/1542984 14:09:18 Launchpad bug 1542984 in Python Fixtures "fixtures 1.2.0 isn't compatible with testtools 2.0.0" [Undecided,New] 14:09:25 until it's fixed, nothing is mergeable 14:09:32 openstack stable team looks into it 14:09:39 so rechecks won't help for now. 14:09:46 no need to overload the gate 14:09:59 for stable/liberty, we have some sporadic failures 14:10:05 some of them suspiciously linuxbridge 14:10:32 but I need to triage logs from several jobs before starting to pull people into looking at the job 14:10:46 ok, that was purely informational. let's move on 14:10:50 #topic Bugs 14:11:23 mhickey: wanna update? 14:11:30 sure 14:11:31 mhickey: anything interesting the prev week? 14:11:57 one critical: https://bugs.launchpad.net/neutron/+bug/1543880; thanks kevinbenton 14:11:57 Launchpad bug 1543880 in neutron "duplicated security groups in test_port_security_disable_security_group" [High,Fix released] - Assigned to Kevin Benton (kevinbenton) 14:12:24 some open high: 14:12:26 https://bugs.launchpad.net/neutron/+bug/1543094 14:12:26 Launchpad bug 1543094 in neutron "[Pluggable IPAM] DB exceeded retry limit (RetryRequest) on create_router call" [High,Confirmed] - Assigned to Salvatore Orlando (salvatore-orlando) 14:12:37 https://bugs.launchpad.net/neutron/+bug/1544548 14:12:37 Launchpad bug 1544548 in neutron "DHCP: no indication in API that DHCP service is not running" [High,In progress] - Assigned to Gary Kotton (garyk) 14:12:47 releeased high: 14:12:49 https://bugs.launchpad.net/neutron/+bug/1542923 14:12:49 Launchpad bug 1542923 in neutron "Linuxbridge agent failed to start when bridge_mappings is given" [High,Fix released] - Assigned to Slawek Kaplonski (slaweq) 14:13:02 https://bugs.launchpad.net/neutron/+bug/1543880 14:13:02 Launchpad bug 1543880 in neutron "duplicated security groups in test_port_security_disable_security_group" [High,Fix released] - Assigned to Kevin Benton (kevinbenton) 14:13:11 https://bugs.launchpad.net/neutron/+bug/1544349 14:13:11 Launchpad bug 1544349 in neutron "neutron-vpnaas db migration breaks in MySQL 5.6" [High,Fix released] - Assigned to Al Miller (al-miller) 14:13:21 thats it. thanks to all for help! :) 14:13:36 cool. I see those assigned to seasoned folks, I believe it's under control 14:14:23 fyi this week it's rossella_s, and next one is dasm who are on bugs 14:14:29 thanks to all volunteers 14:14:32 let's move on 14:14:37 #topic Ongoing reviews 14:15:20 I highly encourage reviewers to focus on deliverables scheduled for MItaka 14:15:23 #link https://launchpad.net/neutron/+milestone/mitaka-3 14:15:49 to ease the pain of searching for the patches that correspond to the bps and bugs from the link above 14:15:59 we have the following link 14:15:59 #link https://goo.gl/Hb3vKu 14:16:02 kudos to rossella_s 14:16:18 that's gerrit dashboard with all Mitaka reviews 14:16:25 thanks ihrachys 14:16:31 it should be updated every day 14:16:42 I wrote to infra to find a common place where to put the link 14:16:51 rossella_s++ 14:16:57 rossella_s, awesome! 14:17:05 I see a lot of blueprints still in flight. 14:17:20 some of them will probably need to move to next cycle, depending on their shape 14:17:46 so please help your PTL to get a clue whether a feature is ready for merge 14:17:59 by updating blueprint whiteboard and RFE bugs 14:18:18 and then our glorious PTL will go and scrub them based on the info 14:18:42 I also remind that delivering a blueprint requires two people actively engaged: the implementer, and the approver 14:19:12 to implementers: if you see your patches are stuck with no reviews, please reach your approver to see how to move forward 14:19:31 to approvers: please go thru the list of blueprints assigned to you and update whiteboards, if not already. 14:19:41 it will be of tremendous help to your PTL 14:20:02 overall, please review as much as you can so that Mitaka is juicy :_ 14:20:13 one more procedural thing 14:20:29 please be mindful when merging stuff that is not scheduled for Mitaka 14:20:45 especially if the patch will evoke a huge number of git conflicts 14:21:02 hint: new gerrit UI shows the list of conflicting patches 14:21:10 so you can rely on it before pressing +W 14:21:33 git conflicts are not fun 14:21:45 especially now when we had 350 jobs in queue the prev week 14:22:18 anyone wants to raise some red flag for ongoing mitaka reviews? 14:23:06 * ajo checks how to see conflicting patches 14:23:10 ok I guess everyone just wants to get back asap to reviews, and coffee 14:23:24 oh, "Conflicts With" 14:23:29 ajo: it's in the right top corner of UI 14:23:29 right 14:23:38 DB migrations will always conflict because of HEAD, I guess :) 14:23:43 #topic Docs 14:23:49 Sam-I-Am: anything you'd like to advertise? :) 14:24:14 hi 14:24:22 not much, really. 14:24:32 still working on mitaka updates for the install guide and network guide 14:24:39 nice to see quite a few contributions from neutron folks 14:25:18 indeed 14:25:39 ok, thanks for the update Sam-I-Am 14:25:47 #topic Open Discussion 14:25:58 we have a lot of time woot! 14:26:02 i do have one thing... a bug i forgot about 14:26:04 https://bugs.launchpad.net/neutron/+bug/1523031 14:26:04 Launchpad bug 1523031 in openstack-ansible trunk "Neighbor table entry for router missing with Linux bridge + L3HA + L2 population" [Medium,Confirmed] 14:26:06 that's a problem 14:26:46 Sam-I-Am: should we raise the Importance? 14:26:58 yes 14:27:06 the other thing is trying to fix this mtu problem 14:27:11 seems like something we need to do before mitaka 14:27:37 Sam-I-Am: ok I will target it to M 14:28:37 Sam-I-Am: what would be the plan for MTU for Mitaka? 14:28:48 ihrachys: fix it? 14:29:00 we know whats wrong, its just making it work 14:29:34 the tl;dr - we need to get the 'calculated' mtu for each network distributed to all the components of that network such as veth pairs, bridges, etc 14:29:35 Sam-I-Am: yeah :) is it just a matter of setting network_device_mtu for all intermediate devices in hybrid setup? 14:29:59 yeah, probably not network_device_mtu but per-network, yes 14:30:19 Is there a bug number? 14:30:24 I guess we may fix it for M, if it's just that. is the bug targeted? what's the link" 14:30:54 network_device_mtu breaks stuff in different ways 14:30:59 more or less just kicks the can down the road 14:31:38 Sam-I-Am: do we have a bug number? 14:31:48 carl_baldwin: there's two bugs - https://bugs.launchpad.net/neutron/+bug/1542108 https://bugs.launchpad.net/neutron/+bug/1542475 14:31:48 Launchpad bug 1542108 in neutron "MTU concerns for the Linux bridge agent" [High,Confirmed] - Assigned to Sean M. Collins (scollins) 14:31:49 Launchpad bug 1542475 in neutron "MTU concerns for the Open vSwitch agent" [High,Confirmed] - Assigned to Sean M. Collins (scollins) 14:32:05 exactly whats wrong and how to fix it is explained there 14:32:18 Sam-I-Am: thanks 14:32:22 solving this once and for all is a big win for neutron 14:32:28 this is a HUGE problem 14:32:51 ihrachys, I also have a short (doc related) topic regarding source format of graphics in the networking guide 14:33:12 Sam-I-Am, is this something we even hope to solve by M-3? 14:33:16 Sam-I-Am: agreed. let's make it happen. I will target bugs to M. 14:33:39 jschwarz: it may be considered a bug fix, so could probably go after M3. 14:33:51 ack 14:34:19 Sam-I-Am: ok bugs are already targeted. let's talk offline on how to make it happen. we'll need Sean too. 14:34:29 scheuran: shoot 14:34:32 thx 14:34:34 do we have a rule how what the source format of graphics for the networking guide should be? Many of the are .graffle which you cannot edit on a linux system.. 14:34:35 Who is impending the mtu fix? 14:34:57 I used draw.io to paint some nice graphics and uploaded the source xml 14:35:09 Just targeting for M won't get it done. 14:35:17 but this causes troubles cause you need to exlcude every source file by hand as some xml parser will complain about it 14:35:28 should it be .svg? I mean is there a rule? 14:35:28 Sam-I-Am: ^ 14:35:29 carl_baldwin: bugs are on Sean, but I gotta help. 14:35:46 ihrachys: ack 14:35:47 carl_baldwin: sean and i fixed a couple of things 14:35:54 namely around path_mtu and advertise_mtu 14:36:08 but we need additional discussion about the best way to make path_mtu less... obscure. 14:36:16 are you at the MC next week? 14:36:36 Sam-I-Am: yes, I'll be the first three days. 14:36:59 I won't be there, in case you ask me. 14:37:25 carl_baldwin: we can either discuss it here real soon now and figure it out, or next week 14:37:32 well, not here, but in -neutron 14:37:44 right 14:37:56 Sam-I-Am: do you have a comment for scheuran questions? 14:38:09 ihrachys: yeah... 14:39:04 scheuran: the docs folks dont care what you use to draw diagrams, as long as they conform to the standards/look and have some sort of 'open source'-ish version of the file 14:39:15 i say open source-ish because it turns out there's no real open source diagram format 14:39:24 svgs are very much vendor-flavored 14:40:02 i draw most of the diagrams for the docs and use omnigraffle because it makes me life a lot easier... so i upload both .graffle and .svg in addition to the .png or whatever 14:40:29 the docs team has been down this road many times and there's just too much bike shedding that prevents work from getting done 14:41:22 scheuran: I guess that addresses the question 14:41:34 ok so svg is the choice 14:42:08 thanks 14:42:14 ok, I guess that's it for today, but I will give 1 minute for anyone to raise another flag 14:42:35 scheuran: for now, yeah :/ 14:43:12 ok let's call it a day 14:43:15 thanks folks! 14:43:15 #endmeeting