14:01:29 #startmeeting networking 14:01:29 o/ 14:01:31 Meeting started Tue Mar 1 14:01:29 2016 UTC and is due to finish in 60 minutes. The chair is ihrachys. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:31 o/ 14:01:32 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:33 hiho 14:01:34 o/ 14:01:34 Hi 14:01:35 The meeting name has been set to 'networking' 14:01:35 o/ :) 14:01:35 o/ 14:01:36 o/ 14:01:36 hi 14:01:38 \o/ 14:01:39 o/ 14:01:39 hi 14:01:42 hello my friends ;) 14:01:44 hi 14:01:47 hi all 14:01:54 o/ 14:01:59 * amuller salutes supreme commander chair 14:02:05 we'll try to have it short today since everyone is deep into mitaka3 pieces ;) 14:02:08 moo. 14:02:16 aloha 14:02:19 hi 14:02:22 #link https://wiki.openstack.org/wiki/Network/Meetings Agenda 14:02:23 * ajo unburies head from code 14:02:33 #topic Announcements 14:02:49 ok, first thing first, we are approaching Mitaka 3 cut-off day 14:03:02 if you wonder when it happens for neutron... it's some time this week 14:03:05 Hello 14:03:12 no exact day known so far 14:03:14 I just pushedt he patch 14:03:18 OR am pushing 14:03:23 So it happens today or tomorrow 14:03:23 our glorious PTL will make the final judgement 14:03:27 mestery: hurry! :) 14:03:27 #link https://review.openstack.org/286609 14:03:53 right, it usually happens some time around Wed, and probably this time we'll stick to the tradition 14:04:07 please wed ;) 14:04:17 I see no reason we wouldn't ;) 14:04:17 though we may accommodate for a bit if there are patches in flight that may be better getting in before the cut-off 14:04:36 hi 14:04:39 so its Wednesday UTC ? 14:04:47 so long story short, you are short on time, so push harder if you want something in 14:04:58 * Sam-I-Am chants mtu 14:05:17 another thing to note is we not only release neutron server packages (neutron + *aas) but also the client 14:05:21 probably this week 14:05:36 4.1.0 is preparing https://review.openstack.org/#/c/286585/ 14:05:43 so if you have pieces in the client for features landed in server side, better hurry 14:05:52 just waiting armando. 14:06:17 right. our glorious PTL will do the final call here too. :) 14:06:24 perhaps 4.1.0 will be shipped soon and if needed we will release 4.2.0 too 14:06:38 armax and mestery will take care of them :) 14:06:47 that's fair. client patches are usually more safe-ish than server side 14:07:00 amotoki: Some of the changes post 4.1.0 look like we may need to bump the version again, so yes, a 4.2.0 would be the one 14:07:40 and the final announcement is... 14:07:40 yeah. as far as i see, it is backward-compat. 14:07:41 as you all probably know, a lot of team members were gathering in Minnesota wilderness the previous week for a midcycle sprint 14:08:01 we have Armando report from the event on the ML: 14:08:02 #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/087693.html 14:08:22 please make yourself aware of that, and reply if you have questions 14:08:33 thanks everyone who got there and made progress 14:08:44 why was it not in europe? :) 14:08:49 it seems very productive :) 14:09:29 garyk: The mid-cycle? 14:09:31 garyk: next time let's bug our PTL to get it somewhere in Europe ;) 14:09:39 mestery: yes :) 14:09:51 garyk: We were not even going to have a mid-cycle, but dougwig and armax decided at the last minute to go ahead with it 14:09:58 So I scrambled to find a location on very short notice 14:10:01 That's why :) 14:10:13 ihrachys: Lets see how Brno goes! :) 14:10:19 mestery: yeah, that's one bad thing about the latest midcycle: short notice :) 14:10:29 Yes 14:10:29 mestery: garyk: True. AFAIR, there was an email in mid-Dec or sth, that there will be no mid-cycle in this release 14:10:29 mhickey: Brno is limited in the scope and attendance, so it's not the same 14:10:33 * mestery still thinks there should be no mid-cycles 14:10:35 Rochester was very nice. 14:10:41 mhickey: It was! 14:10:46 i agree with mestery 14:10:49 mestery++ 14:10:52 it was nice to see everyone in person 14:11:04 we get to beat each other up in person 14:11:06 mestery: I bet then you take part in the thread on splitting the summit, don't you? 14:11:16 ihrachys: I'm too old to take part in that thread ;) 14:11:23 But yeah, split them up! 14:11:46 yeah, it will be like the staidum, split them up and then get them back together and then split them up.... 14:11:49 It was good for me to meet people for the first time. 14:11:57 grays +1 14:12:02 lol 14:12:06 garyk +1 14:12:10 garyk :+1 14:12:11 garyk: it's "agile" 14:12:15 sarcasm is great with garyk 14:12:17 "iterative" 14:12:17 garyk: openstack: it's all about process 14:12:23 :) 14:12:23 garyk: We have no idea what the next PTL will do, do we? 14:12:41 he might run on some stadium platform ;- 14:12:45 i hope that she/he does the right thing 14:12:46 and indeed I'm sure mestery was very supportive of midcycles 2 years ago 14:12:50 mestery: unless we keep the one well known to us ;) 14:12:59 * Sam-I-Am runs for PTL 14:13:00 salv-orlando: I'm nothing if not a flip-flopper, that's for sure 14:13:21 mestery: you should run for president 14:13:21 garyk: runs from the PTL.... 14:13:21 mestery: False, I've never once seen you wear flip flops 14:13:25 mestery: nah it's just that since you had your 4th your wife has threathened you to not live home anymore 14:13:26 leave 14:13:30 amuller: lol 14:13:36 salv-orlando: ha! 14:13:37 garyk : hehe... 14:13:46 ptl sounds like such a fun job lol 14:13:53 amuller: Challenge accepted for Austin :) 14:14:17 i think that it is cheaper to dye ones hair grey 14:14:18 * ihrachys waits until people waste all their jokes for the week 14:14:40 i think that we have digressed a little from the agenda 14:14:51 oh really? ok let's move on then ;) 14:14:58 #topic Bugs 14:15:01 we never digress from anything 14:15:07 +1 14:15:11 its called agile meeting 14:15:18 un conference 14:15:19 we have two things here. one is dasm report for the prev week 14:15:30 dasm: anything mind blowing from the week you served? 14:15:34 ihrachys: True. There was a very calm week. 14:15:49 we had just one big incident with api gate 14:16:04 where tempest_lib broke everything and glorious PTL(TM) fixed it 14:16:28 now that's the Leader! 14:16:32 and second one, when by accident neutron-fwaas broke support for postgres 14:16:47 dasm: ok I guess you tagged all bugs as per rules, and all went smooth 14:16:49 and, about this postgres thing. we lack gate for this, because it would be visible earlier 14:17:03 ihrachys: i'm still buried under the pile of bugs :D 14:17:08 but progressing 14:17:17 dasm: I think we had postgres jobs before, but they never got enough attention so were dissolved 14:17:37 they come and go 14:17:53 I was for removing but... 14:18:10 mhm. in this case at least two different projects noticed problems with db. so, probably we should again think about it 14:18:16 just my opinion 14:18:31 postgres got "sacrificed" - I believe - to limit gate job sprawl 14:18:35 dasm: I guess as long as you drive the job and fix it quickly, no one would object more coverage 14:18:59 if sprawl is the issue, periodic jobs are always there as a mid solution 14:19:12 as long as we don't have postgres-dvr, postgres-lb, postgres-*... 14:19:27 amuller: we should also add python34 to the matrix 14:19:31 and centos vs. ubuntu maybe 14:19:34 that would be fun 14:19:39 ok. 14:19:45 there is some patch out to restore a postgres job somewhere 14:19:50 ihrachys: sure. The sacrified happened before people started moving jobs to the periodic queue. a postgres job is useful because there's always the odd failure due to the different default isolation level 14:20:02 iirc there was an idea of a periodic job rather than gate jobs, but I don't know how it goes. 14:20:06 and why not batman vs superman? 14:20:11 * salv-orlando disappears 14:20:30 the 2nd thing on bugs is... we don't have anyone as the bug deputy for the week, neither for the next one. 14:20:45 PG is back in the check queue, non voting. https://review.openstack.org/283812 14:20:55 so if you feel like you want to please your glorious PTL and the team, it's time to make a step forward 14:21:22 HenryG: oh cool! thanks for the link 14:21:27 HenryG: ++ 14:22:00 anyone for the bug deputy thing? 14:22:26 please, take it... it's fun! 14:22:35 you heard that week was quiet. I bet it's the best time to get in, since I suspect this week to be as quiet as the previous one. ;) 14:23:10 i cant take on any more stuff right now :/ 14:23:29 note we have a table of people who already served the job for the project 14:23:30 #link https://wiki.openstack.org/wiki/Network/Meetings#Bug_deputy 14:24:02 it is now a big table! too wide to my browser 14:24:09 ihrachys, I can take over this week 14:24:16 amotoki: buy bigger monitor 14:24:17 I guess we reach the time when we should wrap up the list 14:24:17 scheuran: great! 14:24:20 scheuran: thanks a lot man. 14:24:29 we also need to cover the next week. 14:24:36 dasm: right solution! 14:24:37 scheuran rocks! 14:25:01 ok, let me trigger the wrap up, I will take the next week after that. 14:25:17 we should probably think of reformating that table to something else :) 14:25:26 ajo++ 14:25:26 go ihrachys! 14:25:48 ajo: so... you're a volunteer for reformatting? 14:25:56 #topic Docs 14:26:01 Why do we need a matrix for the table ?? 14:26:01 i recommend a 3-d table :) 14:26:09 scheuran:++ 14:26:10 Sam-I-Am: anything worth noting before Mitaka3 cut off? 14:26:14 3d +1 14:26:17 dasm, :D next time I go for bug deputy I will fix it , I'll be happy to doit after FF 14:26:24 scheuran:+1 14:26:26 ihrachys: not much. trying to push some patches into the networking guide. 14:26:36 ihrachys: ++ 14:26:36 good to see a lot of contributions from neutrinos 14:27:02 if we get MTU stuff merged, its probably fewer docs because it Just Works as it should 14:27:06 rather than being a mess of workarounds 14:27:12 so, removing docs is nice. 14:27:27 we'll get there, seems like we are on good track with MTU work 14:27:36 ok folks, one thing to note here is 14:27:37 #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/087837.html 14:27:41 yep 14:27:54 ^ Armando posted a reminder on how to handle DocImpact tags 14:28:10 basically, we don't want to have the tag in each and every patch for a feature 14:28:18 a single commit with the tag is enough 14:28:30 And we shouldn't add the tag into spec 14:28:40 and also the tag should give brief clue which changes for docs are expected 14:28:46 you don't need to add DocImpact flag for both server and client changes. 14:29:04 hichihara: yes, adding it to specs patch is too early, since code may not land 14:30:01 as summary we don't need docimpact for -specs and (in most cases) -client 14:30:11 #topic i18n liaisonship 14:30:42 one more thing to note is there was an email from i18n folks that they need a i18n liaison from all projects, including neutron 14:30:47 #link http://lists.openstack.org/pipermail/openstack-dev/2016-February/087757.html 14:30:54 we already have amotoki serving for the role for several cycles 14:31:25 while Akihiro is too kind to serve for more, neither he will be against someone stepping in and freeing him from the obligation 14:31:28 right amotoki ? 14:32:00 i am working as i18n team as well, so it is not an additional job... but if someone volunteers it, feel free to go! 14:32:42 great. so if someone is willing to share some burden there, please reach to amotoki or armax or me 14:32:51 perhaps it would be small burden. 14:33:34 (just an idea) could be a good start for someone not yet too involved into the team business 14:33:38 ok, moving on 14:33:39 #topic Open Discussion 14:33:57 amotoki: wanna reiterate on the client release? or we are good? 14:34:17 i don't want to discuss them again :) 14:34:25 ok great :) 14:34:38 anyone has anything to raise in this tight pre-release time? 14:35:37 ok, I guess the answer is no and we can get back to reviews and coding and drinking coffee 14:35:52 thanks folks for attending, next time it will finally be US time, Monday 14:36:00 thanks all. 14:36:01 keep up the good work ;) 14:36:04 #endmeeting