14:00:08 #startmeeting networking 14:00:09 Meeting started Tue Oct 6 14:00:08 2015 UTC and is due to finish in 60 minutes. The chair is armax. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:10 hi 14:00:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:13 The meeting name has been set to 'networking' 14:00:14 mestery: WAT? but whatever 14:00:15 hi 14:00:19 hi 14:00:19 o/ 14:00:25 o/ 14:00:31 o/ 14:00:37 hi 14:00:42 * dougwig yawns 14:00:57 welcome everybody who made it 14:00:58 hi 14:01:05 hello 14:01:05 Hi! 14:01:10 #link https://wiki.openstack.org/wiki/Network/Meetings Agenda 14:01:18 sleepy but .. :-) 14:01:20 hi 14:01:30 I see lots of people scratching their eyes :) 14:01:39 #topic Announcements 14:01:59 We had an RC2 churning the last couple of days 14:02:14 #link https://launchpad.net/neutron/+milestone/liberty-rc2 14:02:19 o/ (sorry to be late ;) 14:02:20 mestery: what’s the latest? 14:02:24 ajo_: welcome 14:02:31 we good? 14:02:34 armax: We have one patch in the merge queue then RC2 will be cut 14:02:35 yeah, thanks :) 14:02:36 hi 14:02:39 #link https://review.openstack.org/#/c/231362/ 14:02:50 #link https://review.openstack.org/#/q/project:openstack/neutron+branch:stable/liberty,n,z 14:03:03 mestery: ok, let’s hope there’s no last minute scare 14:03:08 we’re really running out of time 14:03:11 Yes 14:03:13 Time is gone now 14:03:14 mestery: actually https://review.openstack.org/#/q/status:open+project:%255Eopenstack/neutron.*+branch:stable/liberty,n,z is better 14:03:20 HenryG: we good as far as DB goes? 14:03:21 ihrachys: Ack :) 14:03:29 armax: yup 14:03:33 ok 14:03:39 ok, that patch, is good to have, but if not merged, we can go 14:03:43 #link https://wiki.openstack.org/wiki/ReleaseNotes/Liberty#OpenStack_Liberty_Release_Notes 14:03:52 I mean, that bug is not actually happening in liberty, but in earlier releases, and off-gate 14:03:54 please can we all check that we haven’t missed anything? 14:04:08 but it's a fix we'd want backported to kilo 14:04:23 ajo: cool, thanks for clearing that up 14:04:25 so, if it doesn't merge, IMHO , feel free to cut, we're safe 14:04:39 231362 has no cherry-picked line. it's a nit though... 14:04:56 * armax looks 14:04:58 oh, true amotoki , I used the gerrit interface, I didn't realize 14:05:03 "Neutron now supports a way that an agent can start without selectable for auto-scheduling but manual-scheduling available so that a deployer can test an agent manually" am I the only one who fails to parse? 14:05:07 I guess the change-id keeps the traceability 14:05:12 um, yeah 14:05:35 yeah. I don't think we need to pull it from the queue. 14:05:36 I don’t see it in the gate queue 14:05:43 amotoki: actually, cherry-picked from line had some tech meaning (there was a bug in gerrit where it merged a patch into wrong branch if it was not present) 14:05:44 at all 14:06:12 but ttx approved it 14:06:15 ihrachys: that's scary 14:06:24 armax, ok, let's edit the commit message and get's the votes fixed quick, that's likely to retrigger the gate 14:06:26 not sure if it was a simple oversight 14:06:31 kevinbenton: it is. now you understand why I pick nits on backports? 14:06:32 ok 14:06:51 * ajo does 14:06:53 Wait, we're pulling it out of the gate to add a cherry-pick line? 14:07:03 Why? 14:07:04 ajo: the queue is 20+ patches deep 14:07:14 * ajo waits 14:07:14 ihrachys: Do you have a link to some info on that bug. I’ve never heard of it. 14:07:17 Please don't do that, ttx and I already approved it and it's the last one 14:07:26 ok 14:07:27 and hte check queue is 60+ patches deep 14:07:30 Right 14:07:31 mestery: ok 14:07:32 * ajo stops 14:07:34 Cool 14:07:35 :) 14:07:37 but I don’t see the change in the queue at all 14:07:37 Whew :) 14:07:47 It's in the merge queue armax 14:07:49 something is up! 14:07:54 ':) 14:07:55 231362? 14:08:15 oh found it 14:08:19 armax: :) 14:08:19 It's building right now 14:08:25 ok, wer’e good 14:08:27 carl_baldwin: I suggest people to read the guidelines: https://wiki.openstack.org/wiki/StableBranch#Proposing_Fixes 14:08:27 see in bold 14:08:28 let’s move on 14:08:28 Yes 14:08:42 yes…let’s accept it as a momentary lapse of the reason 14:09:01 ++ 14:09:01 #info People should remember to do cherry-pick -x on their cherry picks 14:09:08 Sanity prevails! Yay! :) 14:09:09 ajo: capisce? 14:09:18 ihrachys: it's sad I know it, but I trusted gerrit on doing the right thing on the "cherry-pick" button ;( 14:09:24 You can't even trust a machine.. 14:09:29 (it seems) 14:09:35 ajo: it fails to do it until the patch is not merged. 14:09:45 I always just made up the hash after the cherry picked from statement. I didn't know it meant something :) 14:09:45 ihrachys: yikes, understood 14:10:08 kevinbenton: made up like in uuidgen ? 14:10:09 X) 14:10:14 sorry, that's not an uuid... ;D 14:10:23 ok, next 14:10:29 before we do that 14:10:31 mestery: anything else? 14:10:38 anyone else on RC2? 14:10:41 comments? 14:10:52 Well 14:10:52 There is also a stable/kilo release very soon, right? 14:11:00 johnsom__ has indicated there are 3 LBaaS bugs which woudl be good to get in RC2 14:11:04 So I'm doing cherry-picks now 14:11:07 kevinbenton: yes, week+ from now 14:11:08 And I'll work offline with him on those 14:11:16 I am chatting with mestery about a couple of neutron-lbaas issues that fell through the cracks 14:11:31 +1 14:11:37 ok, for kilo, please: http://docs.openstack.org/developer/stevedore/patterns_loading.html#hooks-single-name-many-entry-points 14:11:39 argh.. 14:11:41 sorry :( 14:11:46 https://review.openstack.org/#/c/231363/ 14:11:49 that's what I meant 14:12:01 same ipset fix, for kilo, I will fix the cherry-pick reference 14:13:34 ajo: I'll handle the kilo queue after the meeting 14:13:40 thanks ihrachys 14:13:59 ok cool, I think we’re set then 14:14:13 if not, let’s quickly reconvene on our channel after the call 14:14:29 * armax wants to go to bed again after this ;) 14:14:39 armax: same here! 14:14:50 so, next reminder: 14:14:53 thanks armax & emagana 14:15:18 #link http://docs.openstack.org/developer/neutron/policies/index.html 14:15:27 we review/revise them as they go 14:15:37 we need to see what works and what doesn’t 14:15:56 and make sure we are a well oiled machine as a much as we can be 14:16:17 armax: ++ 14:16:21 I proposed changes to the way we handle bugs last week 14:16:23 #link http://docs.openstack.org/developer/neutron/policies/bugs.html 14:16:25 +1 14:16:38 armax: I have an announcement 14:16:40 +1 :) 14:16:46 the reaction was overwhelmingly positive 14:16:50 Sukhdev: wait your turn please 14:17:18 sorry - still sleepy 14:17:27 please make yourself familiar with the new guidelines 14:17:39 and reach out to us for questions or propose edits yourself 14:17:49 after all, we all own the policies 14:18:03 Sukhdev: you’re up 14:18:30 We are having ML2 sprint today and tomorrow for anybody interested in it 14:18:35 here is the information - https://etherpad.openstack.org/p/Neutron_ML2_Mid-Cycle_Sprint 14:18:41 #link: https://etherpad.openstack.org/p/Neutron_ML2_Mid-Cycle_Sprint 14:18:53 you should call it pre-cycle :D 14:19:00 it will cover few critical issues for ML2 14:19:02 Sukhdev: The secret ML2 illuminati meeting? :) 14:19:04 pre-wash? 14:19:10 lol 14:19:13 * regXboi likes pre-wash 14:19:18 lol 14:19:25 mestery: no secret - it was announced on ML weeks ago 14:19:31 I hope those issues are washed at high temperature 14:19:35 pre-Mitaka 14:19:39 Sukhdev: I'm kidding, mostly 14:19:41 some stains won’t go away otherwise 14:19:44 * regXboi wonders if we should say pre-OH 14:20:12 whats pre-OH? 14:20:16 Sukhdev, thanks, sounds interesting, I will be watching 14:20:17 ok, last announcement/reminder 14:20:25 pre-alcohol 14:20:35 regXboi: Love it! 14:20:37 We started our collictive guide 14:20:39 #link http://docs.openstack.org/developer/neutron/devref/effective_neutron.html 14:21:04 if you want to make changes to it, please tag the change with the ‘effective’ topic 14:21:43 *collective 14:22:06 * regXboi definitely thinks we should be considered a "well c2h5oh-ed" group :) 14:22:44 * dougwig groans 14:22:53 * ihrachys shrugs 14:22:56 * kevinbenton thinks regXboi took long to respond because he was looking that up on the Internet 14:23:13 kevinbenton: lol 14:23:19 * regXboi says - no, he was double checking with the chemist in the house - his wife :) 14:23:27 lol 14:23:38 the outstanding patches to the guide are: 14:23:47 https://review.openstack.org/#/q/status:open+project:openstack/neutron+branch:master+topic:effective,n,z 14:23:55 armax, it's a nice devref guide, there are so many questions/mistakes which repeat over time 14:24:00 feel free to check them out and post comments 14:24:05 ajo: that’s the idea 14:24:14 it is also a good chance to remember openstack common practices :-) 14:24:15 the more useful content we add, the better it is 14:24:29 we should strive to add Neutron-specific content 14:24:39 after all we’re more hopeless than other openstack projects 14:24:44 as some people might think 14:24:52 I thought you told me that the less content I added, the better it would be? 14:25:01 kevinbenton: no I mean, overall 14:25:10 :) 14:25:14 not per bullet point 14:25:39 any other announcement? 14:25:45 armax: I guess a link to another page with broader discussion of a topic is applicable there. 14:25:57 armax: Perception lags reality, but only when people are lucid 14:26:07 sorry, not following 14:26:12 I haven’t had my morning coffee yet 14:26:17 Who is throwing the Tokyo party? amotoki?? 14:26:17 armax: You must not be lucid 14:26:18 :) 14:26:30 kevinbenton: Doesn't matter who's throwing, armax is buying right? 14:26:48 you both armax and mestery should be buying! 14:26:49 ihrachys: you mean this? 14:26:50 #link http://lists.openstack.org/pipermail/openstack-dev/2015-October/076190.html 14:26:52 I want sake! 14:26:53 armax: if it's for me, I mean, let's say I want to cover upgrade, but that's a huge topic, so I create a new page in devref. Should I link to it from effective? 14:27:04 ihrachys: oh, yes 14:27:19 ihrachys: I think that’s mentioned in the introduction to the guide 14:27:20 kevinbenton: I am looking for a place... 14:27:30 ihrachys: if you need to expand feel free to just provide the pointer 14:27:30 oh poor me. ok. 14:27:53 emagana: as I said, since we’re the one providing the service, we’re the one that should be bought drinks ;) 14:28:00 amotoki: punishment for not finding one is you host it in your home 14:28:05 * markmcclain sneaks in late from teh car dealership 14:28:05 :) 14:28:11 ZZelle: ping 14:28:17 armax, pong 14:28:28 #topic Bugs 14:28:38 ZZelle: hi 14:28:39 kevinbenton: my home is not so near to the summit :-( you can enjoy one hour train. 14:28:49 armax: actually that is a good idea... who is working for an operator? :-0 14:28:51 party on the train? 14:29:11 with the changes in bug management procedures 14:29:22 we started looking at the backlog of bugs we have pending and that’s... 14:29:23 huge 14:29:37 armax: ++ for each bug :( 14:29:44 between client and server we have 1000+ open bugs 14:29:53 some may as well be junk 14:29:54 ouch 14:30:05 but we don’t know until we have a good handle on these 14:30:09 oi 14:30:22 so we gotta figure out a plan to bring those down 14:30:49 ZZelle has been looking at expiring some of those 14:30:57 armax: take 100, distribute 10 per core, expect them to be triaged till next week? 14:31:02 I think bug fixes would be good 14:31:07 ihrachys, that could sound good 14:31:08 We have a lot of bugs, a lot of bugs with no activity in the last 12 months: ~ 120 bugs 14:31:15 bugs that are marked incomplete, unassigned and untargeted expire after 60 days 14:31:30 I'd vote for getting those 120 into that state 14:31:40 that should give us enough time to look at them and revive those that might be relevant 14:31:45 for those that aren’t 14:31:56 the LP janitor will eventually evict them 14:32:09 so if everyone is okay with that strategy 14:32:13 ZZelle: fire away! 14:32:13 if we've been that lax on bugs, we might even want to scrub the recently expired (<6 months) 14:32:38 armax: no triage, just move them to the state? 14:32:48 ihrachys: LP has a link 14:32:49 dougwig, expired go when incomplete & not assigned & not targetted if I din't get it wrong 14:33:01 I guess somebody scrub those to become expired 14:33:06 scrubbed 14:33:15 #link https://bugs.launchpad.net/neutron/+expirable-bugs 14:33:27 ihrachys: after ZZelle fired his script 14:33:36 those bugs should appear under that link 14:33:41 but should we triage bugs before marking them for expiration? 14:33:46 ajo: if at Tzero a bug is moved to incomplete/unassigned/untargeted, then at Tzero+60 days, it will be expired by the LP janitor 14:33:48 and then we can go over and see which one is worth triaging 14:33:56 and reviving if we have too 14:34:16 dougwig, 6 months is really aggressive as it implies > 300 bugs in https://bugs.launchpad.net/neutron/+expirable-bugs to triage in 60 days 14:34:16 ihrachys: they won’t actually expire until the Janitor kicks in 60 days from now 14:34:27 ok, as long as we don't blindly close eyes and forget them at all 14:34:32 ihrachys: nope 14:34:54 dougwig: yeah we figured 1y was the best compromise 14:34:57 ok, that sounds good, 14:35:06 1.5y not aggressive enough and .5y too aggressive 14:35:18 if the reporter is not willing to provide more information after 60 days... I guess it's not that important 14:35:24 if 120 bugs seem too large to handle in 60 days, then we can run the script twice 14:35:50 ZZelle: that should be fine 14:35:52 by scrub, i meant, maybe we should take a look at recently expired bugs, lest we let things slip through the cracks. 14:35:55 ZZelle, you don't need to fix them ,just move out from incomplete 14:36:40 #action ZZelle to mark incomplete the bugs older than one year 14:36:46 ajo, i hope so :) 14:36:51 armax, ack 14:36:59 #info they will show up here: https://bugs.launchpad.net/neutron/+expirable-bugs 14:37:18 then I and someone else will look into those to make sure we don’t miss anything ‘important' 14:37:52 this actually brings me to the next point 14:38:04 bug deputy for the forthcoming week 14:38:44 I’ll be shadowing/helping where I can to ensure that we iron out the kinks of the new process 14:38:49 any volunteer? 14:38:50 I can volunteer 14:38:58 armax: I raise my hand as well 14:39:05 happy to take it too 14:39:10 can I go ahead and close from those bugs, what's actually incorrect? 14:39:19 or invalid? 14:39:24 ajo: yes you can 14:39:30 armax: too much volunteers. we need a lottery! 14:39:37 ihrachys: went first 14:39:39 rofl 14:39:41 I'm happy to wait until november :) 14:39:41 sigh 14:39:42 :) 14:39:42 :) 14:39:44 heck - you've got the next three weeks 14:39:48 the first one who replies wins 14:39:50 just hand it out one at a time 14:39:52 :) 14:40:05 I will volunteer after (or even during, if that's useful) summit 14:40:14 Let’s have ihrachys be the deputy for the week of Oct 5 14:40:14 Can you unassign the bugs that haven't had activity for a week? 14:40:20 armax: actually, I really thing you *should* schedule it out a bit further 14:40:20 ihrachys++ 14:40:24 With no patch proposed 14:40:29 but anyone else should be free to help and take stuff off his plate 14:40:42 armax: we select every week? 14:40:43 should we just add our names to the neutron meetings wiki? 14:40:43 if they feel like he’s slipping ;) 14:40:52 ihrachys: either every week or every other week 14:40:54 it might help folks to know a little in advance what week they are "in the slot" 14:40:58 #link https://wiki.openstack.org/wiki/Network/Meetings 14:40:59 ihrachys: depending on how you’re doing 14:41:03 ack 14:41:12 * ihrachys is voted to be the guinea pig 14:41:12 armax, ZZelle , marking as Invalid is enough, or should touch anything else? 14:41:18 ihrachys: you feel like you have bandwidth this week? 14:41:22 armax: If you schedule it out further than anyone who is driven by a sprint schedule can make sure that they have time allocated for bugmaster duty in their sprint. 14:41:23 we need some token of office, like a sceptre. a wand with a rubber cockroach on the end. 14:41:30 armax: yes, I will do bug scrubbing for OSP anyway 14:41:40 ihrachys: ok great 14:41:40 Nice 14:41:50 dougwig: I think you have shopping to do between now and tokyo 14:41:54 I’ll add you to the bug deputy table for the week 14:42:16 armax: see njohnston's comment - he makes a good point 14:42:24 #info ihrachys the bug deputy for Oct 5th, supported by regXboi, markmcclain if necessary 14:42:45 * regXboi acks the #info 14:42:55 ihrachys: I can help as well 14:43:07 folks, thanks and I will reach if needed 14:43:08 njohnston: yes, assumed that their sprint schedule overlaps with ours? 14:43:56 week by week seems the best horizon for things upstream, but we can definitely plan a little further ahead 14:44:05 so for instance, if regXboi wants to take next week 14:44:19 regXboi: do you want to? 14:44:27 we can revise your commitment next week 14:44:30 armax: I can take the next slot 14:44:35 njohnston: is that what you were suggesting? 14:44:54 #info regXboi to provisionally take the bug deputy role for week of Oct 12th 14:45:30 ok, let’s move next 14:45:32 10 mins to choose a volunteers. sick. 14:45:40 lol 14:45:53 ihrachys: too many volunteers! you gotta love it 14:46:08 #topic docs 14:46:11 emagana: any update? 14:46:20 armax: short one 14:46:43 On the networking guide we will be stopping having the old hangout meetings and moving into IRC 14:47:26 emagana: yay! Good call! 14:47:54 Nick Chase was helping on leading the logistics but he will take a break and I will be leading the meeting.. I will send the details on ML because I dont have them yet! 14:47:56 :-) 14:48:15 thank you emagana! 14:48:26 So, folks please join and contribute to the networking guide! 14:48:50 thanks emagana 14:49:03 anything else on docs? 14:49:10 armax: passing the ball to you! 14:49:13 emagana: do we have our docs process documented in devref? that could be a good start for more participants. 14:49:14 bear in mind that whenever DocImpact is flagged on a commit message 14:49:34 a new bug is filed for the openstack manuals 14:50:02 please remember to provide as much info 14:50:15 to help the doc bug submitter with his/her job 14:50:44 #topic Open Discussion 14:50:54 as some of you may know... 14:50:55 #https://etherpad.openstack.org/p/neutron-mitaka-designsummit 14:50:59 the summit is fast approaching 14:51:10 we wanted to discuss https://blueprints.launchpad.net/neutron/+spec/cms-to-test-environment 14:51:15 #link https://bugs.launchpad.net/neutron/+bug/1485422 14:51:15 Launchpad bug 1485422 in neutron "neutron-ovs-cleanup sometimes failed at reboot (Ubuntu 14.04)" [Undecided,Invalid] 14:51:18 I have been working with the drivers team to put together a provisional schedule for the neutorn sessions 14:51:19 #undo 14:51:20 sorry :( 14:51:34 was trying to link it right, my copy & paste does *not* work :( 14:51:50 #link https://etherpad.openstack.org/p/neutron-mitaka-designsummit 14:51:55 * regXboi thinks only chair can #undo? 14:52:00 #undo 14:52:01 Removing item from minutes: 14:52:07 one more armax 14:52:08 undo twice, sorry 14:52:08 #undo 14:52:09 regXboi is correct. 14:52:10 Removing item from minutes: 14:52:16 we good? 14:52:18 #link https://etherpad.openstack.org/p/neutron-mitaka-designsummit 14:52:21 now yes :) 14:52:21 now we are good 14:52:49 expect an email soon 14:53:02 about what the schedule is going to be about and how we are breaking down the various sessions 14:53:26 we’ll have a lightining talk session too 14:53:46 and don’t forget that Friday is full day for community get together 14:53:52 so if you haven’t booked your flights yet 14:53:53 will kevinbenton produce another talk about bad commits? 14:53:55 consider staying late 14:53:57 or maybe BAD REVIEWS! 14:54:03 ihrachys: no, we have the effective guide now ;) 14:54:14 armax: no fun 14:54:18 i think ihrachys is implying that it's time to roast the reviewers. 14:54:25 kevinbenton: unless he wants to, but he doesn’t want to chair any session 14:54:32 dougwig: it's always that time 14:54:35 anyhoo 14:54:44 ihrachys: -1, sentence missing period. 14:55:04 -1 this sentence shouldn't have a period 14:55:10 * regXboi wonders who let the punctuation police into the meeting 14:55:20 yamini, ajo: please consider adding the topic on the on demanda agenda on 14:55:28 https://wiki.openstack.org/wiki/Network/Meetings 14:55:32 * dougwig thinks subtle humor is lost at this hour of the morning. 14:55:38 Monday will be some FWaaS fun 14:55:39 armax, which topic? 14:55:42 sigh. I think I have a flight back on Fri 14:55:56 it it is 14:56:01 it is already there 14:56:10 is it? 14:56:17 I must have wiped it out 14:56:20 * regXboi doesn't see it 14:56:44 blueprints are best discussed in gerring under neutron-specs 14:56:50 btw 14:56:55 or the ML 14:57:05 sc68cal: +1 14:57:41 but yeah, we gotta figure out a way to leave more time for the open discussion section otherwise 14:57:44 are we done? 14:57:46 yamini: we are using RFE bugs rather than filing blueprint http://docs.openstack.org/developer/neutron/policies/blueprints.html i suggest you file a RFE bug so that we can be aware of it. 14:57:54