21:02:11 #startmeeting networking 21:02:12 Meeting started Mon Jun 30 21:02:11 2014 UTC and is due to finish in 60 minutes. The chair is mestery. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:13 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:15 The meeting name has been set to 'networking' 21:02:23 #link https://wiki.openstack.org/wiki/Network/Meetings Agenda 21:02:33 #topic Announcements 21:02:42 The mid-cycle sprint is next week: https://etherpad.openstack.org/p/neutron-juno-mid-cycle-meeting 21:02:52 We'll cover this in detail in a few minutes, but wanted to note this here. 21:03:01 Any other announcements the team should be aware of? 21:03:27 #topic Mid-Cycle Sprint 21:03:39 I wanted to ensure we spent some time on this early in this meeting. 21:03:53 #link https://etherpad.openstack.org/p/lOiqsLmizl Sprint Work Items 21:04:01 I've started that etherpad to collect work items for next week. 21:04:41 I'm hoping those who arrive in beautiful Minnesota next week can all hit the ground running Wednesday morning. 21:04:42 nice 21:04:53 3 intense days of closing nova-network gaps. :) 21:05:06 mestery: +1 21:05:25 We'll go from 9AM-6PM or so, though I will likely be in the office a bit before that if people want to come by earlier. 21:05:42 tough boss! ;) 21:05:43 I hope none of the nationalities there in Minnesota will still be represented at the world cup 21:05:45 One thing to note: If you are planning on attending, please send unicast me your [name, email, company] so I can get guest wifi setup for you 21:05:51 sc68cal: :P 21:06:05 salv-orlando: Ha! 21:06:08 So go Belgium! 21:06:21 salv-orlando: Mexico is out :-( 21:06:22 salv-orlando: we were kicked out by the Dutch yesterday, so i'll be ok 21:06:46 Any other logistical questions I can answer from anyone about next week? 21:06:58 very unfortunate for Mexico! 21:07:17 mestery: *how* early will you be in the office? 21:07:28 regXboi: How early is early for you? Now I'm afraid. :) 21:07:34 * markmcclain notes that neither the England or Italy are left for salv-orlando 21:07:42 mestery: I suppose the unicast of info will get us also in the building? 21:07:45 mestery: I'll probably be able to wander in about 7 am 21:07:54 I never cared for England. Italy is instead on a sabbatical 21:08:08 Italy literally got eaten by Uruguay 21:08:16 so yes, you should be afraid :) 21:08:18 banix: That's a good point. You can get into the building past 8AM, take the elevator to the 4th floor and checkin with the lobby ambassador 21:08:20 salv-orlando: oh that explains their poor performance 21:08:36 regXboi: I don't know if I'll be there by 7AM. 8AM likely, though :) 21:08:38 mestery: thanks 21:08:54 mestery: no worries ... 21:09:06 regXboi: I'll see what I can do :) 21:09:23 I'll likely come in early, I'm an early riser. 21:09:48 Looking forward to working F2F with folks next week to close nova-network parity gaps! 21:09:50 mestery: the session kicks off at 9am though, right? 21:09:58 Sukhdev: Correct. :) 21:10:10 * mestery notes Sukhdev is from PDT timezone :P 21:10:20 mestery: I will be on time :-) 21:10:21 Sukhdev: I see you are worried. You should be when 11m PST is early for you :) 21:10:31 mestery: same here… am excited to close these gaps 21:10:47 banix: :-) 21:10:47 OK, unless there's anything else, we'll proceed with the rest of hte meeting. 21:11:13 #topic Bugs 21:11:20 hi 21:11:20 enikanorov__: Hi! 21:11:38 nothing much new this wee. i'm trying to track gate failures 21:11:47 Anything new on that front (gate failures)? 21:11:57 so far we have a list of rarely hit bugs related to lock wait timeouts 21:12:27 i hope most of the cases should be fixed by the patch 21:12:36 let me find the link 21:12:41 Thanks enikanorov__ 21:12:45 enikanorov_: I think you’seen this already by anyway,… https://review.openstack.org/#/c/103540 21:12:57 that’s ihrachyshka take on the lock wait timeout problem 21:13:10 salv-orlando: yes, i've seen this, it seems as an overkill to me... 21:13:35 Elena Ezhova and me were analyzing different cases of timeouts 21:13:36 ok - I just wanted to raise awareness. comments are better placed on gerrit. 21:13:48 so far we've seen that they are still connected to port deletion 21:14:13 so the patch ihrachyshks has on review should fix most of them 21:14:15 only thing I want to avoid is to have 10 different solutions for the same problem on gerrit and none on git log ;) 21:14:44 #link https://review.openstack.org/#/c/100934/ 21:14:59 salv-orlando: sure 21:15:03 another question i have on this 21:15:08 is the e-r queries 21:15:36 i've pushed a few to gerrit, but infra folks are not happy to merge queries that give 2-4 hits per week 21:15:46 but when summed, those timeouts are several a day 21:15:51 what would you recommend? 21:16:30 enikanorov__: I wonder if they don't want to explode e-r queries for bugs with such small hits. 21:17:17 so e-r queries are weighted for severity on gate hits 21:17:21 enikanorov__: I would say lets bring this up in #openstack-infra and see what we can do. 21:17:22 then it would not be convenient to track hits of particular cases 21:17:31 and those cases are ~10 different ones 21:17:49 mestery: just like with beetlejuice you can try and say Joe Gordon’s name 3 times 21:17:53 he might appear here 21:18:00 salv-orlando: :P 21:18:04 ok. the problem is that i don't know the policies about e-r queries, so i don't quite understand their reasoning 21:18:04 * markmcclain is starting descent…will stay rejoin when on ground 21:18:22 markmcclain: Enjoy SLC for the short while you're there ;) 21:18:33 enikanorov_: usually they’re filed to bring down uncategorized failures to 0 21:18:55 so that’s why infra focuses on failures hitting the gate 21:19:20 anopther concern was that some failures happening on non-voting neutron-full job 21:19:26 but in the past I brought up the argument in openstack-infra or in gerrit and had them merged 21:19:44 enikanorov_: yeah lock wait timeout failures are more frequent there. 21:19:51 salv-orlando enikanorov__: Is the full job not voting yet? The plan is to turn that on and make it voting soon right? 21:20:10 mestery: yes - there is an update on that on the mailing list 21:20:34 anyway, i'll followup with infra folks on their channel 21:20:36 salv-orlando: Thanks! 21:21:22 Anything else bugs related enikanorov__? 21:21:30 no 21:21:36 OK, thanks enikanorov__! 21:21:39 #topic Docs 21:21:41 emagana: Howdy! 21:21:49 mestery: Hi There! 21:22:08 Wiki is updated with the High* priority bugs 21:22:18 (only one open) :-) 21:22:38 I have seen developers adding DocImpact.. and I would like to thank all for that!! 21:23:27 Just a nice extra reminder! If you push a commit with "DocImpact" please, take a look to the Doc Bug and try to fix it as well :P 21:23:54 emagana: One of the gap areas is around documentation of hte open source options, I'm hopeful you can assist those who want to narrow that gap next week in MN! 21:24:07 emagana: You may need to boot strap some new docs contributors :) 21:24:14 mestery: I will do that! 21:24:23 emagana: Awesome, thanks! 21:24:24 * regXboi marvels at mestery's deft delegation 21:24:29 regXboi: :P 21:24:36 emagana: Anything else this week? 21:24:42 mestery: good to know, I will prepare some overview of the Docs for newers! 21:24:54 emagana: That would be even better! /cc regXboi 21:25:07 emagana: :) I'm looking for that info 21:25:12 mestery: Yes, I dont like fake penalties!!!!!! 21:25:18 because I keep banging my head on doc issues 21:25:37 emagana: Maybe we can even look at documentation options for DVR next week since carl_baldwin will be there too. But maybe it's too early, lets see. 21:25:58 mestery: not a bad idea 21:26:03 mestery: the sooner the better... I will discuss that with carl_baldwin 21:26:08 * mestery is bringing everyone together. :) 21:26:27 Thanks for the updates emagana! 21:26:39 #topic Tempest 21:26:42 mlavalle: Hi there! 21:26:49 hi: we are pretty much complete with the api coverage that we planned early this year. I have one test script for VPNaaS and zzelle_ for the provider network extension 21:26:51 no 21:26:59 oops... sorry - wrong window 21:27:06 mlavalle: Nice! 21:27:33 I will update mark_mccclain's etherpad for nova pairty with all the tests pmerged 21:28:39 mlavalle: For next week, we may have some new tempest contributors, perhaps Wednesday you could spend some time boot strapping a few. 21:28:41 I want to devote the next few minutes, though, to gather some feedback for the sprint next week, especially from markmcclain and salv-orlando: what do you see is needed next week for nova-parity and the full neutron job? 21:28:56 in terms of tempest^^^ 21:29:06 mlavalle: I think for enabling the full neutron job we don’t need anything in tempest. 21:29:12 mestery: yes, I will do that 21:29:24 The issues have been identified, bugs filed and root caused. 21:29:29 Patches wil follow up shortly. 21:29:49 So, I guess I will do my best to get the full job switched to voting this week 21:30:02 so that you guys next week can deal with inevitable spike in failure rate 21:30:04 salv-orlando: any tests that should be redesigned? I think you mentioned something in the tempest irc meeting to that effect 21:30:07 salv-orlando: Nice! 21:30:09 salv-orlando: :( 21:30:37 mlavalle: the tempest design issue pertained the lbaas scenario test. Elena Ehzova already addressed that. 21:30:53 salv-orlando: cool 21:31:00 mestery: problems? 21:31:16 how about markmcclain? 21:31:19 salv-orlando: No, just #sadpanda for the inevitable fallout 21:31:23 ah ok! 21:31:25 mlavalle: markmcclain is landing right now :) 21:31:31 mlavalle: I'll circle with him later :) 21:31:41 mestery: ok, i'll follow up directly with him 21:31:47 markmcclain loves neutron meetings at 35k feet 21:31:47 mlavalle: Thanks! 21:32:03 any other input from the team as far as stuff that needs to be covered next week? 21:32:07 salv-orlando: I don't think he takes neutron meetings from anyewher else :) 21:33:06 if there is not more input, I am done. I will work the etherpad over this week 21:33:11 Thanks mlavalle! 21:33:23 #topic L3 21:33:26 carl_baldwin: Hi there! 21:33:28 hi 21:33:39 Frist, DVR. We’ve seen a lot of great progress. 21:34:03 We had a big push to clean up the test failures. We were 4 UT failures away this morning on only one patch. 21:34:17 We’ve been addressing feedback about as quickly as it can come in. 21:34:35 armax posted a demo video that was interesting. 21:34:46 carl_baldwin: Sweet! 21:34:47 Link? 21:34:57 awesome 21:34:58 * carl_baldwin is looking. armax? 21:35:14 hy 21:35:15 hi 21:35:35 I put the link to the video here: https://wiki.openstack.org/wiki/Neutron/DVR/HowTo 21:35:47 #link https://wiki.openstack.org/wiki/Neutron/DVR/HowTo 21:35:51 Thanks armax! 21:35:53 Not sure why, but it looks like Chrome does not render the video tag very well 21:36:12 armax: mestery: that must be why I couldn’t see it. 21:36:19 Anyway, we’ll get that taken care of. 21:36:30 I was going to send a mail to ML later to broaden the announcement 21:36:43 looks fine in Safari 21:36:47 armax: please do 21:36:54 armax: ml post is still good. I will forget about that link in 5 minutes or less 21:36:54 I’ve also put up a temporary branch that merges all of the gerrit patch sets in to one commit a few times a day. 21:37:06 #link https://www.youtube.com/watch?v=p4BwAjLHd0M 21:37:10 #link https://github.com/ecbaldwin/neutron-dvr.git 21:37:10 "The page at 'https://wiki.openstack.org/wiki/Neutron/DVR/HowTo' was loaded over HTTPS, but ran insecure content from 'http://www.youtube.com/embed/p4BwAjLHd0M?showsearch=0&modestbranding=1': this content should also be loaded over HTTPS." 21:37:21 ^ This is just if you want a quick way to get all of the code together. 21:37:47 carl_baldwin: Thanks for that, for people who want to test it out, that will be helpful! 21:37:59 carl_baldwin: Is testing something which people can help with next week around DVR? 21:38:09 Yes, it has helped me out as well. I thought I’d share. 21:38:54 mestery: I think so. I’m working on it myself with a few others. If others are interested in testing please let me know. 21:39:10 mestery: ideally a soon as we merge that we might add a non-voting job that runs neutron with DVR 21:39:16 carl_baldwin: If there are things people can assist the DVR team with next week, we can add them to the etherpad. 21:39:33 salv-orlando: +1 to that, though that would require multi-node tempest testing to be useful, right? 21:39:36 mestery: I’ll have a look. 21:39:41 carl_baldwin: Thanks! 21:39:41 carl_baldwin: i'm interested in testing 21:39:56 salv-orlando: Yes, we’re working on that. 21:40:00 mestery: yeah but even without we’ll add coverage for new paths 21:40:01 enikanorov__: noted. Thanks. 21:40:24 salv-orlando: True, dat. 21:40:52 The rest is up to date on the meeting page. There were a couple of new High bugs but we’re on them. 21:41:05 carl_baldwin: Awesome, thanks! 21:41:14 yw 21:41:17 That’s it. 21:41:32 #topic Advanced Services 21:41:37 SumitNaiksatam: Hi there! 21:41:42 mestery: hi 21:41:57 we had quite an interesting discussion in a special session on flavors on Friday :-) 21:42:18 SumitNaiksatam: Indeed ;) 21:42:29 we thought we were beginning to reach a conclusion at least on the user facing API, but then we ran out of time 21:42:53 so for now, the plan is that markmcclain and enikanorov_ or going to reconcile on the items that we agreed upong 21:42:56 *upon 21:43:07 right 21:43:27 i've left corresponding comment on markmcclain's spec 21:43:28 we might decide to leave features like the “tags” defined in enikanorov_’s proposal as a future enhancement 21:43:31 I think we're close, though I'm hopeful we can close this before the end of hte week. 21:43:33 enikanorov_: great, thanks 21:43:43 mestery: +100! 21:43:53 I'll reply to those on next flight 21:43:58 Flavors is important for all the services, so closing on this will be good. 21:44:04 markmcclain: great! :-) 21:44:05 markmcclain: Awesome! (And welcome back :) 21:44:18 markmcclain: is a pilot, isn't he? 21:44:28 enikanorov__: :) 21:44:36 other than that, we want to focus on the service insertion and chaining blueprint specs 21:44:51 to the extent that we can without distracting people from their focus 21:45:07 so expect this wednesday to have more of a discussion on the service insertion spec 21:45:19 SumitNaiksatam: :) 21:45:25 regXboi: :-) 21:45:41 mestery: thats the short summary 21:45:54 SumitNaiksatam: Thanks for the summary! 21:46:10 SumitNaiksatam: I plan on reading through and adding comments on the other GBP patchsets... just trying to find blocks of time 21:46:18 #topic IPv6 21:46:20 sc68cal: Howdy! 21:46:22 Hello! 21:46:28 regXboi: thats a separate agenda topic, but thanks :-) 21:46:38 So the two big ticket items are the IPv6 radvd spec and review 21:46:43 #link https://review.openstack.org/#/c/101306/ IPv6 radvd spec 21:46:43 #link https://review.openstack.org/102648 IPv6 radvd review 21:47:00 mestery: thanks a bunch for chairing the flavors meeting! 21:47:07 SumitNaiksatam: yw 21:47:21 We also got tempest API tests for ipv6 attributes merged https://review.openstack.org/#/c/93502 21:47:26 sc68cal: I saw those, that's great they are both there, given the movement to radvd 21:47:31 sc68cal: woot! 21:47:46 I also rebased and fixed conflicts on the python-neutronclient side for the ipv6 attributes 21:48:06 #link https://review.openstack.org/#/c/75871/ python-neutronclient v6 support 21:48:17 sc68cal: thanks! 21:48:18 So, eyeballs and reviews appreciate 21:48:20 sc68cal: You've been busy :) 21:48:42 also check out the multi-prefix review https://blueprints.launchpad.net/neutron/+spec/multiple-ipv6-prefixes 21:48:54 We'd appreciate some eyeballs on that, and feedback 21:49:07 That's it for me 21:49:27 sc68cal: Seems like IPv6 is making good headway here, which is good! 21:49:37 #topic ML2 21:49:40 Sukhdev: Hi! 21:49:47 mestery: Hi 21:49:54 Got few items to report 21:50:20 Team has been busy reviewing specs - some of them are ready to go and require core approvals 21:50:41 https://review.openstack.org/#/c/88101/ 21:50:45 https://review.openstack.org/#/c/91811/ 21:50:56 https://review.openstack.org/#/c/89728/ 21:51:10 Sukhdev: Thanks for pre-reviewing those and noting their readiness for core reviews! 21:51:19 mestery: thanks for approving them - need other cores to 21:51:30 review as well 21:51:41 I will help Sukhdev 21:51:50 emagana: tahnks 21:51:55 emagana: awesome! 21:52:13 Another thing to report is that some of the team members met Friday to discuss ML2 Sync 21:52:40 We are beginging to plan the activities around ML2 Sync to solve some of the issues 21:52:57 https://bugs.launchpad.net/neutron/+bug/1193861 This is short term fix 21:52:59 Launchpad bug 1193861 in neutron "ML2 plugin needs to override bulk operations" [Medium,In progress] 21:53:13 but, we are looking to solve it with ML2 - 21:53:35 will work on the specs and BPs with the team members - so look out for that 21:53:50 That is all I have to report for now 21:53:53 Sukhdev: perhaps i am reviewing all three above. will look later today but i have some concern on wording. 21:53:55 Sukhdev: Thanks for the updates! 21:53:57 unless there are questions 21:54:06 amotoki: thanks 21:54:17 amotoki: Awesome and thanks! 21:54:30 #topic Group Based Policy 21:54:33 SumitNaiksatam: Howdy again! 21:54:44 mestery: hi again 21:54:49 short update 21:55:00 we have lots of patches in review now, the model series is almost complete 21:55:21 and the driver series is almost there (rkukura would be posting a patch sometime later today) 21:55:39 and now I'll say what I said before.... I owe you reviews - just need to find a block of uninterrupted time (who knows, maybe friday morning) 21:55:40 with that one should be able to test a subset of the resources all the way to the data path 21:55:56 regXboi: thanks, and appreciate your comments earlier 21:56:06 thanks in general to all the other reviewers as well 21:56:13 mestery: so good progress over all 21:56:25 SumitNaiksatam: Excellent! 21:56:31 CLI patch is also ready 21:56:47 mestery: thanks for the time 21:56:52 SumitNaiksatam: thank you for the updates! 21:56:56 #topic Open Discussion 21:57:00 conflicting views on #link https://review.openstack.org/#/c/93866/ Small issue and patchset. Shouldn’t require too much of your time 21:57:35 https://review.openstack.org/#/c/93866/ 21:57:44 #link https://review.openstack.org/#/c/93866/ 21:57:48 banix: ^^^ 21:58:19 mestery: thx. yeah the first thing didn’t come out right for some reasons. 21:58:27 banix: No worries :) 21:58:30 OK, thanks everyone! 21:58:36 Need help on how to get oslo.messaging 1.4.0.0a2 in global rewuirements https://review.openstack.org/#/c/103536/ 21:58:38 it is a discussion on positional argument of internal class. we have similar discussion several times so far. 21:58:39 Look forward to seeing some of you next week in Minnesota! 21:58:41 is 3rd party ci expected to test and vote stable branches? 21:58:49 pcm__: See reply by markmc on the ML to that one. :) 21:58:56 ok 21:59:09 yamamoto_: If your driver is upstream in a stable branch, yes. I'll clarify the wiki page for that. 21:59:22 amotoki: I know implementations based on that class 21:59:24 ok, thank you 22:00:11 OK, thanks everyone! 22:00:17 Catch you all on the ML and IRC! 22:00:18 #endmeeting