21:02:03 #startmeeting networking 21:02:04 Meeting started Mon Jul 28 21:02:03 2014 UTC and is due to finish in 60 minutes. The chair is mestery. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:07 The meeting name has been set to 'networking' 21:02:37 #link https://wiki.openstack.org/wiki/Network/Meetings Agenda 21:02:37 Hi 21:02:43 aloha 21:02:47 #topic Announcements 21:02:52 Juno-2 has been released! 21:02:57 #link https://launchpad.net/neutron/juno/juno-2/+download/neutron-2014.2.b2.tar.gz 21:03:08 For those who want to play with it, please have a look! 21:03:17 We closed 11 BPs during Juno-2 21:03:19 #link https://launchpad.net/neutron/+milestone/juno-2 21:03:41 so is that only like 150 left for Juno-3 then? 21:03:41 So of course, this must mean that ... Juno-3 and the end of Juno is on the way! 21:03:45 #link https://launchpad.net/neutron/+milestone/juno-3 21:03:51 kevinbenton: Give or take a few. ;) 21:04:04 My advice to anyone with a Juno-3 BP: Please propose yoru code soon. 21:04:13 The sooner you get it out htere, the sooner people can review it and we can iterate. 21:04:27 I expect a massive crush at FPF, so sooner is better. 21:04:38 Any questions on Juno-3 timelines? 21:05:18 OK then, moving along in the agenda for today. 21:05:20 #topic Bugs 21:05:23 enikanorov__: Hi! 21:05:24 hi 21:05:27 hi 21:05:36 hi 21:05:50 hi 21:05:58 nothing major to update with. I'm monitoring gate failures 21:06:18 the amount of lock wait timeouts has decreased from 10-20 to ~1-2 per day 21:06:29 enikanorov__: Are those the three listed in the agenda at the moment? 21:06:37 .. and switching to mysqlconnecter is still blocked on oracle fwiw. 21:06:51 mestery: not really. i'm starting to decrease importance of corresponding bugs 21:07:07 because it's really ahrd to say which one was fixed as they don't show up in days 21:07:32 * mestery magically reappaers. 21:07:51 * aleksandr_null =) 21:08:06 OK, anything else in the bugs arena enikanorov__? 21:08:07 so, no critical bugs appeared during last 2 weeks 21:08:12 enikanorov__: Ack 21:08:18 mestery: what about bug swuash day? 21:08:22 *squash 21:08:31 enikanorov__: I was going to annoucne that during Open Discussion, but can do it now. 21:08:39 So, I'm going to send an email and propose a Neutron Bug Squashing day 21:08:45 Will likely be next week. 21:08:52 Stay tuned for the details as I work with enikanorov__ on those :) 21:09:01 ok, great! 21:09:15 I am a bit concerned about the number of new/untriaged bugs? 21:09:33 markmcclain: are you? 21:09:33 #link https://bugs.launchpad.net/neutron/ 21:09:36 markmcclain: Ack on that, I think perhaps a bug day can help there too. 21:09:59 yeah 1/6 haven't been looked at might be good to make that part of the focus on the day 21:10:08 enikanorov__: if it helps: http://princessleia.com/journal/?p=9558 21:10:16 anteaya: thanks 21:10:19 np 21:10:22 markmcclain: i'm generally try to keep them under 100, some of those I really don't know what to do with 21:11:03 #action mestery to setup Neutron Bug Squash Day and send email to mailing list with details 21:11:47 enikanorov__: maybe on that day we can have a group of folks help you finalize the disposition on that class of bugs 21:12:06 yeah, i hope so 21:12:52 OK, lets move on in the agenda, I'll start a ML thread for the bug squash day 21:12:55 #topic Nova Parity 21:12:57 markmcclain: Hi there! 21:13:07 hi 21:13:50 so at Nova mid-cycle this week to discuss how to get the migration unstuck 21:15:21 otherwise not much other movement to report this week 21:15:37 markmcclain: I look forward to yourself and marun making progress on the migration work, thanks for that! 21:15:57 Also thanks to obondarev for his work here too. Team is #ftw! 21:16:01 #topic Docs 21:16:04 emagana: hi there! 21:16:08 hi 21:16:27 I want to talk about the Juno deliverables for documentation 21:16:53 This release involves a lot of changes and we need to take care of document them properly 21:17:04 specially the Migration and DVR staff 21:17:35 ++ 21:17:44 We also need to review all current guides and make sure they will have the proper documentation.. So, I have this etherpad: https://etherpad.openstack.org/p/neutron-docs-juno 21:18:17 We need VOLUNTEERS to cover some of the guides, basically we need readers and bug reporters 21:18:25 #link https://etherpad.openstack.org/p/neutron-docs-juno 21:18:51 The idea is to have one owner per guide who will report any missing part for the networking changes in Juno.. 21:18:57 Similar to bug day, I wonder if a "docs" dya makes some sense. 21:19:24 mestery, possibly 21:19:31 mestery: probably a docs "week" would be more like it 21:19:38 mestery, the key for docs is the docimpact bug typically isnt enough information for a writer to pick up 21:19:43 mestery: We want to start reporting the potential issues/bugs. Once we are aware of what is needed to be changed, we can organize a docs day 21:19:45 sgordon: Ack on that one. 21:19:59 mestery, for it to be attractive for someone to pick up from the queue someone who knows neutron needs to flesh it out 21:20:02 We are working on a TOC for a Neutron Manual also 21:20:05 #action mestery to work with emagana on a docs day for Neutron in Juno 21:20:13 sgordon: that's correct. This is why we are proposing owners to verify the needed changes 21:20:14 sgordon: Ack on that 21:20:48 Neutron Folks! Propose your self to review some of the guides (it is a good learning experience) 21:21:09 You can email me emagana at gamail.com or just go ahead to the etherpad 21:21:33 Networking guide TOC: https://wiki.openstack.org/wiki/NetworkingGuide/TOC 21:21:41 sorry, domain is gmail (I did not create my own domain) 21:21:53 #link https://wiki.openstack.org/wiki/NetworkingGuide/TOC Networking Guide TOC 21:21:56 Thanks phil_h 21:22:01 yes, thanks phil_h! 21:22:36 Anything else emagana? 21:22:43 By now we have an idea of what will be in Juno regarding to networking, so please read and open BP/bugs 21:23:26 mestery: No nothing else, at least we do not have any volunteers.... I will ask you to nominate candidates :-) 21:23:26 phil_h: that guide, at first glance, looks great! 21:23:38 thanks 21:23:48 emagana: Lets work offline on that. :) 21:23:54 #topic Tempest 21:23:56 mlavalle: hi there! 21:23:59 hi 21:24:24 over the past few days I've been doing tempest testing of the new LBaaS 21:24:55 Last night I sent initial rersults to the ML http://lists.openstack.org/pipermail/openstack-dev/2014-July/041320.html 21:25:04 #link http://lists.openstack.org/pipermail/openstack-dev/2014-July/041320.html 21:25:17 mlavalle: I saw that, very nice and thanks for doing that! 21:25:30 +1 21:25:37 mlavalle: I spoke to Jorge Miramontes this morning and he finds it very useful 21:26:03 Also a few days ago I worked with SumitNaiksatam on developing an api test for policy group 21:26:16 I verified it in devstack. it works fine 21:26:19 mlavalle: thanks much! 21:26:42 mlavalle: it was a great effort over a weekend, and like you said we were able to validate 21:26:46 actually i had one more thing for the previous topic.... 21:26:51 release notes are open for editing: https://wiki.openstack.org/wiki/ReleaseNotes/Juno 21:26:59 thanks sgordon 21:27:01 mlavalle: I think I may have found a flawn in the way tempest tests are designed 21:27:08 and the way they are run 21:27:19 by default all api extensions are exercised 21:27:26 I think these 2 cases illustrate how developing tempest tests cases can help in the early development of a new project 21:27:41 however currently we only keep one tempest codebase to support both master and stable releases 21:27:58 if we add new extensions, gbp and dvr, we won’t be able to address icehouse correctly 21:28:05 as in tests would break 21:28:25 in this patch: https://review.openstack.org/#/c/98235/ 21:28:38 armax: that's where capability flags come in 21:28:39 I dynamically discover the api supported to bypass the problem 21:28:58 marun: right, but it looks like that does not change across realeses 21:29:01 armax: tempest configuration that determine what tests should be run 21:29:19 armax: and since api's should not change, it should be possible to have boolean flags and different configurations across releases 21:29:24 marun: besides keeping different localrc templaces might be a pain 21:29:36 armax: that's pain that can't be avoided 21:29:47 armax: there is always going to be the potential of different config across releases 21:29:50 marun: at the moment the api_extensions flag for icehouse is ‘all' 21:29:56 marun: just as the one for juno 21:30:16 anyhow, food for thought 21:30:22 armax: That will need to be fixed. I'm just pointing out that it should be a configuration fix. 21:30:39 marun: agreed, that said it might not be the desirable one 21:30:44 armax: thanks for the feedback…. let's follow up off line…. good suggestions 21:32:02 mestery: I will continue debugging the new LBaaS and as for policy group, Summit and I have talked about developing a scenario test 21:32:04 mlavalle: sure thing, feel free to poke me offline 21:32:24 that's all I have this week 21:32:29 mlavalle: thanks! 21:32:39 #topic L3 21:32:41 carl_baldwin: hi there! 21:32:43 mestery: hi 21:32:45 mlavalle: yes for GBP, plan to get back to you for the scenario test 21:33:07 I’ve updated status on the page for most projects. Some moved out to Kilo. 21:33:36 We’re still iterating on DVR. There are two more patches to merge to enable it. 21:34:20 We’re waiting to hear from amotoki on this one: https://review.openstack.org/#/c/102398/ 21:34:33 Otherwise, I believe it is ready for merge. 21:34:42 The last one is this: https://review.openstack.org/#/c/89694/ 21:34:51 amotoki’s feedback didn’t sound like a no-no 21:35:12 carl_baldwin: So we're really close with DVR, that's good news! 21:35:37 mestery: We will move in to hardening and cleaning up a few backlog items during juno-3. 21:35:45 carl_baldwin: ack 21:36:10 I am working through a couple of issues on the last patch but it should be close as well. 21:36:20 carl_baldwin: I’ll do another review of the first of those tonight or tomorrow. 21:37:02 carl_baldwin: I hit the first DVR bug during my testing 21:37:14 Merging these will allow us to enable an experimental job to run tempest with DVR and root out issues. 21:37:26 Sukhdev: is it reported? 21:37:30 will be filing a bug later today 21:37:32 carl_baldwin: ++, that will be good! 21:38:06 Sukhdev: thanks for testing! 21:38:40 We’re anxious to start hitting it with more testing. 21:38:55 #link https://bugs.launchpad.net/neutron/+bugs?field.tag=l3-dvr-backlog 21:38:58 sorry, I fell 21:39:06 armax: welcome back ;) 21:39:15 I'm also trying to test it every update of Carl's branch. 21:39:21 It works from time to time :) 21:40:10 Thanks for the updates carl_baldwin and team! 21:40:13 mestery: I think that is all. 21:40:20 #topic IPv6 21:40:22 sc68cal: hi there! 21:40:27 if someone wants to attend infra's meeting tomorrow and ask about the status of multinode that can't hurt 21:40:53 anteaya: I will follow up with them 21:40:54 anteaya: I was actually curious about it 21:40:56 hello 21:41:03 kk, sorry sc68cal 21:41:06 anteaya: Thanks. I’m curious too. 21:41:20 Right now, I think we're working on getting stateful/stateless merged 21:41:36 * sc68cal digs up link 21:41:48 https://review.openstack.org/#/c/106299/ 21:42:04 sc68cal: I think that is close 21:42:08 #link https://review.openstack.org/#/c/106299/ 21:43:09 that's it for me 21:43:18 sc68cal: Thanks! 21:43:22 #topic ML2 21:43:26 rkukura and Sukhdev: Hi there! 21:43:34 mestery: hi 21:43:36 hi 21:43:46 Sukhdev: Do you want to start? 21:43:47 rkukura: go ahead 21:43:49 ok 21:44:28 We are focusing on a couple of BPs needed by several mechanism drivers: hierarchical port binding, and related type driver changes 21:44:58 rkukura: I am actively working on the unified binding model :) 21:45:13 rkukura: today I should have a new patchset 21:45:34 Trying to make sure we have agreement on how all the parts fit together, and that mechanism drivers in juno will be able to support hierarchical networks. 21:45:48 armax: Yes, saw the WIP patch set 21:46:27 Plan is to get the DVR-required changes in first, then the hierarchical port binding changes 21:46:59 We are also tracking a couple bugs, but nothing that needs discussion here AFAIK 21:47:08 Sukhdev: Anything to add? 21:47:29 rkukura: not really - this covers well 21:47:39 Anything else on ML2? 21:47:47 rkukura Sukhdev: Thanks for the ML2 updates! 21:48:10 #topic Group Based Policy 21:48:12 Sukhdev: hi there! 21:48:24 SumitNaiksatam: Hi there 21:48:25 mestery: you mean Sumit 21:48:29 Sukhdev: Sorry, wrong person (autocorrect) 21:48:30 mestery: hi :-) 21:48:49 we had a productive code sprint last thursday 21:49:07 a number of folks from the team participated and there were some new folks joining as well 21:49:20 in general, most of the code has been review for a while 21:49:40 we focussed on bridging the remaining gaps, and the vendor drivers that are planned 21:50:09 i think we should be in a good position to have this stuff in review pretty soon 21:50:29 i already saw some WIP patches from kevinbenton (that came out of the code sprint) 21:50:43 while on that, armax thanks for the latest round of reviews 21:50:51 rkukura: anything else to add on GBP? 21:50:53 SumitNaiksatam: np 21:51:27 mestery: thats it from me, happy to take questions on GBP if there are any 21:51:39 Thanks SumitNaiksatam! 21:51:54 #topic Open Discussion 21:51:59 so let's talk a little about GBP 21:52:44 mestery asked me to open a thread on the ML (which I will do tomorrow) about GP, HK, ATL and the "simpler" approach 21:52:54 regXboi: like the exchange rate with USD and EUR? 21:53:11 salv-orlando: That one always cracks me up ;) 21:53:26 regXboi: link? 21:53:39 regXboi: Yes, based on discussion with you, I wanted your input there. 21:53:50 markmcclain: no link yet - the ML thread will be *started* tomorrow 21:53:51 markmcclain: no thread yet 21:54:01 this is just the "heads up" 21:54:13 so that folks don't think it's from left field 21:54:39 #info OVS and LB plugins will be proposed for removal this week. 21:54:44 You've all be warned :) 21:54:48 markmcclain: Yay! 21:55:01 It also would be great to discuss a DVR a little bit. 21:55:18 markmcclain: would have been nice to have that info in the thrid party meeting today under the deprecations topic 21:55:33 so maybe next time 21:55:34 anteaya: I think I've mentioned this in previous 3rd party meetings. 21:55:41 anteaya: sorry I thought 2 cycles was enough :) 21:55:44 np 21:56:12 repetition is important with this group, I have found 21:56:40 OK, I think we're pretty much done here. 21:56:49 anteaya: Over communication is better than under communication 21:56:58 Thanks everyone, see you all on the ML and in #openstack-neutron! 21:57:00 #endmeeting