21:02:44 #startmeeting Networking 21:02:45 Meeting started Mon Jul 22 21:02:44 2013 UTC. The chair is markmcclain. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:46 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:47 hi All 21:02:49 hi 21:02:49 The meeting name has been set to 'networking' 21:02:53 hi 21:03:15 Ok very different agenda this week 21:03:18 #link https://wiki.openstack.org/wiki/Network/Meetings 21:03:52 I want to spend the bulk of the meeting this week discussing 3 things: bugs, docs, and H3 21:04:01 #topic Announcements 21:04:10 #link http://tarballs.openstack.org/neutron/neutron-milestone-proposed.tar.gz 21:04:28 H2 was released last week. Great job by the entire team 21:04:41 Hi all 21:05:04 we completed 21 blueprints and fixed 99 bugs 21:05:36 #topic bugs 21:05:53 We've got two critical bugs that we're tracking this week 21:06:14 https://bugs.launchpad.net/neutron/+bug/1194026 21:06:16 Launchpad bug 1194026 in neutron "check_public_network_connectivity fails with timeout" [Critical,In progress] 21:06:36 nati_ueno has put a lot of effort into this bug and submitted two patches for it 21:06:52 the first reduced some of the problems but not all of them 21:07:41 until we can fix the false negatives the Infra team will keep the job as non-voting in the gate which less than ideal 21:08:13 nati_ueno has proposed a 2nd fix: https://review.openstack.org/#/c/37576/ 21:08:29 false negatives - you mean the test is succeeding whereas it should fail? 21:08:45 no.. the test is failing when it should succeed 21:09:28 actually false negative is the wrong way to classify it 21:09:36 the test found a race condition 21:09:46 so it's not a false at all 21:09:46 it may not be false negatives .. because floating ip configuration is not finished in 60 sec.. it sounds problem 21:10:16 yeah.. my brain is moving slow today 21:11:02 markmcclain: as slow as it might be, it will still be fast enough to pass and lap mine 21:11:15 haha 21:11:34 I'd really like folks to review and test this fix as it is critical to getting the gating restored 21:12:00 my lap time is third times.. so I need someone's help. 21:12:18 looping call looks sometimes stopped.. I'm not sure why, and I can't reproduce 21:12:48 nati_ueno: I'll have a look at this with you. 21:12:56 mestery: Thanks! 21:13:01 mestery: thanks 21:13:02 nati_ueno: if it looks like it's looked, it might be because of some other green thread blocking it (rpc from server perhaps)? 21:13:15 I dislike eventlet really nowadays.. 21:13:38 nati_ueno: I'm with you on that one.. but that's something we'll have to wait to Icehouse summit to talk about 21:13:38 salv-orlando: it could be possible 21:13:51 markmcclain: I agree. 21:14:40 the other bug I'd like ot highlight involves recycling of IPs 21:14:40 Another critical bug is this one: https://bugs.launchpad.net/neutron/+bug/1201692 21:14:41 Launchpad bug 1201692 in neutron "can not create port after some port update operations" [Critical,In progress] 21:14:56 https://review.openstack.org/#/c/37384/ 21:15:03 we're short one core to review it 21:15:21 I claim it 21:15:25 salv-orlando: thanks 21:15:49 Any other bugs that should be brought up to the team? 21:16:35 #topic Docs 21:16:40 #link https://wiki.openstack.org/wiki/Network/Meetings#Docs_.28emagana.29 21:17:37 We have both doc blueprints for the manuals and I've added new bugs for the netconn-api changes too 21:18:10 Good docs are critical for new features and it is best that we start writing them now 21:18:28 just pushed this for netconn-api: https://review.openstack.org/#/c/38199/ 21:18:53 salv-orlando: awesome 21:19:25 markmcclain: did you tag the netconnapi bugs you filed on openstack-api-site? 21:19:32 yes 21:20:05 https://bugs.launchpad.net/openstack-api-site/+bug/1203864 21:20:07 Launchpad bug 1203864 in openstack-api-site "Neutron FWaaS API docs" [Undecided,New] 21:20:14 https://bugs.launchpad.net/openstack-api-site/+bug/1203865 21:20:15 Launchpad bug 1203865 in openstack-api-site "Neutron VPNaaS API Docs" [Undecided,New] 21:20:58 markmcclain: thanks. I am not just slow, but blind too. 21:21:32 no worries it is late there 21:22:18 The docs segment will be a regular item on our agenda so if there are major doc items we need to track feel to add to agenda 21:22:49 I assigned the vpn docs for me 21:23:00 nachi/sumit: can you make sure there is a member of your subteams assigned to the netconn-api bugs? 21:23:27 markmcclain: yeah, i assigned it to me (for now at least) 21:23:34 SumitNaiksatam: thanks 21:23:36 nati_ueno: I can help with VPN docs... just let me know what to do :) 21:24:00 pcm_: Ah Thanks! Ok could you take this task? 21:24:11 nati_ueno: yes 21:24:21 pcm_: thanks for taking it on 21:24:23 pcm_: OK I'll reassign it for you 21:24:31 Any question on docs before we move on? 21:24:49 #topic H3 21:24:52 nati_ueno: Offline we can talk about what I need to do 21:25:02 pcm_: surew 21:25:10 #info Last milestone to add features to the Havana release 21:25:24 #info H3 release date is Sept 5th 21:25:52 #info Neutron feature cutoff is end of day August 23rd 21:26:14 H3 follows a long holiday weekend here in the US 21:27:11 so I've move the cutoff to ensure we have the same number of review days as we had for Grizzly 21:27:53 The third milestone is usually the most taxing on the gate and gating/merges can take much longer than normal 21:28:52 we have 31 days and that time will fly by, so try not to wait until the deadline to propose code 21:29:45 Any questions on H3 scheduling? 21:30:15 #link https://wiki.openstack.org/wiki/Network/Meetings#Havana_Design_Summit_Priorities_Reminder 21:30:45 I wanted to reminder everyone of where we stand on the major priorities that were developed at the Portland sumit 21:30:51 I'm fine with it. Shall we try and enforce a blueprint cap? 21:31:01 salv-orlando: getting to that :) 21:31:05 k 21:31:32 markmcclain: Are we going to deplicate linuxbridge and ovs in this release? 21:32:05 nati_ueno: I'd like to, but want to make sure that ML2 is ready as a drop in replacement 21:32:05 nati_ueno: I thought that would happen in Icehouse? 21:32:27 markmcclain: By deprecate, would LB and OVS still be around for Icehouse? 21:32:33 Or gone completely? 21:33:01 mestery: we'd originally talked about deprecating in Havana and removing in Icehouse 21:33:22 markmcclain: OK, that makes sense, I agree with that. We'll work with that assumption on the ML2 team for the next 31 days then. :) 21:33:23 the nice thing about deprecating is that we can make that decision during the RC phase 21:33:30 markmcclain: Perfect. 21:33:38 Ok thanks. I got it 21:33:42 sounds good 21:34:12 the warning will be such that it will be remove at the earliest in Icehouse, but if a show stopper pops up we can delay until J 21:35:12 Now as salv-orlando pointed out we probably need to start considering a cap of blueprints for H3 21:35:34 currently have 63 blueprints targets and we only completed 21 during H2 21:36:27 #link https://launchpad.net/neutron/+milestone/havana-3 21:36:43 considering that we typically land more blueprints in the 3rd milestone, I respectfully wish to suggest 35 blueprints as a reasonable target 21:36:53 (on the optimistic side) 21:37:04 salv-orlando: you were reading my mind :) 21:38:38 our velocity will be aided somewhat by the fact that we have several blueprints (VPN, FW, service types) that are in progress from H2 and should land relatively soon 21:38:42 Looking at that page, salv-orlando is currently assigned almost half of his own specified target quota. :) 21:38:57 I will cleanup 21:39:24 mestery: you're noticing another trend 21:39:47 which is a lot of the blueprints are for vendor plugin features 21:39:57 the trend of me postponing the stuff from one milestone to another and never getting it done? :) 21:40:58 markmcclain: What are your thoughts on the vendor plugin situation? 21:41:09 with regard to all the BPs in H3? 21:41:15 I'll be reaching out to the vendor contacts looking for their priorities 21:41:23 some we can mark as medium 21:41:38 some we will mark as low which means if they make it great if not that ok 21:41:55 and others we'll assign to the next milestone which are items for Icehouse 21:42:14 markmcclain: That makes sense. 21:42:54 vendor blueprints are really a random mix of patch set sizes.. some are small and easy to review and others are large 21:43:30 may be it is good to let them report expected patch size 21:44:21 nati_ueno: that will definitely be one of the factors 21:45:01 but there are also business factors too, so I want to be a collaborative process 21:45:08 also I'm not sure all of proposer is still working on the bps. 21:45:38 markmcclain: I agree 21:46:13 the end goal is by next week to have ~35 blueprints that are medium and above 21:46:21 markmcclain: Pragmatism is a great goal to have here. 21:47:04 those are the items that we will track for release planning 21:47:29 items marked low will still be accepted but it will all depend on reviewer bandwidth 21:48:20 I think blueprint owners should also explicitly note whether they expect to drop huge commits. 21:48:52 in that case - either split the blueprint or pull the deadline forward (i.e. : one week before) 21:49:45 salv-orlando: that's a good idea too 21:50:18 Any questions about the H3 plans? 21:50:20 I know because I have one which is huge and I'm considering what to do with it... 21:51:25 yeah.. I get the feeling we'll have a few in that situation 21:51:34 #topic Open Discussion 21:52:21 so we spent a lot of time on planning any items that need to be brought up, but were skipped because of the special agenda today? 21:53:37 would like to understand where we are on https://review.openstack.org/#/c/30441/ & https://review.openstack.org/#/c/30447/, would love to get them merged 21:54:15 dkehn: I've been traveling and will go back and look at the latest revision 21:54:25 markmcclain, thx 21:54:36 I'll also ping amotoki to take another look since he's the 2nd core on it 21:55:00 thx again 21:55:35 Hello, as I anticipated in the mailing list I'm going to request a review for the first part of the Embrane's Neutron plugin soon (by today maybe)... Your feedback will be extremely appreciated when you have some time :) (I see that you really are full of work to do!!) 21:56:30 markmcclain, ivarlazzaro: we would still need a core maintainer right? 21:56:31 I'm using the open discussion of this meeting just to "anticipate" this and answering any question if any :) 21:56:54 salv-orlando: correct 21:57:28 ivarlazzaro: we can talk offline, but plugins submitted by non-core team members require a core team member to sponsor 21:57:40 markmcclain, salv-orlando: I red in the wiki that the company could propose the plugin developer itself for maintaining it 21:57:56 pointer? 21:58:12 proposed core developers? 21:58:29 Seems like a mistake on the wiki if it's there. 21:58:51 * markmcclain searches for link too 21:59:18 markmcclain, salv-orlando: https://wiki.openstack.org/wiki/NeutronDevelopment 21:59:35 at Adding a Plugin to the Public Quantum Repository 22:00:15 ivarlazzaro: thanks 22:00:22 ivarlazzaro: You may also want to consider the option of developing drivers for the ml2 plugin as an alternative to a new monolithic plugin 22:00:40 markmcclain salv-orlando: "or by having one of the developers of the plugin become a member of the core developer team" its probably a misunderstanding then :) 22:00:40 ivarlazzaro: I think there is confusion there, that means you can become core by doing all the duties a core dev does, not by proposing to be a core dev. markmcclain, please correct me if I'm wrong. 22:01:17 ivarlazzaro rkukura: I think this plugin is an L3 plugin, not an L2 plugin, based on a mailing list thread I remember reading. 22:01:29 mestery: that's correct 22:02:07 well let's at least glance at blueprint spec and code first 22:02:27 salv-orlando ivarlazzaro: I think this also depends on Bob Melander's patch, no? 22:02:28 ivarlazzaro: that person has been acceptable as a core team member in the general sense first which is why we require a current team sponsor first 22:02:55 because nearly all new plugin contribs do not have a developer in house who is already a core team member 22:02:58 salv-orlando: I agree, anyway I'm willing to improve it a lot as we go on 22:03:12 #action markmcclain to clarify working on plugin contribution 22:03:31 Anything else for this week? 22:04:18 Thanks again for everyone's hardwork on H2 and I'm really excited on the Havana release is shaping up 22:04:25 #endemeeting 22:04:44 adieu 22:04:53 #endmeeting