19:00:37 #startmeeting Dragonflow 19:00:38 Meeting started Mon Sep 11 19:00:37 2017 UTC and is due to finish in 60 minutes. The chair is oanson. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:39 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:41 The meeting name has been set to 'dragonflow' 19:00:49 First spot of good news - this looks like the right time and place :) 19:00:58 Who's here for the Dragonflow weekly meeting? 19:01:04 Hi 19:01:14 dimak, lihi, leyal, irenab, I'm looking at you! 19:01:17 leyal, hi. 19:01:20 o/ 19:01:27 lihi, hi. 19:02:06 Let's give dimak another minute, and then we can start 19:02:52 All right then 19:02:58 #topic Roadmap 19:03:27 Let's see now: LBaaS - the spec is up. I saw pino added some comments. I'll respond to them this week 19:03:38 No other progress there. 19:04:00 L3 flavour - I've added our open questions to the PTG. I hope it will be more clear next week. 19:04:23 When is the vPTG? 19:04:24 I'll also try to catch mlavalle off-line to see if video conferencing the thing is an option 19:04:42 I wanted to raise that to a vote. I was thinking 2-3 Oct. 19:04:55 There's also an etherpad 19:05:16 #link vPTG etherpad https://etherpad.openstack.org/p/dragonflow-queens 19:05:28 We need to select items to discuss 19:05:45 I was thinking making the first day a scheduled discussions day, and the second day an open day 19:06:15 The etherpad is a list of optional topics? 19:06:19 Yes 19:06:30 We need to synthesise a schedule from that. 19:06:56 Not everything really needs a discussion. Features, for instance, should be straight forwards. 19:07:05 So should the L3 flavour once the Neutron side is figured out 19:07:11 Same for interop 19:07:24 I'm not sure I'll be able to attend on the 2-3 of Oct, but I'll wait for the vote :) 19:07:43 lihi, you're 1/3 of the voting party currently :) 19:08:15 I guess we'll do the vote off-line, since I want dimak, irenab, mlavalle, and pino to be there. 19:08:21 (At least) 19:08:34 And kkxue may want to join, or at least drop in. 19:09:26 Mostly, I think the following needs discussion: API, southbound API, troubleshooting, and documentation. 19:09:46 The rest is straight forwards and selecting priorities, which we do during the weeklies anyway 19:10:40 Next is RPM packaging - I'm updating the patch (link will follow in a sec). I have to ask one of the RPM guys how to debug their gates. 19:11:10 #link RPM packaging patch https://review.openstack.org/#/c/500360/ 19:11:31 dimak isn't here, and I think he is responsible for the upgrade path. 19:11:58 the db-upgrade ? 19:12:00 I haven't seen any updates to the patches, but that's all I know. 19:12:03 leyal, yes 19:12:10 db-upgrade, grenade gate, etc. 19:12:28 The spec has comments on it, and I guess it will be answered shortly (either by me or dimak). 19:13:14 This leaves ETCD publisher. lihi ? 19:13:28 I'm investigating an issue in the etcd client we use, in hope to help release the next version soon. 19:13:48 Currently this is the only thing blocking it 19:14:23 I need to update the patcha bit, but it haven't changed much. 19:14:47 So basically, the publisher is review-ready, and we're waiting for a version bump from etcd3gw? 19:16:00 Yes. Just need to remove my debugs from the code, so I would wait for that (will be tommorow) 19:16:23 Very cool! 19:16:37 Anything else for roadmap? 19:16:40 btw , i got reviews from dimak and irenab about ironic draft - i think the first part (describing the ironic networking) , can be published . i will be happy for reviews about the second parts (suggestion for implementation inside DF) 19:17:02 leyal, yes. Sorry. I should have done that already. First thing tomorrow! 19:17:12 Same for me :) 19:17:32 This is exactly the reason our docs look like they do :( 19:17:33 lihi, oanson 10x :) 19:17:40 Anything else for roadmap? 19:17:58 it's here - https://docs.google.com/document/d/1kAFRRBSQQbKjrfxvD8UYZMZeyBG4avxnY3TjqRP8SpE/edit 19:18:34 leyal, great, thanks for the link 19:18:37 leyal, cool. Thanks! 19:18:41 #topic Bugs 19:19:05 If you all turn your screen to https://bugs.launchpad.net/dragonflow/ , you'll note I unassigned almost all the bugs 19:19:14 Well, High bugs 19:19:37 That's because they were mostly assigned to people not working on them, or people not working on the project anymore 19:19:48 :( 19:19:49 This should give us a clearer picture of who does what. 19:20:06 If you are working on a bug, just grab one, and start bashing that keyboard. 19:20:28 Note there are quite a few 'incomplete' that have to be verified and can probably be closed, and a couple of others which should be fairly easy 19:21:09 Anything else for bugs? 19:21:13 Yes 19:21:20 lihi, take the floor. 19:21:28 Shachar found bugs in our deployment 19:21:39 etcd is broken, again 19:22:05 And redis + fedora26 is also broken 19:22:46 Fedora 26 is broken in other places too. Ansible group install doesn't work 19:23:21 And as I recall the gate passes on etcd except that one thing we're waiting on etcd3gw 19:23:59 All right. I'll fire up a couple of setups tomorrow to see how things go 19:24:04 Yes, that will solve it. The problem is with the ZMQ pub/sub 19:24:25 Yes. We want to continue supporting that. 19:24:45 Non-broker pub/sub is important if only for completeness 19:24:55 So that is another bug 19:25:20 I'll open them tommorow 19:25:25 All right. 19:25:36 Once you do, shoot me a message, and I'll verify and confirm them 19:25:46 ok 19:25:52 Anything else for bugs? 19:26:15 All right. 19:26:19 #topic Open Discussion 19:26:31 Anyone wants the floor? 19:26:51 No thanks, I've got my own floor at home 19:27:04 just one small thing 19:27:09 lihi, Lucky you. 19:27:13 leyal, go ahead. 19:28:12 i also mention it in the previous meeting - i started a small document - that when i encounter a trem/exection-flow that i need to investigate - or ask someone - i wirte it in the doc .. 19:28:44 I must have missed that in the previous meeting. 19:28:55 Is this a public document? Do you have a link? 19:29:32 I think that can help-me as newcomers -for understanding basic terms and flow - you invited to comment/edit/add-content 19:29:41 https://docs.google.com/document/d/1o_8Zl08xJynb1GtLYJPpXORRvoCovqdDaz4rlDXpA9E/edit?usp=sharing 19:29:55 One sec. Skimming 19:30:20 This looks like another blog post 19:30:59 leyal, fancy making it into a post? 19:31:22 A contributing.rst can also fit 19:31:50 I don't sure if it's interesting as a blog post.. 19:31:51 I think 19:32:11 A blog post has the benefit of not needing to be updated. 19:32:31 If we make it a contributing.rst, we have to make sure it's up-to-date. 19:32:53 Then again, our entire docs is so far out of date, we chiselled it in rock 19:33:24 leyal, up to you. 19:33:42 yep , but on the other hand - it's just a help doc. it's not something that interest people that not contributing .. 19:34:29 Sure. I'm convinced and out-voted :) 19:34:51 sure , currently i will continue to update it.. 19:35:21 I'm confused - you plan to push it to master, right? As a patch? 19:36:27 well - currently - i don't have specific plan for that .. 19:36:52 All right. Just wanted to align our expectations 19:36:59 Still a cool document. 19:37:13 Anything else for open discussion? 19:37:25 10x 19:38:21 All right. Then let's knock off early! 19:38:35 Thanks for coming out at this late hour. 19:38:56 #endmeeting