09:01:58 #startmeeting Dragonflow 09:01:59 Meeting started Mon Aug 22 09:01:58 2016 UTC and is due to finish in 60 minutes. The chair is oanson. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:02:00 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 09:02:02 The meeting name has been set to 'dragonflow' 09:02:03 Hello. 09:02:08 hola 09:02:08 Who is here for the Dragonflow meeting? 09:02:08 Hello. 09:02:22 Already better than last time :) 09:02:29 o/ 09:02:29 hello 09:03:09 Let's wait another minute to see if anyone is late 09:03:13 \o 09:03:45 All right 09:04:03 #info oshidoshi gsagie yuli_s oanson in meeting 09:04:25 yuval is in the meeting as well.. just fyi 09:04:26 :) 09:04:37 #info yuval is in meeting as well 09:04:45 Thank you. I missed him in all the excitement 09:04:58 #topic Newton Roadmap 09:05:00 Well... he was hiding.. 09:05:05 hidding 09:05:18 who's hiding? 09:05:34 its Ocata roadmap.. 09:05:37 ML2 has been merged a while back. I have been testing it, but it isn't going very well. 09:05:59 Ocata is next version, no? Isn't October Newton is for October? 09:06:18 Ocata is next version, no? Isn't October Newton? ** 09:06:26 yes, i thought we arent talking about this version sorry 09:06:32 Welcome lihi!!!! 09:06:44 Welcome 09:06:59 We'll do Ocata in the Barcelona summit. We can discuss there what features we want and which directions to take. 09:07:00 :) 09:07:04 So everyone should be there 09:07:10 #info lihi is also in meeting 09:07:16 Welcome. 09:07:47 As I was saying, I have been testing ML2. I couldn't get the fullstack to run. The L3 test + sec groups fail. 09:08:18 Which is strange, because when I run it with real VMs, it works. So I'm looking into it, unless someone volunteers to take it of my hands 09:08:50 Additionally, Neutron have added transaction guards on {create,update,delete}_port commands, so our L3 service plugin started throwing exceptions. 09:09:13 This patch is supposed to be a workaround. https://review.openstack.org/#/c/355736/ 09:09:45 I know midonet is having similar issues, so I will see if I can contact the relevant people and learn the correct solution 09:10:07 @oanson, the patch you mentioned is about ML2, not L3 09:10:26 oshidoshi, it's the L3 service plugin in ML2. 09:10:30 So technically, it is both 09:10:34 ok 09:11:27 The VLAN network plugin is also up. I was trying to test it before merge which is when I ran into some issues. If someone can test it and then +1 the patch, that would help a lot. 09:12:00 I was hoping DuanKebo would be here and they could maybe take it. 09:12:24 Metadata is in. That's nice. 09:12:51 I am also working on the packaging, but I am running into strange pip behaviours, so I have a lot of untested ansible code :) 09:13:00 Anything I left out? 09:13:32 i saw you merged the DPDK code 09:13:33 there are several merge conflicts, what about those? 09:13:42 anyone tested it by any chance? 09:13:55 oshidoshi, patch owners have to fix it. 09:14:15 oanson: of course they do, this is a reminder 09:14:42 oshidoshi, you out-polited me. I was caught by surprise. 09:14:53 * oshidoshi blushing 09:15:34 #topic Bugs 09:15:41 Hello~ DuanKebo is not here right now. But I can help about the problem of L3 + secgroup fullstack cases fail. I will try this on my local environment:) 09:16:12 yuanwei1, thank you. 09:16:33 All right, yuli_s, as our bug master, any word? 09:16:45 sec. 09:17:01 we need ownership of this bug: 09:17:07 https://bugs.launchpad.net/dragonflow/+bug/1611707 09:17:07 Launchpad bug 1611707 in DragonFlow "Changing tunnel type for existing chassis does not work" [High,New] 09:17:22 and here related to ml2 09:17:24 https://bugs.launchpad.net/dragonflow/+bug/1605535 09:17:24 Launchpad bug 1605535 in DragonFlow "auto-topology-allocation doesn't work with ml2" [Medium,New] 09:17:47 i am working on this bug https://bugs.launchpad.net/dragonflow/+bug/1615351 09:17:47 Launchpad bug 1615351 in DragonFlow "Bug in subnet create command" [High,New] - Assigned to Yuli (stremovsky) 09:18:35 yuli_s, bug 1280522 doesn't have a priority. 09:18:35 bug 1280522 in Blazar "Replace assertEqual(None, *) with assertIsNone in tests" [Undecided,In progress] https://launchpad.net/bugs/1280522 - Assigned to zhangguoqing (474751729-o) 09:18:56 will fix it 09:19:05 Great. Thanks. 09:19:38 So who's taking bug 1611707 09:19:38 bug 1611707 in DragonFlow "Changing tunnel type for existing chassis does not work" [High,New] https://launchpad.net/bugs/1611707 09:19:38 ? 09:20:09 What about 1605535? 09:20:31 Great. 09:20:36 Any other bug-related stuff? 09:21:02 nothing for now 09:21:15 #topic Performance 09:21:23 yuli_s, anything interesting on this front? 09:21:43 hm, I made few more tests 09:21:49 and send the results, 09:22:13 i was working on setting up a new lab with real dragonflow, but here got a problem to create a network 09:22:42 yuli_s, great. No worries. 09:22:50 Did you get a chance to see if https://review.openstack.org/#/c/355237/ made a difference? 09:23:17 i have not tested it 09:23:38 No worries. I was just curious. 09:23:54 ok 09:24:43 Anything else here? 09:25:06 #topic Open Issues 09:25:33 weizj isn't here, but his patch https://review.openstack.org/#/c/356860/ raised an interesting question. 09:25:45 Do we want to use argparse (or something similar) in df-db? 09:25:50 Are there any recommendations? 09:26:14 Obviously the CLI must remain identical. 09:27:34 Don't all shout at once. 09:27:51 Well, let's move on. 09:27:56 Anyone wants to raise a question? 09:28:44 Thanks for coming. 09:28:47 #endmeeting