09:00:16 #startmeeting dragonflow 09:00:17 Meeting started Mon Jul 25 09:00:16 2016 UTC and is due to finish in 60 minutes. The chair is oanson. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:00:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 09:00:20 The meeting name has been set to 'dragonflow' 09:00:24 Hi. 09:00:32 Who's here for the Dragonflow weekly meeting? 09:00:38 Hello 09:00:45 o/ 09:01:12 Looks like a skeleton crew. 09:01:19 #info yuli_s, oshidoshi in meeting 09:01:32 Let's wait another minute to see if others join. 09:02:02 ok, i will be back in a minute 09:02:14 yuli_s, I'm counting 59 seconds 09:02:57 All right, let's start 09:03:03 #topic Newton Features 09:03:14 Firstly, DB consistency. 09:03:36 hujie uploaded a patch. Patchset 19. It has to be reviewed. 09:03:44 ML2 and L3 Service Plugins 09:04:09 The patches were merged. Last week we discussed adding a gate job for the ML2 configuration. 09:04:12 what's the patch #? 09:04:24 (refering to the db consistency) 09:04:24 DB consistency: 09:04:30 #link https://review.openstack.org/#/c/336377/ 09:04:44 oanson, ;) 09:04:57 I suggest waiting another week with the gate job, since it currently fails. I'll get to that in the bugs section. 09:05:20 Packaging - this one was on me. There was no progress, but I promise to have something for next week. 09:05:33 VALN networking - liuhaixia uploaded a patch 09:05:36 VLAN* 09:06:00 It should be reviewed as well. While we're at it, she also uploaded a flat networking patch. 09:06:03 Links in a second: 09:06:36 #link spec: https://review.openstack.org/#/c/288270/ 09:06:48 #link VLAN: https://review.openstack.org/#/c/340789/ 09:07:09 #link flat: https://review.openstack.org/#/c/340846/ 09:07:33 I'll also have the metadata service updated for next week. 09:07:54 #link metadata https://review.openstack.org/#/c/309746/ 09:08:05 #topic performance testing and bugs 09:08:09 yuli_s, take it away 09:08:25 ok, 09:08:46 I assigned one bug to myself 09:08:50 related to redis 09:08:53 https://bugs.launchpad.net/dragonflow/+bug/1597332 09:08:53 Launchpad bug 1597332 in DragonFlow "df dropall doesn't work for redis" [Medium,New] - Assigned to Yuli (stremovsky) 09:09:02 #link https://bugs.launchpad.net/dragonflow/+bug/1597332 09:09:05 we have few more bugs needed to take care 09:09:25 Last week we said bug owners need to verify their bugs. 09:09:25 but they are not that high priority 09:09:30 Check geneve and vxlan 09:09:30 https://bugs.launchpad.net/dragonflow/+bug/1571551 09:09:30 Launchpad bug 1571551 in DragonFlow "Kernel module vport_geneve.ko fails to load on ubuntu" [High,New] 09:09:36 I doubt that happened, but I guess we can say that for this week as well. 09:09:52 #action everyone Verify assigned bugs. Close invalid bugs. 09:09:54 another bug: 09:09:56 Check that unstack works correctly: 09:09:57 https://bugs.launchpad.net/dragonflow/+bug/1582122 09:09:58 https://bugs.launchpad.net/dragonflow/+bug/1597568 09:09:58 Launchpad bug 1582122 in DragonFlow "unstack couldn't unload openvswitch module" [Medium,New] - Assigned to lei zhang (zhang-lei) 09:09:59 Launchpad bug 1597568 in DragonFlow "Devstack can not start openvswitch when it is running" [Medium,In progress] - Assigned to wang jian (wangjian011) 09:10:07 Bug related to instllation command sequence: 09:10:08 https://bugs.launchpad.net/dragonflow/+bug/1605839 09:10:08 Launchpad bug 1605839 in DragonFlow "devstack starts NB DB driver between ova_init and ovs_start" [Medium,New] 09:10:13 yuli_s, what about these bugs? Are they new? Are they without owners? 09:10:22 Are they critical? 09:10:23 without owners 09:10:48 1582122 and 1597568 are assigned to lei zhang and wang jian 09:11:06 bug 1605839 has an open review 09:12:09 (but I can't find it now) 09:12:26 ok, 09:13:16 #link https://review.openstack.org/#/c/342530/ 09:13:23 yuli_s, anything else? 09:13:34 nop 09:13:40 Great. Thanks! 09:13:47 I have an update about the fullstack tests. 09:13:59 They appear to fail in gate due to the selective topology distribution 09:14:17 I see the ports created in q-svc and ovs, but the dragonflow controller doesn't handle them. 09:14:33 I also see that the relevant flows (table 0, ARP responders) are not installed. 09:14:49 I think the issue is that we don't use real VMs, and we need to fool the system. 09:15:08 I am still reading the code (Topology.py) but it might take a while longer. 09:15:16 #topic Open Issues 09:15:23 The floor is for the taking. 09:16:07 Anyone? 09:16:33 regarding the installation 09:16:40 oshidoshi, yes? 09:16:41 which tool are you checking? 09:16:48 opentack-ansible 09:17:05 #link https://github.com/openstack/openstack-ansible 09:17:06 did you compare other tools? 09:17:34 I looked into fuel, and one other (who's name I forgot), but I think this is the best direction. 09:18:04 did you check popularity? 09:18:32 oshidoshi, yes. There was a popularity parameter in Austin. openstack-ansible and openstack-puppet were very close together at the top 09:18:38 I prefer ansible, since it is agentless. 09:18:59 fuel is puppet-based. 09:20:08 oshidoshi, anything else? 09:21:07 not from me 09:21:13 All right. Thanks everyone for participating. 09:21:16 cheers 09:21:22 Let's do this again sometimes, like next week. 09:21:24 thanks ! 09:21:26 #endmeeting