15:00:37 #startmeeting neutron_l3 15:00:39 Meeting started Thu Jul 12 15:00:37 2018 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:40 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:42 The meeting name has been set to 'neutron_l3' 15:00:46 hi 15:02:38 * mlavalle waiting a little for others to join 15:03:59 ok, let's get moving 15:04:11 #topic Announcements 15:04:43 Rocky-3 is a little more that a week away 15:04:54 July 23 - 27 15:05:25 thanks to manjeets or keeping an eye on the targeted bluperints and community goals ;-) 15:05:53 :) 15:06:28 also to haleyb who took the time yesterday to go through my big multiple port bindings patch. I'll address your comments today ;-) 15:07:05 Please also keep in mind that next week we are going to release neutron-lib again 15:07:09 mlavalle: sorry, totally forgot 15:08:07 should this https://review.openstack.org/#/c/577245/ wait then for 1.18 ? 15:08:30 boden sent a message yesterday. If you have a patch that you want to be included in next week's release, please leave a comment in it with "rocky-candidate" 15:09:16 we are especially interested to get this patch in that release: https://review.openstack.org/#/c/319328/ 15:10:05 also this one: https://review.openstack.org/#/c/580786 15:10:32 although this one will be affected by haleyb's comments to the port binding patch, so expect an update for it 15:11:04 mlavalle: were you going to fix the "returns" in that second one ? 15:11:21 i know it was a nit... 15:11:47 the first one i hadn't taken a look at... quite large 15:11:57 oh, and old 15:13:34 haleyb: yes, in the get_port_binding_by_status_and_host I am going to let the caller request an exception to be raised if the expected binding (active or inactive) is not found 15:14:04 I'll explain further the reasoning behind that in my response to youir comments 15:14:52 mlavalle: ok, i'll wait for the comments, i didn't remember callers raising if not found but you would know best 15:15:42 mlavalle: for the lib change were you going to address boden's nit? otherwise it's good to go 15:16:12 missing ":returns: ..." in docstring 15:16:19 haleyb: yes and also, if we adopt the solution I propose to your comments, the code will change a bit 15:17:56 before the change, I'll ping you in channnel to discuss 15:18:11 ack 15:18:31 This coming Tuesday 17th is the deadline to propose talks for Berlin 15:18:49 any other announcements from the team? 15:20:43 ok 15:20:48 #topic Bugs 15:22:00 First https://bugs.launchpad.net/neutron/+bug/1757482 15:22:00 Launchpad bug 1757482 in neutron "IP address for a router interface allowed outside the allocation range of subnet" [High,In progress] - Assigned to Miguel Lavalle (minsel) 15:22:39 hongbin had some comments on the fix last week: https://review.openstack.org/#/c/575444/ 15:22:59 o/ 15:23:27 I responded to them but I would like haleyb to take a look. If you are still ok with the patchj, pleaase merge it 15:23:37 mlavalle: ack, will look 15:24:01 mlavalle: your respond is fair enough, i am fine with it 15:24:17 hongbin: cool, thanks ;-) 15:24:57 Next bug is https://bugs.launchpad.net/neutron/+bug/1711463 15:24:57 Launchpad bug 1711463 in neutron "TestNetworkBasicOps.test_network_basic_ops failed with "Timed out waiting for 172.24.5.8 to become reachable"" [High,Confirmed] - Assigned to Miguel Lavalle (minsel) 15:25:30 For this one, I spent time preparing for this meeting searching Kibana for hits 15:26:14 the bug, as filed by Ihar and later also updated by boden, referred to failures with greandes jobs 15:27:25 I created searches (which I added in #3) for neutron-grenade-dvr-multinode, neutron-grenade-multinode and neutron-grenade 15:28:36 I don't get any hist over the past 7 days 15:29:16 If I change the search for any build in openstack/neutron, I get 28 hits over the past 7 days 15:29:48 24 for of those, though, are with neutron-tempest-dvr-ha-multinode-full, so it is probably expected 15:30:10 There are 4 with build legacy-grenade-dsvm-neutron-linuxbridge-multinode 15:30:25 do you guys know where are we running that job? 15:30:38 it's not in master check 15:32:25 i don't know then, strange 15:32:47 if recall correctly this used to be there 2 cycles back, since then, zuul v3 and other changes may have lost this job 15:33:57 or job name may have changed 15:33:58 http://codesearch.openstack.org/?q=legacy-grenade-dsvm-neutron-linuxbridge-multinode&i=nope&files=&repos= 15:34:17 shows up in kibana over the past 7 days 15:35:34 I'll dig in 15:35:38 it's not in any zuul listing on changes 15:35:56 does kibana records experimental jobs ? 15:36:06 we can also discuss in the CI meeting next week. maybe slawek has seen it 15:37:07 haleyb: would it be a waste of time if we try to see why the failures in the neutron-grenade-dvr-multinode job? 15:37:52 mlavalle: yes, that linuxbridge job is in experimental, which only runs on-demand 15:38:16 ok 15:38:38 and it's been run a few times this past week 15:39:02 * njohnston checks graphite... 15:39:28 * haleyb even remembers njohnston running experimental :-p 15:40:12 https://review.openstack.org/#/c/578886/ I did check experimental on this 15:40:47 ahh yeah 15:40:52 mlavalle: regarding neutron-grenade-dvr-multinode - is that failing on master? 15:41:12 haleyb: no, I didn't get any hits over the past 7 days 15:41:36 not for the bug we are talking about 15:42:32 so I won't worry about the hits in the linuxbridge job 15:42:55 The next bug that I have is https://bugs.launchpad.net/neutron/+bug/1766701 15:42:55 Launchpad bug 1766701 in neutron "Trunk Tests are failing often in dvr-multinode scenario job" [High,Confirmed] - Assigned to Miguel Lavalle (minsel) 15:43:23 I haven't made much progress on this one 15:43:40 But the next step is to do the same that I am doing with the previous one in Kibana 15:44:06 I'll update the bug and report next week in this meeting 15:44:25 Any other bugs we should discuss today? 15:44:45 there might have been one new one 15:44:56 https://bugs.launchpad.net/neutron/+bug/1780094 15:44:56 Launchpad bug 1780094 in neutron "The transition of a router from distributed to centralized may mistakenly get HA enabled" [Medium,In progress] - Assigned to Kailun Qin (kailun.qin) 15:45:08 currently being debugged 15:45:43 right, I forgot. swami mentioned it last week 15:46:00 should we raise its priority so it is visible? 15:46:19 in the searches we have in the meeting etherpad 15:46:25 it's being worked on, i'm watching it 15:46:33 ah ok, perfect 15:46:45 ok, let's move on 15:46:57 #topic Port Forwarding 15:47:09 I have been reviewing the series 15:47:26 https://review.openstack.org/#/q/topic:bp/port_forwarding+(status:open+OR+status:merged) 15:47:43 mlavalle, first patch adding db and ovo now looks good to me ! 15:47:44 manjeets and haleyb have also take a look at some of the patches 15:47:46 thanks 15:48:22 This past Tuesday I spent time with this one https://review.openstack.org/#/c/574673 15:48:52 and was having a hard time understanding why we need locks to handle "race conditions" 15:49:33 in principle, the db layer transactions must handle most of that 15:49:39 what do you guys think? 15:49:58 am I being too simplistic? 15:50:40 What patch introduced the locks to prevent race conditions? Does it reference a bug? 15:50:52 Hello btw :) 15:50:54 mlavalle: and i'm sure you saw the response when asked about it in PS8 15:51:11 davidsha: https://review.openstack.org/#/c/574673 15:51:22 i have seen another change recently doing similar things with @runtime.lock... 15:51:49 haleyb: to tell you the turth, I didn't understand it, so I asked again in PS11, this past Tuesday 15:52:31 oh, the answer was even longer this time, i haven't parsed it all yet 15:53:04 ok, please let's take a look and try to steer this patch towards a simpler approach 15:53:23 hi I want to know how we can make 528336 to rocky release? 15:53:27 I feel relieved that I am not the only one having trouble undertanding the reasoning behind it 15:54:09 also 472289 15:54:11 patch looks cheesy, but I need to spend more time to understand it as well and need to catch up with zhaobo's replies 15:54:18 mlavalle, ^^ 15:54:40 xbzhang: gives us a moment to finish this topic 15:54:56 ok 15:55:09 davidsha: https://review.openstack.org/#/c/574673/ 15:55:21 This is the patch,. Take a look if you have some time 15:55:49 other than the locks business, the 3 server side patches seem in reasonably good shape to me 15:56:02 mlavalle, Thanks, I'll check it out! 15:56:52 davidsha: you might find useful looking at the spec: https://specs.openstack.org/openstack/neutron-specs/specs/rocky/port-forwarding.html 15:57:07 #topic OpenFlow DVR 15:58:03 xbzhang: https://review.openstack.org/#/c/528336 seems like a fair bet to make it to Rocky 15:58:36 That is actually a great first step for OpenFlow DVR 15:58:38 ok thanks 15:58:56 472289 15:59:02 seems more complicated 15:59:15 but I will add to my pile and see what we can do 15:59:29 ok, ran out of time 15:59:40 Thanks for attending 15:59:47 #endmeeting