15:00:04 #startmeeting neutron_l3 15:00:04 Meeting started Thu Aug 13 15:00:04 2015 UTC and is due to finish in 60 minutes. The chair is carl_baldwin. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:07 hi 15:00:08 The meeting name has been set to 'neutron_l3' 15:00:16 moo 15:00:17 hi 15:00:20 #topic Announcements 15:00:36 hi 15:00:39 hi 15:00:42 hi 15:00:58 #link https://wiki.openstack.org/wiki/Liberty_Release_Schedule 15:01:22 Liberty continues to march on. Feature freeze will come at the end of the month. 15:01:26 hi 15:01:38 Just three weeks away. 15:01:42 Any other announcements? 15:01:47 hi 15:02:35 #topic Bugs 15:02:37 mlavalle: ping 15:02:42 pong 15:02:54 today I want to highlight two bugs 15:03:10 https://bugs.launchpad.net/neutron/+bug/1404743 15:03:10 Launchpad bug 1404743 in neutron "sporadic test failures due to VMs not getting a DHCP lease" [Critical,Confirmed] - Assigned to Ryan Moats (rmoats) 15:03:31 This one was discussed this week during the weekly Neutron IRC meeting 15:03:36 it became critical 15:04:12 any feedback on this regXboi, since it is assigned to you? 15:04:25 I have one more update for the first failure signature 15:04:45 the tempest case *IS* in fact using nova for the floating-ip setup 15:04:56 and not handling the 202 response back correctly 15:05:07 I still need to look at the second case some more 15:05:07 Are the 1400 hits from 1400 difference failed runs? 15:05:25 carl_baldwin: no - when the failures happen, they happen in groups 15:05:35 so a single run may trigger 20+ hits 15:06:14 carl_baldwin: the problem is that the test code in one case is poorly written and creates a race condition 15:06:39 regXboi: what's the name of the test scriupt? 15:06:42 I still am triaging the second case and then need to re-visit to see if there are more 15:07:05 regXboi: Is there any way to write the logstash to isolate the various failure modes? 15:07:18 carl_baldwin: I haven't come up with a good query yet 15:07:28 mlavalle: I'll update the bug with that information 15:07:31 regXboi: ok, thought I’d ask. 15:07:45 regXboi: ok, i'll look for your update 15:07:51 carl_baldwin: I've been trying to come up with better queries - that's a bit of a blunderbuss 15:08:26 mlavalle: TestVolumeBootPattern:test_volume_boot_pattern 15:08:34 is the nova one 15:08:46 regXboi: understood 15:09:01 the other signature looks to be either a stale or lost port id, which frankly, scares me 15:09:04 regXboi: if i rememebr correctly, that one has been problematic for aloooooong time 15:09:24 mlavalle: I suspect so - it associates the FIP via nova client, which gives back a 202 15:09:32 and then never checks the FIP status to be sure it's active 15:09:37 that's a race condition waiting to happen 15:09:43 of course 15:10:19 anyway, that's all I have on this for now 15:10:25 regXboi: Thanks for the updates. 15:10:26 I'll keep the bug updated as I dig more out 15:10:43 regXboi: Let us know if you need a hand. 15:10:51 carl_baldwin: ack 15:10:57 the second bug is https://bugs.launchpad.net/neutron/+bug/1461172 15:10:57 Launchpad bug 1461172 in neutron "neutron.tests.functional.agent.test_l3_agent.MetadataL3AgentTestCase.test_access_to_metadata_proxy times out intermittently" [High,Confirmed] - Assigned to Carl Baldwin (carl-baldwin) 15:11:23 this one was assigned to us by mestery during the weekly neutronmeeting 15:11:58 I see it assigned to carl_baldwin. I don't know if you've had time to take a look 15:12:00 ? 15:12:18 I have not looked at it. 15:12:28 do you need help? 15:12:38 triaging? 15:12:44 Yes, I could use a hand. 15:12:51 ok, i'll assign it to me 15:13:01 i'll dig and report back soon 15:13:37 regXboi: any other bug you want to highlight? 15:13:58 mlavalle: I have a few from the dvr backlog 15:14:21 regXboi: you have the floor. 15:14:56 the first is http://bugs.launchpad.net/neutron/+bug/1405910 15:14:56 Launchpad bug 1405910 in neutron "distributed router can not add routes" [High,In progress] - Assigned to Brian Haley (brian-haley) 15:15:09 this has a patch set that could use review so we can get it merged 15:15:33 the second is http://bugs.launchpad.net/neutron/+bug/1369721 15:15:34 Launchpad bug 1369721 in neutron "manually moving dvr-snat router fails" [High,In progress] - Assigned to Oleg Bondarev (obondarev) 15:15:41 it also has a patch set that needs reviewing 15:15:54 i think that first one is ready to go 15:16:04 the third is http://bugs.launchpad.net/neutron/+bug/14365473 15:16:05 regXboi: Error: malone bug 14365473 not found 15:16:16 let's try that again 15:16:27 the third is http://bugs.launchpad.net/neutron/+bug/1365473 15:16:28 Launchpad bug 1365473 in neutron "Unable to create a router that's both HA and distributed" [High,In progress] - Assigned to Adolfo Duarte (adolfo-duarte) 15:16:43 I'm a little nervous on this one as I don't see any activity on the bug in the last week 15:17:02 lastly, there is http://bugs.launchpad.net/neutron/+bug/1407889 15:17:02 Launchpad bug 1407889 in Percona Server 5.6 "Default config contains deprecated directives" [Medium,Triaged] 15:17:09 regXboi: Can you clarify which patch set needs review for 1369721? A quick glance through is confusing. 15:17:12 that's wrong 15:17:22 carl_badlwind 15:17:31 ack - my fingers aren't working 15:17:42 regXboi: regarding 1365473, we could pin Adolfo. I can do it if you want 15:17:54 carl_baldwin: the patchset for 1369721 is the last one in the bug 15:18:03 mlavalle: please 15:18:09 ok, will do 15:18:37 carl_baldwin: the patch set for 1369721 is https://review.openstack.org/#/c/122024/ 15:18:53 the other patch set has been -2'd and can be abandoned as not being necessary 15:19:17 regXboi: thanks. Which is the one that has been -2d? 15:19:50 looking 15:20:14 https://review.openstack.org/#/c/123815/ 15:20:17 carl_baldwin: ^^ 15:20:30 thanks, obondarev 15:21:09 obondarev: thanks. 15:21:10 as https://review.openstack.org/#/c/211939 should cover agent side changes needed to handle rescheduling 15:21:42 obondarev is correct - so we can abandon 123815 in favor of 122024 and 211939 15:21:53 I can plan to begin reviewing these today. 15:21:59 carl_baldwin: thanks 15:22:00 obondarev: Do you want to abandon that one? 15:22:02 carl_baldwin: thanks 15:22:15 carl_baldwin: I can 15:22:51 #action mlavalle to ping assignee of bug 1365473 15:22:51 bug 1365473 in neutron "Unable to create a router that's both HA and distributed" [High,In progress] https://launchpad.net/bugs/1365473 - Assigned to Adolfo Duarte (adolfo-duarte) 15:23:12 obondarev: can you provide an update on https://bugs.launchpad.net/neutron/+bug/1470889 ? 15:23:12 Launchpad bug 1470889 in neutron "DVR: should not remove routers form down l3 agents on compute nodes" [High,In progress] - Assigned to Oleg Bondarev (obondarev) 15:23:25 regXboi: sure 15:23:30 there is a patch on it 15:23:41 let me find a link 15:23:42 obondarev: patch #? the bug doesn't show that 15:23:58 https://review.openstack.org/#/c/198009/ 15:24:21 not sure why launchpad stopped showing patches in bugs lately 15:24:36 yeah, I don't know either - so I'm updating by hand as needed 15:24:50 regXboi: thanks 15:24:57 all: that patchset ^^^^ also needs review 15:25:02 and that's the high list from the backlog 15:25:28 carl_baldwin: I'm done - returning the floor 15:26:02 regXboi: Thanks for your work here. I feel the momentum building on this due to some new very capable hands. 15:26:11 Any other bugs? 15:26:21 carl_baldwin: that's it from me 15:26:26 carl_badlwin: thanks 15:27:15 #topic Routed network segments 15:27:45 Nothing new here. I still have the action item to write a new proposal. 15:27:53 I will devote tomorrow to that. 15:28:34 carl_baldwin: if you need help - ping me 15:28:48 I've got folks internally beating me up over this topic :) 15:29:36 Also, looking forward to the discussion next week that we discussed last week. 15:29:44 #topic BGP dynamic routing 15:29:50 tidwellr: vikram_ : hi 15:29:57 hi 15:29:58 hi 15:30:43 https://review.openstack.org/#/c/201621 15:31:10 start of the review chain 15:31:53 not ready to throw in the towel yet, feature freeze is still a couple weeks away, but getting concerned 15:32:28 tidwellr: understood. I will review it today and encourage others to jump in. 15:32:32 I'm getting bogged down in testing, every time I try testing a corner case I find something horribly wrong 15:32:53 tidwellr: Well, that’s progress. 15:32:55 :) 15:33:02 https://etherpad.openstack.org/p/bgp-dynamic-routing 15:33:16 Here we are maintaing our open items 15:33:41 again, the struggle is in identifying what routes should be advertised, DVR makes it a nightmare 15:34:16 vikram_: That looks like overview. 15:34:49 request initial glance for "https://review.openstack.org/#/c/207607/", "https://review.openstack.org/#/c/207625/", and "https://review.openstack.org/#/c/207635/" 15:34:51 tidwellr: Can we separate the dvr case from the simpler case? 15:35:06 I have already punted on DVR host routes 15:35:22 unfortunately, DVR is so intertwined that it's hard to ignore 15:36:34 I've made some good progress, but it has delayed RPC's, automated tests, etc. 15:37:15 the DB queries are over the hump, but we're now behind a little bit in terms of bringing the agent and server-sides together 15:38:40 tidwellr: Can we do anything to help you out? 15:40:03 it's a big ask, but reviews on what we have now would be helpful so we can address things before we start pushing final code 15:40:21 tidwellr:+1 15:40:23 we're going down to the wire on this 15:40:37 tidwellr: ack. Let’s get reviewers on it. I can start devoting more effort to reviews. 15:41:02 it's a big ask I know, this stuff is all still WIP 15:41:04 tidwellr:+1-baldwin: thanks 15:41:14 carl_baldwin: thanks 15:41:43 tidwellr: do you expect needing help on getting voting jobs to pass? 15:42:55 I have some tempest tests, unforatunately I've got a chicken and egg problem where we need to enable the svc plugin in the gate for them to be useful and even pass 15:43:26 I pulled them API tests out into their own review, but abandoned that 15:43:42 tidwellr: that isn't quite what I was asking - I was more asking about keeping the pipeline flowing on the current WiP patches 15:45:00 regXboi: I understand now, I haven't been blocked by it yet 15:45:31 tidwellr: I'm just noticing that the check pipeline isn't entirely happy with the patch set chain... 15:45:45 that's all 15:46:14 ok, the dsvm-functional tests are failing because of that chicken-egg problem I mentioned 15:46:26 yep 15:46:28 sorry, dsvm-api 15:46:35 we can take it off line 15:46:42 dsvm-functional is failing and I don't know why 15:46:48 offline 15:46:52 I’m going to jump ahead a bit in the agenda. I trust tidwellr, regXboi, vikram_, and others can connect out of band. 15:47:00 carry on 15:47:15 Thanks, guys. Let’s get some attention on this since it is new functionality and must make Liberty-3 to be considered. 15:47:19 #topic DNS 15:47:22 mlavalle: hi 15:47:35 carl_baldwin: https://review.openstack.org/#/c/200952 got a -1 from armax 15:47:53 there is some debate as to why dns_name in ports was not implemented as an extension 15:48:10 mlavalle: well, calling me out is unfair, you got 4 -1! 15:48:12 the help i need here is for you weigh in herehttps://review.openstack.org/#/c/200952/12/neutron/api/v2/attributes.py 15:48:13 :) 15:48:38 armax: sorry for that, I am just trying to be quick 15:48:57 mlavalle: no worries…IMO this should be easy to fix 15:49:08 mlavalle: I gave you a pointer on how to do it 15:49:11 carl_baldwin: and then I will continue the conversation with armax and the others 15:49:14 armax: You just matter more than everyone else. ;) 15:49:31 carl_baldwin: that's right, armax is paramount 15:49:38 * armax blushes 15:50:32 carl_baldwin: so if you can give us your opinion in gerrit, i'll take it from there 15:50:56 carl_baldwin: I also pushed https://review.openstack.org/#/c/212213/ for the external dns 15:51:03 mlavalle: Are you asking specifically about doing an extension? 15:51:13 carl_baldwin: yes 15:51:41 I'd like some guidance as to how to move forward. whichever way you think is best to move forward 15:52:01 We should probably do some sort of an extension. I should have thought of that before given I’ve run in to this before. 15:52:28 carl_baldwin: ok, if you are time constrained, that's enough. i'll take it from here 15:52:34 mlavalle: Thanks. 15:52:41 as for https://review.openstack.org/#/c/212213/ 15:53:04 I mostly pushed it because I want early feedback from johnbelamaric on the driver interface 15:53:30 mlavalle: ok, I will look at it today 15:53:34 I will popuplate that interface today and then ask for his feedback, maybe tomorrow morning 15:53:43 mlavalle: What do you think our realistic changes are for getting this part done in Liberty? 15:53:51 mlavalle: I am on PTO next week so have to do it soon :) 15:53:51 … the second part I mean. 15:54:10 carl_baldwin: I think we have a shot. let's revisit this next week 15:54:28 I am going to be working full throttle on this over the next few days 15:54:47 I've asked my boss to remove any barriers between now and the end of the month 15:55:07 mlavalle: Okay. Let’s be sure to get the first part in. 15:55:15 of course :-) 15:55:38 so that's all I have 15:56:16 mlavalle: Thanks. 15:56:19 #topic Address Scopes 15:56:57 I have finished the functionality in the L3 agent that I wanted to get to. But, I have to admit, I’m in over my head to get this tested in the next three weeks with my travel schedule and other things. 15:57:21 I can have my work posted by the end of this week. Is anyone interested in lending a hand? 15:58:03 vikram_ and numan are probably the most familiar with it at this point. 15:59:10 Please let me know if you can lend a hand. 15:59:15 carl_baldwin, is there a review link ? 15:59:17 #topic Open Discussion 15:59:22 I've got one thing 15:59:39 #link http://goo.gl/Y7LIYX graphite link for multinode full and DVR jobs 15:59:48 numan: Look at the address-scopes gerrit topic. I will have everything posted as WIP before the end of tomorrow. 16:00:17 regXboi: nice. 16:00:18 there is a pending infra WiP patch for replacing -full with mutlinode-full already 16:00:21 We’ve got to end the meeting. 16:00:29 I'll be pushing another one for DVR next week 16:00:30 thanks 16:00:31 that's it 16:00:34 regXboi: ack Let’s move to the neutron room if needed. 16:00:36 #endmeeting