15:02:04 #startmeeting neutron_l3 15:02:05 Meeting started Thu Mar 2 15:02:04 2017 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:06 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:09 The meeting name has been set to 'neutron_l3' 15:02:13 o/ 15:02:17 hi 15:02:35 sorry for the delay, got distracted analyzing a bug 15:02:56 #topic Annoucements 15:02:59 * john-davidge is also currently distracted by an internal meeting 15:03:51 We had a succesful PTG last week in Atlanta. A lot was discussed 15:04:19 kevinbenton sent an email with the summary earlier this week 15:05:19 i heard there was some good dvr work, still have to digest all of that 15:06:16 #link http://lists.openstack.org/pipermail/openstack-dev/2017-February/113032.html 15:07:02 any other annoucements? 15:08:52 ok, let's move on 15:08:58 #topic Bugs 15:10:07 First up is https://bugs.launchpad.net/neutron/+bug/1627424 15:10:07 Launchpad bug 1627424 in neutron "FlushError on IPAllocation" [High,In progress] - Assigned to Miguel Lavalle (minsel) 15:10:35 I have been reviewing a couple of patchsets by kevinbenton 15:11:23 They refactor the the way ml2 and the db plugin delete networks and subnets 15:11:44 and as a consequence, the way network owned ports are deleted 15:12:19 we are going to wait for those patchsets to merge and re-examine how this bug is affected 15:13:02 The patchsets in question are: 15:13:09 #link https://review.openstack.org/#/c/423584 15:13:31 #link https://review.openstack.org/#/c/428774 15:13:38 chime in if you can 15:14:01 i will take another look today 15:14:18 i added them to the etherpad under the bug as well 15:14:47 Thanks haleyb! 15:16:34 Next up is https://bugs.launchpad.net/neutron/+bug/1627480 15:16:34 Launchpad bug 1627480 in neutron "create_port can succeed without returning fixed_ips on all requested subnets" [High,Confirmed] - Assigned to James Anziano (janzian) 15:16:53 janzian is off on vacation this week 15:17:18 But I spent some time with him taking a look at this bug the Friday before the PTG 15:17:49 And earlier today I spent some time digging in Kibana (that's how I got late to this meeting) 15:18:03 * john-davidge is now giving this meeting his full attention :) 15:18:53 It has to do with ports being created on subnets being deleted. So it's starting to look to me very similar to the previous bug 15:19:19 I will dig a little bit more over the next few days before reaching a final conclusion, though 15:20:02 any questions, suggestions? 15:20:54 Next up is https://bugs.launchpad.net/neutron/+bug/1610483 15:20:54 Launchpad bug 1610483 in neutron "Pluggable IPAM rollback mechanism is not robust" [High,In progress] - Assigned to Aliaksandr Dziarkach (aliaksandr-dziarkach) 15:22:03 there were some comment by ihar yesterday, but he didn't -1, think they can be addressed though 15:22:12 ihrachys reviewed the proposed fix earlier this week 15:22:26 #link https://review.openstack.org/#/c/390594/ 15:23:33 he seems to be happy overall with it. He suggested a couple of changes and add some more testing 15:24:10 with the review that haleyb did a couple of weeks ago, it seems that we are converging to a solution 15:24:14 any comments? 15:24:56 i will run a quick test with that test back and see what happens 15:25:47 cool, thanks haleyb 15:26:10 might just go :) 15:26:29 hopefully not :-) 15:27:39 Last one today is https://bugs.launchpad.net/neutron/+bug/1509004 15:27:39 Launchpad bug 1509004 in neutron ""test_dualnet_dhcp6_stateless_from_os" failures seen in the gate" [High,Confirmed] 15:28:19 that still fails occasionally, just looked in kibana, ~1-2 a day 15:28:37 usually in multi-node jobs 15:29:50 although a lot is in one of the non-voting jobs 15:30:15 cool, let's keep an eye on it. thanks haleyb 15:30:27 don't have much to add besides that, seems more noise now than critical 15:31:00 great. that's good to know 15:31:30 21 in last 30 days 15:32:01 keep in mind we only keep 10 days of data 15:32:06 so its 21 in 10 15:32:21 clarkb: ahhh, that's good to know 15:32:48 clarkb: so if I query kibana for the last 30 days, it is really only showing the last 10 days 15:32:55 yes 15:32:56 ah, then i am corrected, always just widen to 30 15:34:31 any other bugs? 15:35:28 ok, moving on 15:35:40 #topic Prefix Delegation 15:35:52 I don't see baoli on 15:36:14 I have a minor update to mention on PD testing 15:36:28 john-davidge: go ahead 15:36:49 Had a conversation with the openstack ansible team at the PTG about doing some end-to-end testing in their gate where we can easily install and configure the dibbler server 15:37:20 A couple of people here at Rackspace are ding the preliminary work on that now 15:37:35 hopefully they can join this meeting in the coming weeks to give progress updates 15:37:37 thats it 15:37:49 *doing 15:38:00 that's great! 15:38:26 john-davidge: that would be great since would give us some functional testing potential 15:38:55 haleyb: Yeah that's the hope. PD has suffered a large number of regressions over the last few cycles 15:40:21 any other comments? 15:41:14 ok 15:41:23 #topic Routed Networks 15:41:41 john-davidge: do you want to comment on the progress made in Atlanta? 15:41:49 mlavalle: Sure 15:42:59 So mlavalle and I had a productive conversation about putting the plan onto paper. The result was this RFE https://bugs.launchpad.net/neutron/+bug/1667329 15:42:59 Launchpad bug 1667329 in neutron "[RFE] Floating IP Subnets on Routed Provider Networks" [Undecided,Triaged] 15:43:33 It's not a long read, but the summary is use Service Subnets to allow FIP subnets on routed networks 15:43:48 that's cool 15:44:06 Then use BGP Dynamic Routing to advertise the ToR as the next hop depending on the segment where the instance lives 15:44:17 Hopefully it will be discussed in the drivers meeting today 15:44:33 yeah, that is my expectation 15:44:48 I'm ready to get started with the implementation as soon as its approved 15:44:53 and/or swrite a spec if needed 15:45:07 me too 15:45:54 mlavalle: Excellent :) 15:46:23 anything else? 15:46:47 not from me 15:47:12 thanks for the update john-davidge 15:47:22 #topic Open Agenda 15:47:43 any other topics we should discuss today? 15:48:10 I added this to the agenda: https://review.openstack.org/#/c/433234 15:48:24 Creating subnets without an allocation pool 15:48:45 it could use some review attention so we can agree on the format of the api extension 15:49:04 thanks john-davidge. I commited to take a look at this patchset before flying to Mexico City 15:49:22 I will do it today or tomorrow 15:49:46 mlavalle: Make sure you send me a Margharita in the mail 15:49:53 mlavalle: I'll send you my postal address 15:50:05 will do :-) 15:50:49 or just some "Mexican water" as our waiter called it :) 15:51:05 that's pretty good 15:51:27 ok guys. Thanks for attending :-) 15:51:46 Daughter getting married this coming Saturday. Yaay! 15:51:57 mlavalle: Have a great time!! 15:52:07 #endmeeting