15:00:21 <mlavalle> #startmeeting neutron_l3
15:00:22 <openstack> Meeting started Thu Mar 16 15:00:21 2017 UTC and is due to finish in 60 minutes.  The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:25 <openstack> The meeting name has been set to 'neutron_l3'
15:00:56 <mlavalle> Hi
15:00:56 <haleyb> mlavalle: hi
15:01:33 <john-davidge> o/
15:01:33 <mlavalle> #topic Announcements
15:02:19 <mlavalle> Pike-1 is less that a month away
15:02:29 <mlavalle> #link https://releases.openstack.org/pike/schedule.html
15:03:15 <mlavalle> Any other announcements from the team?
15:04:11 <mlavalle> ok, moving on
15:04:20 <mlavalle> #topic Bugs
15:04:34 <mlavalle> First up is https://bugs.launchpad.net/neutron/+bug/1627424
15:04:34 <openstack> Launchpad bug 1627424 in neutron "FlushError on IPAllocation" [High,In progress] - Assigned to Miguel Lavalle (minsel)
15:04:59 <mlavalle> I spent some time this morning checking Kibana for more hits for this bug
15:06:01 <mlavalle> Unfortunatelly, the refactoring of delete_subnet and delete_network in the plugin (ml2 and DB) didn't eliminate the bug
15:06:31 <haleyb> :(
15:06:46 <mlavalle> The shape of what I found in the logs changed a bit
15:07:47 <mlavalle> But we still have a delete_subnet tripping with a DHCP port update operation.
15:08:10 <mlavalle> I'll continue analyzing after the end of this meeting
15:09:11 <mlavalle> any other comments?
15:10:49 <mlavalle> Next up is https://bugs.launchpad.net/neutron/+bug/1610483
15:10:49 <openstack> Launchpad bug 1610483 in neutron "Pluggable IPAM rollback mechanism is not robust" [High,In progress] - Assigned to Aliaksandr Dziarkach (aliaksandr-dziarkach)
15:11:22 <mlavalle> kevinbenton reviewed the proposed solution for this
15:12:11 <mlavalle> He didn't see much value in it. He goes as far as removing the rollback mechanism
15:12:46 <mlavalle> any other comments?
15:13:50 <mlavalle> ok, next up is https://bugs.launchpad.net/neutron/+bug/1627480
15:13:50 <openstack> Launchpad bug 1627480 in neutron "create_port can succeed without returning fixed_ips on all requested subnets" [High,Confirmed] - Assigned to James Anziano (janzian)
15:14:17 <janzian> I've been looking through a few different occurrences of this bug on Kibana
15:14:55 <janzian> For the most part, the window appears to be around 0.6 seconds from port creation to deletion to bug
15:15:44 <mlavalle> from what I saw during our meeting, it's related also the DHCP port creation or update, right?
15:15:49 <mlavalle> janzian: ^^^^
15:16:07 <janzian> I believe so, yes
15:16:24 <janzian> I'm still trying to figure out an optimal solution. The obvious answer is locking the relevant parts of the database during the process, but I feel like that's much too big a performance hit
15:16:46 <mlavalle> mhhh, before going there/////
15:17:51 <mlavalle> have you seen this failure being concurrent with an operation on the port's subnet, specifically deletion?
15:18:45 <janzian> Yes. Here's one of the examples I've been looking at: http://paste.openstack.org/show/602933/
15:19:46 <janzian> The subnet is deleted very soon after the DHCP port was requested to be created, causing the bug
15:20:21 <mlavalle> ok, so it is very similar to what I am seeing with the Flusherror bug
15:20:44 <mlavalle> let's look at this together
15:21:29 <janzian> Sounds good to me
15:21:37 <mlavalle> any other comments?
15:22:58 <mlavalle> ok, moving on
15:23:05 <mlavalle> https://bugs.launchpad.net/neutron/+bug/1509004
15:23:05 <openstack> Launchpad bug 1509004 in neutron ""test_dualnet_dhcp6_stateless_from_os" failures seen in the gate" [High,Confirmed]
15:23:29 <mlavalle> It seems this one is related to failures in the DVR job now
15:23:53 <mlavalle> which seems to have a failure reate of ~25%
15:24:19 * haleyb didn't know of any DVR issues
15:24:41 <mlavalle> maybe I misunderstood from the CI meeting
15:24:48 <mlavalle> I will go back to the logs
15:25:14 <haleyb> i've also been offline a few days :)
15:27:27 <mlavalle> haleyb: take a look here: http://eavesdrop.openstack.org/meetings/neutron_ci/2017/neutron_ci.2017-03-14-16.00.log.html#l-38
15:28:16 <mlavalle> Let's follow up after the meeting
15:28:33 <haleyb> sure
15:29:10 <mlavalle> any other bugs from the team?
15:29:49 <mlavalle> ok, moving on
15:30:04 <mlavalle> #topic Prefix Delegation
15:30:40 <mlavalle> A copuple of days ago we merged https://review.openstack.org/#/c/413137/
15:30:45 <mlavalle> couple^^^
15:31:15 <mlavalle> which I belive was the last PD related patchset up for review
15:31:52 <mlavalle> any other comments?
15:32:38 <mlavalle> ok, moving on
15:32:41 <haleyb> yes, that was the last one
15:32:46 <haleyb> sorry, distracted
15:33:00 <mlavalle> thanks
15:33:13 <mlavalle> #topic Routed Networks
15:33:46 <mlavalle> I had a brief irc chat with kevinbenton regarding the floating ip RFE last Friday
15:34:09 <mlavalle> He mentioned the drivers couldn't review the RFE last week
15:34:21 <mlavalle> But that they will give it a try this week
15:34:38 <mlavalle> I thought john-davidge might be interested on this :-)
15:34:52 <john-davidge> mlavalle: That's good news, thanks :)
15:35:08 <mlavalle> any other comments?
15:36:11 <mlavalle> #topic Open Agenda
15:36:27 <mlavalle> Any other topics that should be discussed here today?
15:37:17 * john-davidge will be on PTO next week
15:37:31 <mlavalle> john-davidge: enjoy!
15:37:47 <john-davidge> mlavalle: Thank you :) Much needed ski holiday
15:38:04 <mlavalle> ok, guys, thanks for attending
15:38:13 <mlavalle> have a nice rest of your days :-)
15:38:18 <mlavalle> #endmeeting