15:03:37 <haleyb> #startmeeting neutron_dvr 15:03:38 <openstack> Meeting started Wed Feb 8 15:03:37 2017 UTC and is due to finish in 60 minutes. The chair is haleyb. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:03:39 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:03:42 <openstack> The meeting name has been set to 'neutron_dvr' 15:03:45 <haleyb> #chair Swami 15:03:46 <openstack> Current chairs: Swami haleyb 15:03:51 <haleyb> #topic Announcements 15:03:54 <Swami> may be then it would be a short meeting. 15:04:18 <haleyb> yes, maybe just the two of us :( 15:05:05 <haleyb> RC1 was cut, RC2 open for critical things, not sure we have anything on the list for it 15:05:31 <Swami> haleyb: at this point we don't have any. 15:05:47 <haleyb> also, master is essentially pike now, with rc2 code picked to stable/ocata 15:06:02 <Swami> Is pike already open 15:07:02 <haleyb> unless i mis-understood, master is open, but we're trying not to merge too much until we can figure out the memory consumption issue 15:07:19 <Swami> haleyb: understood. 15:07:37 <haleyb> #topic Bugs 15:07:44 <ravichandran> Hello 15:08:05 <Swami> ravichandran: hi 15:08:09 <Swami> haleyb: thanks 15:08:21 <Swami> Regarding bugs, there are no new reported bugs for this week. 15:08:48 <Swami> There was one bug that was reported last week in newton for floatingip not working on a OS reboot. 15:09:07 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1657981 15:09:07 <openstack> Launchpad bug 1657981 in neutron "FloatingIPs not reachable after restart of compute node (DVR)" [Undecided,New] 15:09:28 <Swami> I was not able to reproduce it in the master branch. But will try to see if the issue is still seen in newton. 15:09:59 <haleyb> Swami: thanks, can you add that to the bug? 15:10:24 <Swami> Yes I will. 15:10:33 <haleyb> i just noticed the last comment shows he's running the vpn-agent 15:10:38 <Swami> I will add my findings in the bug. 15:11:05 <Swami> He also mentioned that it is not easy to reproduce, so I am not sure if we would be able to see it. 15:11:59 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1658060 - This bug I have not triaged it yet and will try to triage it this week. 15:11:59 <openstack> Launchpad bug 1658060 in neutron "FirewallNotFound exceptions when deleting the firewall in FWaaS-DVR" [Undecided,In progress] - Assigned to Yaohua Yan (sleepdream) 15:13:41 <Swami> The next one in the list is 15:13:45 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1647432 15:13:45 <openstack> Launchpad bug 1647432 in neutron "Multiple SIGHUPs to keepalived might trigger re-election" [High,In progress] - Assigned to Jakub Libosvar (libosvar) 15:14:05 <Swami> This patch needs review - #link https://review.openstack.org/#/c/407099/ 15:14:34 <haleyb> it was just updated this morning, i'll look again 15:14:54 <Swami> haleyb: ok thanks 15:15:42 <Swami> The next in the list is 15:15:45 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1644231 15:15:45 <openstack> Launchpad bug 1644231 in neutron "fip router config is not created if the vm ports attached to FIPs have no device_owner" [Low,Triaged] - Assigned to Brian Haley (brian-haley) 15:16:08 <Swami> #link https://review.openstack.org/#/c/425919/ - This patch is almost ready for merge, but probably needs worflow. 15:16:47 <haleyb> I just need to get another core to look at that one, armando should have some free time now :) 15:17:00 <Swami> haleyb: ok thanks 15:17:35 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1629539 15:17:35 <openstack> Launchpad bug 1629539 in neutron "Broken distributed virtual router w/ lbaas v1" [Undecided,Expired] 15:18:43 <Swami> This is again reported against the LBaasv1 and still have to triage it. 15:19:13 <haleyb> Swami: the lbaas tag was removed from the bug, wondering if it's still valid 15:20:33 <haleyb> i'm fine with not working on it any more, unless there is something of interest 15:20:45 <Swami> haleyb: yes I did see that the tag was removed, but the bug description is still the same. 15:20:56 <Swami> haleyb: ok no problem 15:21:24 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1577488 15:21:24 <openstack> Launchpad bug 1577488 in neutron "[RFE]"Fast exit" for compute node egress flows when using DVR" [Wishlist,In progress] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 15:21:53 <Swami> Since pike is open, once the memory consumption issues settles down, we should try to merge these two patches. 15:22:03 <Swami> #link https://review.openstack.org/#/c/283757/ 15:22:14 <Swami> #link https://review.openstack.org/#/c/355062/ 15:22:34 <haleyb> Yes, they would be first on our list 15:23:24 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1583694 15:23:24 <openstack> Launchpad bug 1583694 in neutron "[RFE] DVR support for Allowed_address_pair port that are bound to multiple ACTIVE VM ports" [Wishlist,In progress] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) 15:23:37 <Swami> haleyb: I have a quesstion on these patches. 15:23:46 <Swami> #link https://review.openstack.org/#/c/320669/ 15:24:03 <Swami> #link https://review.openstack.org/#/c/323618/ 15:24:33 <Swami> The previous solution that we had to address the unbound ports with floatingip is on its way while try to address this issue. 15:25:14 <Swami> haleyb: my question to you is can we change that behavior and by default go with this approach or will it be an issue if someone is using it in its current behavior. 15:26:16 <Swami> haleyb: we used to inherit the parent port properties of the device_owner and host_id for the allowed_address_pair, but that did not satisfy the Octavia and other regular VRRP use cases. 15:27:00 <haleyb> Swami: just so i'm clear, today this is done in fip namespace, but not fast enough, right? we were looking at centralizing it 15:27:41 <Swami> Yes, in order to do it in fip namespace we inherit the properties of the parent VM port for the allowed_address_pair port and also we only do it when the port is in active state. 15:28:13 <Swami> We do not support active-active configuration for the allowed_address_pairs attached VMs. 15:28:28 <Swami> So it is kind of a narrow use case. 15:29:10 <haleyb> right. has octavia found a way around it, or is just living with it 15:30:33 <Swami_> sorry got disconnected. 15:31:02 <Swami_> Octavia currently does not work with DVR because of this issue. 15:32:12 <haleyb> Swami: is this something to just add to the PTG list? If we can't get approval we can't fix it 15:32:35 <Swami_> haleyb: yes I have already added this to the PTG list 15:32:52 <Swami_> haleyb: I can have a discussion in the PTG meeting. 15:33:36 <haleyb> Swami_: ok, thanks 15:33:51 <Swami_> #link https://etherpad.openstack.org/p/neutron-ptg-pike 15:34:04 <Swami_> Yes I have added a section under DVR to discuss about the unbound ports issue. 15:35:51 <Swami_> that's all I had for bugs today. 15:36:23 <haleyb> ok, thanka 15:36:59 <haleyb> is there anyone here working on HA that has additional bugs to talk about? 15:37:59 <haleyb> #topic Gate 15:38:48 <haleyb> I put the DVR-multinode job on the back burner until current memory issue is addresses, so it will be a Pike thing 15:39:22 <Swami_> haleyb: understood. 15:39:30 <haleyb> #topic Stable backports 15:40:08 <Swami_> haleyb: I may have spin this patch out of the dependency https://review.openstack.org/#/c/421989/ 15:40:09 <haleyb> Swami_: i think your recent change was proposed to stable/ocata? 15:40:35 <Swami_> haleyb: I also proposed a patch to stable/newton 15:40:50 <Swami_> #link https://review.openstack.org/#/c/430384/ 15:41:06 <Swami_> #link https://review.openstack.org/#/c/429985/ 15:41:14 <Swami_> The later one is for /stable/ocata 15:41:48 <Swami_> Ihar mentioned that there was some dvr failures and so he had put a -2 on that patch until the root cause is identified in stable/ocata. 15:41:53 <haleyb> yes, i hadn't added myself to that one 15:43:22 <haleyb> Swami_: you need help with anythign? https://review.openstack.org/#/c/421989/ for example? 15:44:06 <Swami_> haleyb: if you have some time you can churn this to get rid of the dependency and I can focus on the unbound ports validation. 15:44:45 <haleyb> Swami_: sure, i can try and rebase it to master instead of the dependent 15:45:20 <Swami_> haleyb: thanks 15:45:59 <haleyb> #topic Open Discussion 15:47:02 <haleyb> anyone have other topics? 15:47:37 <Swami> sorry got disconnected again, I am not sure if this is my wireless or the hexchat issue. 15:49:00 <haleyb> Swami: np, think we're done 15:49:12 <Swami> haleyb: I don't have any other topics today. 15:49:48 <haleyb> ok, thanks for getting up early Swami 15:49:50 <haleyb> #endmeeting