13:59:32 <liuyulong> #startmeeting neutron_l3 13:59:33 <openstack> Meeting started Wed Feb 5 13:59:32 2020 UTC and is due to finish in 60 minutes. The chair is liuyulong. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:59:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:59:36 <openstack> The meeting name has been set to 'neutron_l3' 14:00:26 <liuyulong> Hi there 14:02:18 <liuyulong> #topic Announcements 14:04:08 <liuyulong> #link https://releases.openstack.org/ussuri/schedule.html 14:04:20 <liuyulong> Next week is U-2. 14:06:24 <liuyulong> OK, let's move on. 14:06:28 <liuyulong> #topic Bugs 14:06:42 <liuyulong> #link http://lists.openstack.org/pipermail/openstack-discuss/2020-February/012332.html 14:08:16 <liuyulong> First one: 14:08:19 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1861674 14:08:19 <openstack> Launchpad bug 1861674 in neutron "Gateway which is not in subnet CIDR is unsupported in ha router" [Undecided,In progress] - Assigned to Zhengdong Wu (zhengdong.wu) 14:08:32 <slaweq> hi, sorry for being late 14:09:55 <liuyulong> slaweq, hi 14:10:52 <liuyulong> From the response of the bug reporter, the HA router will lose one route rule in the namespace comparing to none-HA router. 14:14:07 <liuyulong> So the test could be something like this: if the external network has subnet with IP range(CIDR) which does not include the gateway IP address. 14:16:15 <liuyulong> The HA router namespace will lose the route rule to gateway IP. 14:16:45 <liuyulong> And then floating IP or SNAT traffic may not work. 14:19:08 <liuyulong> Here is my test: 14:19:09 <liuyulong> openstack network create --external public-test-1 14:19:15 <liuyulong> openstack subnet create --subnet-range 192.168.5.128/25 --gateway 192.168.4.129 --network public-test-1 public-test-subnet-1 14:19:40 <liuyulong> According to the information of the bug 14:25:35 <liuyulong> OK, we should do some more test in detail. 14:25:48 <liuyulong> Let's move one. 14:27:19 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1861032 14:27:19 <openstack> Launchpad bug 1861032 in neutron "[RFE] Add support for configuring dnsmasq with multiple IPv6 addresses in same subnet on same port" [Undecided,In progress] - Assigned to Harald Jensås (harald-jensas) 14:27:44 <liuyulong> And its related rfe 14:27:46 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1855854 14:27:47 <openstack> Launchpad bug 1855854 in neutron "[RFE] Dynamic DHCP allocation pool" [Wishlist,New] 14:27:57 * haleyb is really late 14:28:58 <liuyulong> IMO these IPv6 related requests make sense to me. 14:30:21 <haleyb> when i talked to harald yesterday, he was waiting for a response from the dnsmasq maintainer on a proposed patch to make this possible 14:30:41 <haleyb> without that it would be a large change to neutron to support multiple ipv6 addresses per port 14:31:25 <liuyulong> #link http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2020q1/013743.html 14:31:34 <liuyulong> haleyb, this patch right? 14:31:40 <slaweq> there is new rfe proposed related to that one https://bugs.launchpad.net/neutron/+bug/1861032 14:31:40 <openstack> Launchpad bug 1861032 in neutron "[RFE] Add support for configuring dnsmasq with multiple IPv6 addresses in same subnet on same port" [Undecided,In progress] - Assigned to Harald Jensås (harald-jensas) 14:32:46 <haleyb> liuyulong: no, there was another one, and simon responded today saying he was going to update 14:33:41 <liuyulong> http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2020q1/013671.html 14:33:45 <haleyb> http://lists.thekelleys.org.uk/pipermail/dnsmasq-discuss/2020q1/013763.html 14:33:51 <liuyulong> [Dnsmasq-discuss] [PATCH] DHCPv6 - Multiple reservations for single host 14:34:14 <liuyulong> Yep 14:34:42 <haleyb> either way, i'd defer to harald, and he needs to wait on dnsmasq to see what work is needed in neutorn 14:36:18 <liuyulong> This is interesting now, neutron is trying to request its dependency to make something 'new'. 14:38:48 <liuyulong> OK, we can continue the discuss in the LP. 14:39:23 <liuyulong> Next one is related to OVN: 14:39:26 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1861502 14:39:26 <openstack> Launchpad bug 1861502 in neutron "[OVN] Mechanism driver - failing to recreate floating IP" [High,Triaged] - Assigned to Maciej Jozefczyk (maciej.jozefczyk) 14:40:28 <liuyulong> According to the LOG trace in the description, looks like a function param missing. 14:41:42 <haleyb> maciejjozefczyk: are you working on that or need someone to take it? 14:41:42 <liuyulong> Next also related to OVN 14:42:22 <maciejjozefczyk> haleyb, I need to take PTO on next day. If someone will have a minute to take a look, I'm ok with that :) 14:43:05 <haleyb> maciejjozefczyk: thanks, i'll see if it's just a simple change 14:43:30 <maciejjozefczyk> haleyb, this is realted only to maintanance task, perhaps it needs an additional unittest 14:45:41 <haleyb> maciejjozefczyk: ack 14:46:01 <liuyulong> It could be a easy one, " ) 14:46:10 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1861510 14:46:10 <openstack> Launchpad bug 1861510 in neutron "[OVN] GW rescheduling mechanism is triggered on every Chassis updated unnecessarily" [High,In progress] - Assigned to Maciej Jozefczyk (maciej.jozefczyk) 14:47:08 <liuyulong> This one could be related to the gateway rescheduling performance. 14:47:33 <liuyulong> Looks like a "thundering herd". 14:47:40 <maciejjozefczyk> liuyulong, yes, Im working on it right now. 14:48:34 <liuyulong> maciejjozefczyk, cool 14:48:37 <liuyulong> #link https://review.opendev.org/#/c/705660/ 14:48:43 <liuyulong> ^ this is the patch. 14:49:10 <maciejjozefczyk> liuyulong, The fix is about: 1) filtering unrelated events that are generated by ovn sbdb 14:49:45 <maciejjozefczyk> 2) verify and pass only needed logical router ports to be rescheduled 14:49:46 <liuyulong> Maybe test cases should be added for it. : ) 14:49:59 <maciejjozefczyk> liuyulong, actually I'm workign on unit/functional tests now :) 14:51:47 <liuyulong> Honestly, these OVN DB related changes are new to me, I may need sometime to learn it. : ) 14:52:43 <liuyulong> I'm in the reviewer list now 14:53:18 <liuyulong> So is this worthy to do the stable branch backport? 14:53:51 <maciejjozefczyk> liuyulong, Yes, that is needed. We found that actual solution could lead to kind of 'storm' of events. 14:54:05 <liuyulong> OK 14:54:32 <liuyulong> But this could be a cross project backport now. 14:55:02 <maciejjozefczyk> liuyulong, yes, We would need to backport it to networking-ovn. 14:55:54 <liuyulong> Next one: 14:55:57 <liuyulong> #link https://bugs.launchpad.net/neutron/+bug/1861509 14:55:57 <openstack> Launchpad bug 1861509 in neutron "[OVN] GW rescheduling logic is broken " [High,In progress] - Assigned to Maciej Jozefczyk (maciej.jozefczyk) 14:56:20 <liuyulong> maciejjozefczyk, This is only in neutron right? 14:56:48 <liuyulong> networking-ovn does not have such issue? 14:57:12 <maciejjozefczyk> The fix that I'm proposing in previous patch also is related to the bug 1861509. Those need to be pushed together. 14:57:12 <openstack> bug 1861509 in neutron "[OVN] GW rescheduling logic is broken " [High,In progress] https://launchpad.net/bugs/1861509 - Assigned to Maciej Jozefczyk (maciej.jozefczyk) 14:57:23 <maciejjozefczyk> liuyulong, networking-ovn also has this issue. 14:57:40 <maciejjozefczyk> I added two closes-bug tags to 14:57:42 <maciejjozefczyk> #link https://review.opendev.org/#/c/705660/ 14:58:00 <liuyulong> Yes, it is pointing to same fix 14:58:18 <maciejjozefczyk> Both needs to be then cherry-picked to stable branches. 14:58:54 <liuyulong> Sure 14:59:01 <liuyulong> We are out of time. 14:59:09 <liuyulong> Let's end here. 14:59:16 <liuyulong> #endmeeting