| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements https://review.openstack.org/483522 | 00:42 |
|---|---|---|
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork master: Updated from global requirements https://review.openstack.org/484555 | 00:45 |
| *** kiennt has joined #openstack-kuryr | 00:47 | |
| *** hongbin has joined #openstack-kuryr | 01:20 | |
| *** caowei has joined #openstack-kuryr | 01:56 | |
| *** longfei_zhang has joined #openstack-kuryr | 02:17 | |
| longfei_zhang | irena, do you know what these codes do? | 02:18 |
| longfei_zhang | https://github.com/openstack/kuryr-kubernetes/blob/master/kuryr_kubernetes/cni/binding/bridge.py#L35-L37 | 02:18 |
| *** hongbin has quit IRC | 04:01 | |
| *** limao has joined #openstack-kuryr | 05:08 | |
| *** janki has joined #openstack-kuryr | 05:15 | |
| *** caowei has quit IRC | 05:23 | |
| *** kiennt has quit IRC | 05:37 | |
| *** kiennt has joined #openstack-kuryr | 06:05 | |
| *** caowei has joined #openstack-kuryr | 06:10 | |
| *** Guest14 has joined #openstack-kuryr | 06:10 | |
| *** kzaitsev_ws has joined #openstack-kuryr | 06:17 | |
| *** Guest14 has quit IRC | 06:17 | |
| *** kiennt has quit IRC | 06:19 | |
| *** yedongcan has joined #openstack-kuryr | 06:27 | |
| *** vikasc has quit IRC | 06:34 | |
| *** vikasc has joined #openstack-kuryr | 06:49 | |
| apuimedo | longfei_zhang: irenab is on holidya | 07:02 |
| apuimedo | *holiday | 07:02 |
| *** aojea has joined #openstack-kuryr | 07:06 | |
| apuimedo | longfei_zhang: that codes moves it to the netns IIRC | 07:08 |
| *** aojea has quit IRC | 07:10 | |
| longfei_zhang | apuimedo, could you please give me more details about that code | 07:14 |
| apuimedo | longfei_zhang: sure. In the previous block it creates a veth pair | 07:14 |
| apuimedo | and on that block, it moves one side inside the network namespace of the pod | 07:14 |
| longfei_zhang | yes | 07:14 |
| *** kiennt has joined #openstack-kuryr | 07:15 | |
| *** ashishb_ has quit IRC | 07:15 | |
| longfei_zhang | thank you very much | 07:15 |
| longfei_zhang | we found some errors " NetlinkError: (17, 'File exists') " | 07:16 |
| *** kiennt_ has joined #openstack-kuryr | 07:18 | |
| longfei_zhang | we add some fix like this http://paste.openstack.org/show/617638/ | 07:19 |
| *** aojea has joined #openstack-kuryr | 07:19 | |
| longfei_zhang | we check it first | 07:20 |
| longfei_zhang | to avoid the NetlinkError | 07:20 |
| *** kiennt has quit IRC | 07:20 | |
| apuimedo | longfei_zhang: could you report it as a bug so I can get to it later? | 07:24 |
| apuimedo | and thanks for unearthing this | 07:24 |
| longfei_zhang | sure | 07:24 |
| *** aojea has quit IRC | 07:25 | |
| longfei_zhang | apuimedo, i got the reply from our k8s team member , they tested we can use the same subnet for service and pod , sorry for the delay | 07:26 |
| apuimedo | longfei_zhang: perfect | 07:26 |
| apuimedo | there's going to be another way to do it as well, if you want to use octavia L2 connectivity mode | 07:26 |
| apuimedo | (which we don't support yet, but we could add) | 07:27 |
| longfei_zhang | apuimedo, we are thinking about one thing recently , due to the performance of LB, we plan to use Kuryr only monitor pod events | 07:30 |
| longfei_zhang | use the default K8S kube-proxy to provide the LB func | 07:30 |
| apuimedo | longfei_zhang: so the performance is not good enough even when sharing the subnet? | 07:30 |
| longfei_zhang | yes, usually we may have thousands of lbs instance, so the network node's pressure is very high | 07:32 |
| longfei_zhang | we have some concerns about that | 07:33 |
| apuimedo | right | 07:34 |
| apuimedo | longfei_zhang: to address this I'm trying to get the OVN people to expose their native l4 load balancer to networking-ovn | 07:34 |
| apuimedo | so if you use ovn and this was exposed the performance would be much, much better | 07:35 |
| apuimedo | running thousands of haproxys is naturally a concern | 07:35 |
| apuimedo | longfei_zhang: are you doing baremetal or pod-in-vm scenario? | 07:35 |
| longfei_zhang | pod-in-vm | 07:36 |
| longfei_zhang | one more thing is we found there r some special k8s service that kuryr can not support it for now | 07:36 |
| apuimedo | longfei_zhang: we only support clusterip | 07:37 |
| apuimedo | but yboaron is adding loadbalancer type | 07:38 |
| apuimedo | probably next week we'll merge that | 07:38 |
| longfei_zhang | like one kind of service that the ClusterIP is defined to None in yaml ,another case is for the kubenetes service there is no selector defined for it, kuryr will skip it | 07:38 |
| apuimedo | longfei_zhang: what is the clusterIP defined to 'None' used for? | 07:42 |
| apuimedo | longfei_zhang: if there is no selector, what do you want a LB for? | 07:42 |
| apuimedo | we maybe missed something, how do you think the behavior should be? | 07:42 |
| *** aojea has joined #openstack-kuryr | 07:54 | |
| *** aojea has quit IRC | 08:19 | |
| *** aojea_ has joined #openstack-kuryr | 08:19 | |
| *** aojea_ has quit IRC | 08:29 | |
| *** aojea has joined #openstack-kuryr | 08:29 | |
| *** aojea has quit IRC | 08:39 | |
| *** aojea has joined #openstack-kuryr | 08:47 | |
| *** kiennt_ is now known as kiennt | 08:48 | |
| *** c00281451__ is now known as zengchen | 08:52 | |
| *** Guest14 has joined #openstack-kuryr | 09:16 | |
| *** limao has quit IRC | 09:23 | |
| *** limao has joined #openstack-kuryr | 09:23 | |
| *** aojea has quit IRC | 09:26 | |
| *** limao has quit IRC | 09:28 | |
| *** yamamoto_ has quit IRC | 09:44 | |
| *** caowei has quit IRC | 10:02 | |
| *** yamamoto has joined #openstack-kuryr | 10:02 | |
| *** kiennt has quit IRC | 10:09 | |
| apuimedo | vikasc: ping | 10:10 |
| vikasc | apuimedo, hi | 10:10 |
| apuimedo | vikasc: ptal https://review.openstack.org/490793 | 10:10 |
| vikasc | apuimedo, done | 10:12 |
| *** yamamoto has quit IRC | 10:16 | |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Replace six.itervalues() with dict.values() in kuryr-kubernetes https://review.openstack.org/490793 | 10:39 |
| *** Guest14 is now known as ajo | 10:43 | |
| *** aojea has joined #openstack-kuryr | 10:57 | |
| *** aojea has quit IRC | 10:59 | |
| *** aojea has joined #openstack-kuryr | 11:00 | |
| *** yamamoto has joined #openstack-kuryr | 11:13 | |
| *** atoth has joined #openstack-kuryr | 11:42 | |
| *** limao has joined #openstack-kuryr | 12:12 | |
| *** limao has quit IRC | 12:16 | |
| *** limao has joined #openstack-kuryr | 12:19 | |
| *** limao has quit IRC | 12:20 | |
| *** limao has joined #openstack-kuryr | 12:21 | |
| *** limao_ has joined #openstack-kuryr | 12:28 | |
| *** limao has quit IRC | 12:29 | |
| *** aojea_ has joined #openstack-kuryr | 12:38 | |
| *** aojea has quit IRC | 12:40 | |
| *** longfei_zhang has quit IRC | 12:52 | |
| *** aojea_ has quit IRC | 13:55 | |
| *** aojea has joined #openstack-kuryr | 13:55 | |
| *** ajo has quit IRC | 14:03 | |
| *** Guest14 has joined #openstack-kuryr | 14:04 | |
| *** kural has joined #openstack-kuryr | 14:05 | |
| *** kural_ has joined #openstack-kuryr | 14:05 | |
| *** Guest14 has quit IRC | 14:09 | |
| *** zengchen1 has joined #openstack-kuryr | 14:10 | |
| *** aojea has quit IRC | 14:11 | |
| *** yedongcan has left #openstack-kuryr | 14:11 | |
| *** aojea has joined #openstack-kuryr | 14:25 | |
| *** Guest14 has joined #openstack-kuryr | 14:39 | |
| *** Guest14 is now known as ajo | 14:40 | |
| *** kural_ has quit IRC | 14:46 | |
| *** kural has quit IRC | 14:46 | |
| *** hongbin has joined #openstack-kuryr | 14:54 | |
| *** aojea has quit IRC | 14:56 | |
| *** limao_ has quit IRC | 15:12 | |
| *** limao has joined #openstack-kuryr | 15:12 | |
| *** zengchen1 has quit IRC | 15:14 | |
| *** limao has quit IRC | 15:14 | |
| *** limao has joined #openstack-kuryr | 15:22 | |
| *** limao_ has joined #openstack-kuryr | 15:25 | |
| *** limao has quit IRC | 15:27 | |
| *** limao_ has quit IRC | 15:29 | |
| *** aojea has joined #openstack-kuryr | 15:35 | |
| *** ajo has quit IRC | 15:53 | |
| *** yamamoto has quit IRC | 15:59 | |
| *** vikasc has quit IRC | 16:00 | |
| *** vikasc has joined #openstack-kuryr | 16:12 | |
| *** aojea has quit IRC | 16:20 | |
| *** kural_ has joined #openstack-kuryr | 16:47 | |
| *** kural has joined #openstack-kuryr | 16:47 | |
| *** kural has quit IRC | 16:48 | |
| *** kural_ has quit IRC | 16:48 | |
| *** janki has quit IRC | 16:51 | |
| *** yamamoto has joined #openstack-kuryr | 17:00 | |
| *** yamamoto has quit IRC | 17:04 | |
| *** pcaruana has quit IRC | 17:15 | |
| *** livelace2 has joined #openstack-kuryr | 17:19 | |
| livelace2 | Hello. Does anybody use kuryr in production ? :) | 17:20 |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-libnetwork master: Make the name of existing net tag unique https://review.openstack.org/490308 | 17:22 |
| *** hongbin has quit IRC | 17:25 | |
| *** hongbin has joined #openstack-kuryr | 17:26 | |
| *** hongbin has quit IRC | 17:50 | |
| *** gouthamr has joined #openstack-kuryr | 17:57 | |
| *** livelace2 has quit IRC | 18:20 | |
| *** atoth has quit IRC | 19:03 | |
| *** reedip_afk has quit IRC | 19:33 | |
| *** reedip_afk has joined #openstack-kuryr | 19:45 | |
| *** gouthamr has quit IRC | 20:17 | |
| *** gouthamr has joined #openstack-kuryr | 20:24 | |
| *** gouthamr has quit IRC | 20:36 | |
| *** yamamoto has joined #openstack-kuryr | 21:20 | |
| *** yamamoto has quit IRC | 21:25 | |
| *** yamamoto has joined #openstack-kuryr | 21:56 | |
| *** gouthamr has joined #openstack-kuryr | 22:29 | |
| *** hongbin has joined #openstack-kuryr | 22:35 | |
| *** aojea has joined #openstack-kuryr | 22:47 | |
| *** aojea has quit IRC | 22:56 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!