| *** celebdor has quit IRC | 00:04 | |
| *** threestrands has joined #openstack-kuryr | 00:18 | |
| *** tosky has quit IRC | 00:27 | |
| *** rh-jelabarre has joined #openstack-kuryr | 00:41 | |
| *** gianpietro has joined #openstack-kuryr | 02:11 | |
| *** gianpietro has quit IRC | 02:18 | |
| *** premsankar has quit IRC | 02:43 | |
| *** hongbin has joined #openstack-kuryr | 03:02 | |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Support network policy update https://review.openstack.org/605150 | 03:31 |
|---|---|---|
| *** hongbin has quit IRC | 03:56 | |
| *** janki has joined #openstack-kuryr | 04:03 | |
| *** threestrands has quit IRC | 05:20 | |
| *** pcaruana has quit IRC | 05:35 | |
| *** ccamposr has joined #openstack-kuryr | 07:07 | |
| *** pcaruana has joined #openstack-kuryr | 07:22 | |
| *** maysams has joined #openstack-kuryr | 07:51 | |
| *** celebdor has joined #openstack-kuryr | 08:13 | |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: Wait for the k8s api to be healthy https://review.openstack.org/618545 | 08:26 |
| *** tosky has joined #openstack-kuryr | 08:42 | |
| *** celebdor has quit IRC | 08:49 | |
| *** celebdor has joined #openstack-kuryr | 08:51 | |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: parametrize openshift version fetching https://review.openstack.org/619018 | 09:06 |
| celebdor | dulek: could it be that openshift doesn't like your https://review.openstack.org/#/c/620045/ | 09:07 |
| celebdor | ? | 09:07 |
| celebdor | ah no. It seems like flakiness | 09:08 |
| celebdor | :/ | 09:08 |
| celebdor | http://logs.openstack.org/45/620045/4/check/kuryr-kubernetes-tempest-daemon-openshift-octavia/378b9b2/testr_results.html.gz | 09:09 |
| celebdor | or maybe not | 09:09 |
| celebdor | http://paste.openstack.org/show/736468/ | 09:10 |
| dulek | celebdor: I'll look in a minute. | 09:11 |
| celebdor | ;-) | 09:11 |
| dulek | celebdor: Ha, how could I break namespace isolation? | 09:12 |
| dulek | ltomasbo: ? ^ | 09:12 |
| celebdor | dulek: yeah | 09:13 |
| celebdor | that's what I'm wondering | 09:14 |
| ltomasbo | I think that test is broken in openshift? | 09:14 |
| celebdor | *taking out pokéball* ltomasbo, I choose you | 09:14 |
| celebdor | ltomasbo: damn. Not the best platform to have isolation tests broken | 09:14 |
| ltomasbo | celebdor, not isolated test, just the svc one | 09:19 |
| ltomasbo | celebdor, it is the octavia sg thing... | 09:19 |
| celebdor | ltomasbo: this test is pod to pod | 09:20 |
| celebdor | not pod to service | 09:20 |
| ltomasbo | let me see, I may have look at a different one | 09:20 |
| celebdor | ltomasbo: test_namespace | 09:20 |
| celebdor | look for # check no connectivity from NS1 to NS2 | 09:20 |
| ltomasbo | I see this in your paste: in test_namespace_sg_svc_isolation | 09:20 |
| ltomasbo | celebdor, ahh, you meant that the part failing is on the pod to pod part | 09:22 |
| ltomasbo | ok, but the failure is weird... | 09:22 |
| ltomasbo | testtools.matchers._impl.MismatchError: 'kuryr-pod-615029703: HELLO! I AM ALIVE!!!\n' matches Contains('HELLO! I AM ALIVE!!!') | 09:22 |
| ltomasbo | why there is a mismatch there? | 09:22 |
| ltomasbo | celebdor, ^^ | 09:22 |
| ltomasbo | If I'm seeing it right, it does contain the string, right? | 09:22 |
| celebdor | ltomasbo: you are right | 09:24 |
| celebdor | It is that test | 09:24 |
| celebdor | and it is the octavia issue | 09:24 |
| celebdor | and of course, since we only test it in Octavia | 09:24 |
| celebdor | that's the only gate that fials | 09:24 |
| celebdor | s/Octavia/OpenShift/ | 09:25 |
| ltomasbo | yes, the line is checking that pod from NS1 cannot reach service on NS2, and that is not blocked at the moment | 09:25 |
| celebdor | yes, yes | 09:27 |
| celebdor | you are right | 09:27 |
| *** takamatsu has quit IRC | 09:41 | |
| *** pcaruana has quit IRC | 10:44 | |
| *** pcaruana has joined #openstack-kuryr | 10:50 | |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-tempest-plugin master: Enhance svc namespace isolation with LoadBalancer type svc https://review.openstack.org/618751 | 11:19 |
| openstackgerrit | Danil Golov proposed openstack/kuryr-kubernetes master: Introduce test case document for SRIOV functionality https://review.openstack.org/600022 | 11:20 |
| *** pcaruana has quit IRC | 12:11 | |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Add security groups driver for NP https://review.openstack.org/608616 | 12:40 |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Fix definition of CNIRunner.get_container_id() https://review.openstack.org/619996 | 12:40 |
| *** gianpietro has joined #openstack-kuryr | 13:36 | |
| *** takamatsu has joined #openstack-kuryr | 13:55 | |
| *** aperevalov has quit IRC | 14:21 | |
| celebdor | dulek: ltomasbo: dmellado: https://review.openstack.org/#/c/619018/ | 14:24 |
| dulek | celebdor: awk used as format operator? Do I see correctly? That's insane… | 14:26 |
| dulek | No, I don't understand what's going on there. :P | 14:27 |
| *** janki has quit IRC | 14:30 | |
| celebdor | dulek: that's just finding the line in checksum that defines the name of the server tarball | 14:31 |
| celebdor | and when it has it, it constructs the url to download the tarball | 14:31 |
| dulek | celebdor: Oh… | 14:31 |
| celebdor | cause they have a non predictable commit string there in the filename | 14:32 |
| celebdor | so that's the only way I knew how to find it, so that our devstack users only need to specify the v3.10.0 and not have to lookup the whole url | 14:32 |
| celebdor | I think it is quite convenient | 14:32 |
| celebdor | dulek: SonicSong is having some issues on devstack 16.04.3 with cni https://www.irccloud.com/pastebin/0TIEJjlE/ | 14:36 |
| SonicSong | Sorry, ubuntu 16.04.5 instead | 14:37 |
| dulek | SonicSong: Can you check out kuryr-daemon logs? | 14:38 |
| SonicSong | okay | 14:38 |
| dulek | SonicSong: It's either devstack@kuryr-daemon in systemd or a container on kube-system namespace. | 14:39 |
| dulek | SonicSong: Depending if you're trying to run Kuryr in containers. | 14:39 |
| SonicSong | i'm using systemd | 14:39 |
| SonicSong | https://www.irccloud.com/pastebin/es1uA352/ | 14:39 |
| SonicSong | here is a part of it | 14:40 |
| dulek | SonicSong: "Timed out waiting for requested pod to appear in registry" - okay, so seems like kuryr-controller is having troubles with something and doesn't annotate the pod. | 14:41 |
| dulek | SonicSong: It's devstack@kuryr-kubernetes. | 14:41 |
| dulek | (I know, we should rename it one day) | 14:41 |
| SonicSong | https://www.irccloud.com/pastebin/JjyxFT82/ | 14:42 |
| SonicSong | i think it is the key part of the error | 14:43 |
| SonicSong | it's from kuryr-controller | 14:43 |
| *** takamatsu has quit IRC | 14:48 | |
| *** gianpiet_ has joined #openstack-kuryr | 14:52 | |
| dulek | SonicSong: Hm, no idea at the moment. Seems like annotation is added correctly. | 14:53 |
| *** gianpie__ has joined #openstack-kuryr | 14:53 | |
| SonicSong | dulek: I'm curious about the VIFOpenVSwitch part | 14:53 |
| dulek | SonicSong: But that's for pod svc-vdu1-e493cd-b54fd7f95-6whrz. And previous error was about the same? | 14:54 |
| dulek | Yeah, that's the same. | 14:54 |
| dulek | SonicSong: What's wrong with that one? | 14:54 |
| *** gianpiet_ has quit IRC | 14:54 | |
| SonicSong | dulek: Seems that it's inactive? | 14:54 |
| dulek | SonicSong: Yeah, it will be inactive until kuryr-daemon plugs it. | 14:55 |
| *** gianpietro has quit IRC | 14:55 | |
| SonicSong | dulek: oh, I see | 14:55 |
| dulek | SonicSong: You can try restarting kuryr-daemon. Maybe it wasn't watching when pod was created, but that's rather unlikely. | 14:55 |
| SonicSong | I will have a try | 14:56 |
| SonicSong | dulex: it seems the same | 14:59 |
| SonicSong | oops...sorry | 14:59 |
| SonicSong | dulek: it seems the same | 14:59 |
| dulek | SonicSong: Can you mail me full logs? mdulko@redhat.com | 15:01 |
| SonicSong | dulek: sure, which part do you want | 15:01 |
| SonicSong | dulek: both kuryr-controller and daemon? | 15:02 |
| dulek | SonicSong: Yup! | 15:02 |
| SonicSong | okay | 15:02 |
| dulek | SonicSong: Maybe just kuryr-daemon first. :P | 15:02 |
| SonicSong | dulek: okay | 15:03 |
| dulek | SonicSong: Okay, I have a hypothesis, just a second… | 15:13 |
| dulek | SonicSong: Oh my, you've discovered a really sneaky bug. :D | 15:14 |
| dulek | SonicSong: kuryr-daemon watches for pods based on hostname. | 15:15 |
| dulek | And it looks like K8s nodeName is lowercased hostname. | 15:15 |
| dulek | We're not doing lowercasing. | 15:15 |
| dulek | SonicSong: Let me check out if there's a workaround for you. | 15:15 |
| dulek | SonicSong: Yeah, for a quick fix you can set KUBERNETES_NODE_NAME=kuryr-demo-kvm in the env that starts kuryr-daemon, e.g. in systemd servicefile. | 15:17 |
| dulek | SonicSong: I'll fix that on upstream. | 15:17 |
| *** gianpie__ has quit IRC | 15:20 | |
| *** ccamposr has quit IRC | 15:27 | |
| *** takamatsu has joined #openstack-kuryr | 15:29 | |
| openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Lowercase hostname when using it as K8s nodeName https://review.openstack.org/621188 | 15:36 |
| dulek | SonicSong: The fix is above, you can apply it to your repo and restart kuryr-daemon - it should help. | 15:37 |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-tempest-plugin master: Enhance svc namespace isolation with LoadBalancer type svc https://review.openstack.org/618751 | 15:48 |
| *** maysams has quit IRC | 15:52 | |
| *** takamatsu has quit IRC | 15:53 | |
| *** premsankar has joined #openstack-kuryr | 16:14 | |
| *** pcaruana has joined #openstack-kuryr | 16:26 | |
| *** celebdor has quit IRC | 16:51 | |
| *** pcaruana has quit IRC | 17:04 | |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Fix quota readiness check https://review.openstack.org/620883 | 17:35 |
| *** openstackgerrit has quit IRC | 17:51 | |
| *** celebdor has joined #openstack-kuryr | 19:31 | |
| *** rh-jelabarre has quit IRC | 20:03 | |
| *** rh-jelabarre has joined #openstack-kuryr | 20:05 | |
| *** jistr has quit IRC | 20:11 | |
| *** jistr has joined #openstack-kuryr | 20:13 | |
| *** takamatsu has joined #openstack-kuryr | 20:53 | |
| *** celebdor has quit IRC | 21:23 | |
| *** tosky has quit IRC | 23:42 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!