| *** hongbin has joined #openstack-kuryr | 02:32 | |
| *** takamatsu has quit IRC | 03:25 | |
| *** spsurya has joined #openstack-kuryr | 04:14 | |
| *** gkadam has joined #openstack-kuryr | 04:55 | |
| *** takamatsu has joined #openstack-kuryr | 05:05 | |
| *** takamatsu has quit IRC | 05:10 | |
| *** hongbin has quit IRC | 05:16 | |
| *** janki has joined #openstack-kuryr | 06:15 | |
| *** gcheresh_ has joined #openstack-kuryr | 06:16 | |
| *** gkadam has quit IRC | 06:27 | |
| *** maysams has joined #openstack-kuryr | 06:46 | |
| *** ccamposr has joined #openstack-kuryr | 07:01 | |
| *** gcheresh_ has quit IRC | 07:10 | |
| *** gcheresh_ has joined #openstack-kuryr | 07:14 | |
| *** aperevalov has quit IRC | 07:19 | |
| *** gkadam has joined #openstack-kuryr | 08:02 | |
| *** celebdor has joined #openstack-kuryr | 08:29 | |
| *** pcaruana has joined #openstack-kuryr | 08:32 | |
| *** ccamposr has quit IRC | 09:15 | |
| *** ccamposr has joined #openstack-kuryr | 09:18 | |
| *** celebdor has quit IRC | 09:26 | |
| *** celebdor has joined #openstack-kuryr | 09:26 | |
| *** celebdor has quit IRC | 09:31 | |
| *** premsankar has quit IRC | 09:32 | |
| *** garyloug has joined #openstack-kuryr | 09:42 | |
| *** garyloug has quit IRC | 09:43 | |
| *** yboaron has joined #openstack-kuryr | 09:45 | |
| dulek | maysams: This doesn't seem like a lot of tests: https://github.com/openshift/origin/blob/master/vendor/k8s.io/kubernetes/test/e2e/network/network_policy.go | 09:58 |
|---|---|---|
| dulek | maysams: Do you know if that's all? | 09:58 |
| *** alisanhaji has joined #openstack-kuryr | 10:03 | |
| *** yboaron_ has joined #openstack-kuryr | 10:09 | |
| *** yboaron has quit IRC | 10:12 | |
| maysams | dulek: It will probably not be enough. From the quick look a took, I saw that is missing to test with matchExpressions on the podSelector of the NP, for example | 10:18 |
| maysams | dulek: also, to update the label on a pod/svc/np and check that this is reflected in the traffic allowed to resources related | 10:19 |
| dulek | maysams: Hah, I probably should have checked if it's worth it before spending whole day trying to run that damned thing. :D | 10:20 |
| maysams | dulek: But, I do believe it's worth it. It checks the main cases | 10:21 |
| maysams | I was actually talking with ltomasbo yesterday regarding this | 10:22 |
| dulek | maysams: Okay, wonderful. :) "FAIL! -- 0 Passed | 8 Failed | 0 Pending | 1938 Skipped --- FAIL: TestE2E (1009.04s)" - that's the result without NP enabled in Kuryr. :D | 10:22 |
| maysams | dulek: hahah.. really? | 10:23 |
| maysams | dulek: seems we have somethings to fix | 10:24 |
| dulek | maysams: Yes, but I'm only running the NP tests, so it's totally expected. | 10:24 |
| dulek | maysams: Nah, I haven't enabled NP yet in Kuryr. | 10:24 |
| maysams | dulek: ahhhh... good! | 10:24 |
| maysams | dulek: I pinged the guy who replied when a asked about the tests they run.. Just to confirm that those ones are the only ones | 10:38 |
| dulek | maysams: I see, thanks! | 10:38 |
| maysams | np! | 10:39 |
| dulek | maysams: Okay, so now with NP enabled tests fail due to readiness probe being unreachable… | 10:45 |
| dulek | maysams: So that might be the first bug. :D | 10:45 |
| maysams | dulek: I though we were skipping pods with hostNetwork | 10:46 |
| maysams | dulek: but sure, seems so :P | 10:46 |
| dulek | maysams: That pod is not on host networking. | 10:47 |
| maysams | dulek: but the readiness probe runs in the controller and cni pods, right? So, seems that sg of the respective ports are being changed | 10:51 |
| dulek | maysams: At the moment I'm testing it with Kuryr deployed as a service, not in containers. | 10:52 |
| maysams | dulek: aaah, I see | 10:52 |
| dulek | maysams: It's the readiness probe of test pod that's failing. | 10:52 |
| maysams | dulek: okay | 10:53 |
| dulek | maysams: This might be more of a DevStack issue - we run kubelet on the host. | 10:53 |
| maysams | dulek, true | 10:57 |
| *** yboaron_ has quit IRC | 11:15 | |
| *** yboaron_ has joined #openstack-kuryr | 11:19 | |
| *** celebdor1 has joined #openstack-kuryr | 11:20 | |
| *** yboaron_ has quit IRC | 12:14 | |
| *** rh-jelabarre has joined #openstack-kuryr | 12:52 | |
| dulek | maysams: Okay, that was my DevStack's issue, I manually reconfigured kuryr.conf to have NP enabled, so I haven't this: https://github.com/openstack/kuryr-kubernetes/blob/3e3ed9dbb31133b5175ff792a9636603b2df95e1/devstack/lib/kuryr_kubernetes#L84 | 13:00 |
| dulek | Test run still doesn't look to great, but there's progress. | 13:01 |
| maysams | dulek: great that you solved the issue | 13:02 |
| maysams | dulek: so, the tests are still failing? | 13:03 |
| dulek | maysams: http://paste.openstack.org/show/746494/ | 13:03 |
| dulek | maysams: Doesn't seem like our issue. :P | 13:03 |
| dulek | maysams: I'm waiting for the run to finish. | 13:04 |
| maysams | dulek, yup | 13:04 |
| maysams | dulek, btw the guy just confirmed that those are all the tests they run | 13:05 |
| dulek | maysams: Okay. Seems like OOM kills Octavia VM's on my setup… | 13:07 |
| maysams | :/ | 13:10 |
| ltomasbo | hi dulek maysams! sorry I was stuck in meetings the whole day... | 13:10 |
| maysams | dulek, how much memory do you have? | 13:10 |
| ltomasbo | reading the backlog | 13:10 |
| maysams | dulek, 8? | 13:10 |
| dulek | maysams: Yup, it's 8. | 13:10 |
| ltomasbo | dulek, perhaps time to give it a try to the ovn-octavia driver instead of amphora... | 13:10 |
| dulek | ltomasbo: BTW Octavia is freaking dumb that when a VM goes into SHUTDOWN state, the LB will just hang on PENDING_CREATE without any way of deleting it. | 13:13 |
| ltomasbo | dulek, yep! I know... | 13:13 |
| ltomasbo | dulek, and there is no way to get out of that unless you go on the db and hack the status to CREATE or something | 13:14 |
| maysams | XD | 13:14 |
| dulek | ltomasbo: Really? Awesome. | 13:14 |
| ltomasbo | dulek, if you find any easier trick, let me know!!! | 13:15 |
| *** yboaron_ has joined #openstack-kuryr | 13:26 | |
| *** rh-jelabarre has quit IRC | 13:31 | |
| *** gcheresh_ has quit IRC | 13:43 | |
| *** gcheresh_ has joined #openstack-kuryr | 13:59 | |
| *** jistr is now known as jistr|call | 14:11 | |
| *** yboaron_ has quit IRC | 14:19 | |
| *** yboaron_ has joined #openstack-kuryr | 14:19 | |
| *** yboaron_ has quit IRC | 14:21 | |
| *** yboaron_ has joined #openstack-kuryr | 14:21 | |
| celebdor1 | ~/q danken | 14:30 |
| *** rh-jelabarre has joined #openstack-kuryr | 14:54 | |
| *** gcheresh_ has quit IRC | 15:08 | |
| *** maysams has quit IRC | 15:08 | |
| *** maysams has joined #openstack-kuryr | 15:10 | |
| *** maysams has quit IRC | 15:22 | |
| *** maysams has joined #openstack-kuryr | 15:35 | |
| *** yboaron_ has quit IRC | 15:40 | |
| *** yboaron_ has joined #openstack-kuryr | 15:41 | |
| *** alisanhaji has quit IRC | 15:43 | |
| *** janki has quit IRC | 15:52 | |
| *** jistr|call is now known as jistr | 16:00 | |
| *** premsankar has joined #openstack-kuryr | 17:03 | |
| *** gkadam has quit IRC | 17:08 | |
| *** ccamposr has quit IRC | 17:43 | |
| *** yboaron_ has quit IRC | 17:57 | |
| *** yboaron_ has joined #openstack-kuryr | 17:57 | |
| *** rh-jelabarre has quit IRC | 18:40 | |
| *** rh-jelabarre has joined #openstack-kuryr | 18:41 | |
| *** pcaruana has quit IRC | 19:14 | |
| *** dougbtv has quit IRC | 19:33 | |
| *** celebdor1 has quit IRC | 20:48 | |
| *** celebdor1 has joined #openstack-kuryr | 21:39 | |
| *** rh-jelabarre has quit IRC | 22:12 | |
| openstackgerrit | Maysa de Macedo Souza proposed openstack/kuryr-kubernetes master: Ensure only affected services are updated on Pod/NetworkPolicy events https://review.openstack.org/640216 | 22:52 |
| *** maysams has quit IRC | 23:18 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!