| *** rh-jelabarre has quit IRC | 00:53 | |
| *** aojea has joined #openstack-kuryr | 01:02 | |
| *** aojea has quit IRC | 01:07 | |
| *** hongbin has joined #openstack-kuryr | 02:24 | |
| *** hongbin has quit IRC | 02:58 | |
| *** hongbin has joined #openstack-kuryr | 03:01 | |
| *** hongbin has quit IRC | 03:16 | |
| *** hongbin has joined #openstack-kuryr | 03:17 | |
| *** hongbin has quit IRC | 03:17 | |
| *** hongbin has joined #openstack-kuryr | 03:17 | |
| *** gkadam has joined #openstack-kuryr | 03:19 | |
| *** hongbin has quit IRC | 03:20 | |
| *** hongbin has joined #openstack-kuryr | 03:20 | |
| *** kiseok7 has joined #openstack-kuryr | 03:20 | |
| *** hongbin has quit IRC | 03:21 | |
| *** hongbin has joined #openstack-kuryr | 03:21 | |
| *** hongbin has quit IRC | 03:22 | |
| *** hongbin has joined #openstack-kuryr | 03:22 | |
| *** gkadam has quit IRC | 04:21 | |
| *** hongbin has quit IRC | 05:10 | |
| *** gcheresh has joined #openstack-kuryr | 06:08 | |
| *** spsurya has joined #openstack-kuryr | 06:20 | |
| *** yboaron_ has joined #openstack-kuryr | 06:25 | |
| *** yboaron_ has quit IRC | 06:33 | |
| *** ccamposr has joined #openstack-kuryr | 06:50 | |
| *** pcaruana has joined #openstack-kuryr | 07:41 | |
| *** gkadam has joined #openstack-kuryr | 07:51 | |
| *** openstackgerrit has joined #openstack-kuryr | 08:03 | |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Avoid octavia default sg rules collision with kuryr ones https://review.openstack.org/632126 | 08:03 |
|---|---|---|
| *** yboaron has joined #openstack-kuryr | 08:32 | |
| *** celebdor has joined #openstack-kuryr | 08:34 | |
| *** yboaron_ has joined #openstack-kuryr | 08:36 | |
| *** yboaron has quit IRC | 08:39 | |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Fix CRD update when NP has namespaceSelectors https://review.openstack.org/631230 | 08:41 |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Removing lbaasv2 related code https://review.openstack.org/631261 | 08:44 |
| *** yboaron_ has quit IRC | 09:07 | |
| *** yboaron_ has joined #openstack-kuryr | 09:07 | |
| celebdor | ltomasbo: why did you remove the fake k8s service in https://review.openstack.org/#/c/631261/5/devstack/lib/kuryr_kubernetes ? | 09:53 |
| ltomasbo | celebdor, seems that was only used for the lbaasV2 haproxy code | 09:58 |
| ltomasbo | fake service is there, that is the router_fake_service | 09:58 |
| celebdor | ok | 10:07 |
| celebdor | dulek: waiting for the bp to +2 https://review.openstack.org/#/c/631278/ | 10:10 |
| dulek | celebdor: Yes, yes, forgot about it. Will update after I'll finish the call. | 10:10 |
| celebdor | :-) | 10:10 |
| celebdor | thanks dulek ! | 10:10 |
| dulek | celebdor: And if ltomasbo is first I'll need to rebase anyway. ;) | 10:10 |
| celebdor | oh, true | 10:11 |
| celebdor | so approve his | 10:11 |
| celebdor | :-) | 10:11 |
| *** livelace has joined #openstack-kuryr | 10:11 | |
| ltomasbo | celebdor, it will be great if you could review the other NP related patch sets too... | 10:12 |
| dulek | ltomasbo: Can you update the commit message according to my comment? Sorry for being nitpicky, but to really address irenab's comment you should od it that way. | 10:12 |
| ltomasbo | dulek, sure! | 10:12 |
| ltomasbo | let me check! | 10:13 |
| dulek | celebdor: We'll probably see some race over which patch passes both gates first. :D | 10:13 |
| celebdor | dulek: for sure | 10:13 |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Removing lbaasv2 related code https://review.openstack.org/631261 | 10:14 |
| ltomasbo | dulek, done! | 10:14 |
| dulek | celebdor: Okay, so just +2, +W ltomasbo's patch and let's wish him luck. ;) | 10:16 |
| ltomasbo | :) | 10:16 |
| celebdor | may the gates be with you | 10:17 |
| ltomasbo | dulek, should not be that hard to rebase I guess... | 10:17 |
| openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Remove way of running without kuryr-daemon https://review.openstack.org/631278 | 10:32 |
| dulek | ltomasbo, celebdor: ^ | 10:33 |
| celebdor | thanks dulek | 10:35 |
| celebdor | ltomasbo: since you pushed the last few revisions of https://review.openstack.org/#/c/629856/ should we expect maysa to be the reviewer or should you be? | 10:38 |
| ltomasbo | celebdor, I don't know actually... | 10:39 |
| celebdor | xD | 10:39 |
| ltomasbo | celebdor, I check it and it is working on my nested env | 10:39 |
| celebdor | well, with the PTL in paternity leave | 10:39 |
| ltomasbo | probably we could wait for her to review it tomorrow | 10:39 |
| celebdor | I'd like to see both your and maysa's +1/+2 | 10:39 |
| ltomasbo | and then if see agrees, I'll +2? | 10:39 |
| celebdor | tbh | 10:39 |
| ltomasbo | yes yes, I'm fine with the code already, I review it, found a couple of issues and tested it | 10:40 |
| ltomasbo | celebdor, ok, I'll give my +2, but wait to merge it until she reviews it, ok? | 10:41 |
| ltomasbo | celebdor, this should be simple to review: https://review.openstack.org/#/c/632071/ | 10:42 |
| celebdor | ltomasbo: of course | 10:43 |
| ltomasbo | celebdor, and this one: https://review.openstack.org/#/c/631781/ | 10:43 |
| dulek | Any ideas how to push on DevStack patch? | 10:46 |
| celebdor | dulek: the etcd one? | 10:47 |
| dulek | celebdor: Yup! | 10:49 |
| celebdor | dulek: I pinged Andrea on #openstack-dev | 10:51 |
| dulek | celebdor: Thanks, I've added him to the review, but I guess pinging is better. ;) | 10:52 |
| celebdor | dulek: I'll try to get more reviewers on it | 10:53 |
| ltomasbo | dulek, yboaron_: https://review.openstack.org/#/c/632126/ | 11:17 |
| dulek | ltomasbo: Uh, that depends from Octavia's internal behavior? Like how it identifies the SG… | 11:19 |
| dulek | Bottom line is that Octavia has a lot to be done in terms of SG and general access management. | 11:19 |
| ltomasbo | dulek, not octavia, just the amphora driver | 11:20 |
| ltomasbo | dulek, for the ovn-octavia driver, nothing in handle on that side | 11:20 |
| dulek | ltomasbo: Octavia too, it should have API's to set SG's for everything Octavia is creating. | 11:20 |
| dulek | If I understand the issues we face correctly. | 11:21 |
| ltomasbo | dulek, ahh, yes, but those are not on the tenant project, so kuryr will not see them | 11:21 |
| ltomasbo | dulek, problem is that when creating listeners, octavia (the amphora driver actually) is creating security groups to allow the listeners port into the amphora VM | 11:22 |
| ltomasbo | dulek, if kuryr is faster than octavia, octavia will not add them (as there is already a rule on the given listener port/protocol | 11:22 |
| dulek | ltomasbo: Mhm, I get this. | 11:22 |
| ltomasbo | dulek, however, if octavia is faster (as it happens sometimes in some of our gates) | 11:22 |
| dulek | ltomasbo: So this particular case is really specific to Amphora driver. | 11:22 |
| dulek | ltomasbo: But our general issues with access management are problem of Octavia itself. :) | 11:23 |
| ltomasbo | dulek, then, specific to the amphora driver, and specific to the namespace isolation | 11:23 |
| ltomasbo | dulek, ownership of those security group rules is a different story yes, on the octavia roof | 11:24 |
| ltomasbo | actually octavia is pretty much like the neutron API, just the api, then is up to the drivers to handle how the loadbalancer actually works/behave | 11:24 |
| ltomasbo | amphora driver is based on the VMs, allow-address-pairs and security groups, but other drivers may not (as the ovn driver) | 11:25 |
| yboaron_ | ltomasbo, I was out for lunch, I saw that you got +W :-) | 11:39 |
| ltomasbo | yboaron_, ok! that what the thing we were discussing the other day about waiting for the octavia rules to be there | 11:40 |
| yboaron_ | ltomasbo, Yep, LGTM | 11:40 |
| ltomasbo | great! | 11:40 |
| *** rh-jelabarre has joined #openstack-kuryr | 12:53 | |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Avoid octavia default sg rules collision with kuryr ones https://review.openstack.org/632126 | 13:00 |
| celebdor | yay! | 13:28 |
| *** oanson has quit IRC | 13:31 | |
| celebdor | dulek: I added a few more reviewers | 13:33 |
| dulek | celebdor: Thanks! | 13:33 |
| *** aojea_ has joined #openstack-kuryr | 13:59 | |
| ltomasbo | dulek, https://review.openstack.org/#/c/632464/ (backport of the previous one) | 14:06 |
| dulek | ltomasbo: Done. | 14:11 |
| ltomasbo | dulek, thanks! | 14:11 |
| *** aojea_ has quit IRC | 14:46 | |
| *** aojea_ has joined #openstack-kuryr | 14:48 | |
| livelace | Hello. Does Kuryr support "externalIP" of a k8s "Service" ? | 14:52 |
| celebdor | livelace: it supports the loadbalancer service type assigning FIPs | 15:01 |
| *** aojea_ has quit IRC | 15:05 | |
| livelace | celebdor, I tried to deploy jenkins helm chart recently, LB was created, but during deletion of the Service I got error from kuryr-controller, because in the Service definition was externalIP which belong to different from Pod/Service subnets. | 15:06 |
| *** yboaron_ has quit IRC | 15:07 | |
| *** yboaron_ has joined #openstack-kuryr | 15:08 | |
| *** gcheresh has quit IRC | 15:26 | |
| ltomasbo | celebdor, dulek: I resubmitted https://review.openstack.org/#/c/632464/ (dummy mistake doing the backport) | 15:35 |
| celebdor | livelace: could you please file a bug on that? | 15:36 |
| livelace | celebdor, Yes, I'm going to try to reproduce that behaviour (it should be simple) with latest version of kuryr (need to pull docker images). | 15:40 |
| celebdor | thanks livelace! | 15:45 |
| *** yboaron_ has quit IRC | 15:47 | |
| *** yboaron_ has joined #openstack-kuryr | 15:50 | |
| *** openstackgerrit has quit IRC | 15:51 | |
| ltomasbo | celebdor, can you merge this so that I can rebase the follow up patch sets: https://review.openstack.org/#/c/631781/ | 15:56 |
| *** dims has quit IRC | 16:15 | |
| *** dims has joined #openstack-kuryr | 16:20 | |
| *** yboaron_ has quit IRC | 16:28 | |
| *** openstackgerrit has joined #openstack-kuryr | 16:29 | |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Ensure host to pod connectivity for NP https://review.openstack.org/632503 | 16:29 |
| ltomasbo | celebdor, the fix for host to pod connectivity when using network policies ^^ | 16:31 |
| openstackgerrit | hyunsik Yang proposed openstack/kuryr-kubernetes master: Fixed variables error with 'public-subnet' https://review.openstack.org/632510 | 16:48 |
| *** ccamposr has quit IRC | 16:52 | |
| *** ccamposr has joined #openstack-kuryr | 16:52 | |
| *** ccamposr has quit IRC | 16:53 | |
| *** pcaruana has quit IRC | 17:02 | |
| *** aojea has joined #openstack-kuryr | 17:41 | |
| *** aojea has quit IRC | 17:44 | |
| *** aojea has joined #openstack-kuryr | 17:44 | |
| *** ccamposr has joined #openstack-kuryr | 17:47 | |
| *** pcaruana has joined #openstack-kuryr | 18:01 | |
| *** pcaruana has quit IRC | 18:23 | |
| *** gkadam has quit IRC | 18:26 | |
| *** aojea has quit IRC | 18:46 | |
| *** ccamposr has quit IRC | 18:50 | |
| *** aojea has joined #openstack-kuryr | 18:56 | |
| *** aojea has quit IRC | 19:07 | |
| *** aojea has joined #openstack-kuryr | 19:08 | |
| *** aojea has quit IRC | 19:27 | |
| *** takamatsu has quit IRC | 19:27 | |
| *** pcaruana has joined #openstack-kuryr | 19:50 | |
| *** pcaruana has quit IRC | 20:11 | |
| *** yboaron_ has joined #openstack-kuryr | 20:19 | |
| *** pcaruana has joined #openstack-kuryr | 20:24 | |
| *** dims has quit IRC | 20:40 | |
| *** dims has joined #openstack-kuryr | 20:42 | |
| *** dims has quit IRC | 20:50 | |
| *** dims has joined #openstack-kuryr | 20:52 | |
| *** celebdor has quit IRC | 20:53 | |
| *** pcaruana has quit IRC | 21:19 | |
| *** aojea has joined #openstack-kuryr | 21:46 | |
| *** yboaron_ has quit IRC | 22:05 | |
| *** aojea has quit IRC | 23:24 | |
| *** livelace has quit IRC | 23:33 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!