| *** yamamoto has joined #openstack-kuryr | 00:36 | |
| *** atoth has joined #openstack-kuryr | 00:56 | |
| *** maysamacedos has quit IRC | 01:16 | |
| *** caowei has joined #openstack-kuryr | 01:25 | |
| *** pmannidi has joined #openstack-kuryr | 01:29 | |
| *** caowei has quit IRC | 01:32 | |
| *** kiennt26 has joined #openstack-kuryr | 01:46 | |
| *** hongbin has joined #openstack-kuryr | 01:56 | |
| *** threestrands_ has joined #openstack-kuryr | 02:02 | |
| *** jappleii__ has quit IRC | 02:04 | |
| *** hongbin has quit IRC | 02:53 | |
| *** hongbin has joined #openstack-kuryr | 02:54 | |
| *** atoth has quit IRC | 03:06 | |
| *** kiseok7 has quit IRC | 03:08 | |
| *** hongbin has quit IRC | 03:19 | |
| *** hongbin has joined #openstack-kuryr | 03:20 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements https://review.openstack.org/519094 | 03:21 |
|---|---|---|
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr master: Updated from global requirements https://review.openstack.org/516936 | 03:26 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork master: Updated from global requirements https://review.openstack.org/523740 | 03:27 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-kubernetes master: Updated from global requirements https://review.openstack.org/509781 | 03:27 |
| *** yamamoto_ has joined #openstack-kuryr | 03:39 | |
| *** yamamoto has quit IRC | 03:43 | |
| *** janonymous has joined #openstack-kuryr | 04:16 | |
| *** phuoc has quit IRC | 04:29 | |
| *** hongbin has quit IRC | 05:04 | |
| *** janki has joined #openstack-kuryr | 05:10 | |
| *** phuoc has joined #openstack-kuryr | 05:27 | |
| openstackgerrit | wanghui proposed openstack/kuryr master: fix Neutron misspelling https://review.openstack.org/534087 | 05:47 |
| *** yboaron has joined #openstack-kuryr | 06:12 | |
| *** yboaron has quit IRC | 06:25 | |
| *** gcheresh has joined #openstack-kuryr | 06:26 | |
| *** yboaron has joined #openstack-kuryr | 06:40 | |
| *** c00281451_ is now known as zengchen | 06:52 | |
| *** juriarte has joined #openstack-kuryr | 07:18 | |
| *** threestrands_ has quit IRC | 07:18 | |
| *** juriarte has quit IRC | 07:22 | |
| *** juriarte has joined #openstack-kuryr | 07:37 | |
| *** janki has quit IRC | 07:39 | |
| openstackgerrit | Yossi Boaron proposed openstack/kuryr-kubernetes master: Kuryr-Kubernetes openshift route integration devref https://review.openstack.org/530095 | 07:44 |
| *** aojea_ has joined #openstack-kuryr | 07:45 | |
| *** kaliya has joined #openstack-kuryr | 08:22 | |
| openstackgerrit | Merged openstack/kuryr master: fix Neutron misspelling https://review.openstack.org/534087 | 08:23 |
| leyal | yboaron , hi | 08:25 |
| yboaron | leyal, hi | 08:26 |
| leyal | yboaron, i am trying to review the ingress spec - and something that i think that i am not understanding | 08:27 |
| yboaron | leyal, OK | 08:28 |
| leyal | yboaron- in use use case example (line 205) , the spec describe that the ingress-spec is triggering the creation of the lbass polls and listeners | 08:28 |
| leyal | but isn't the service s1 should be there before the ingress is created ? | 08:29 |
| leyal | and the creation of the service should trigger the creation of lbaas pools ? | 08:30 |
| yboaron | leyal, yes , it;s a bit confusing , take a look at the last figure in spec , the one that describes the neutron lbass entities of both the L7 router and the service | 08:31 |
| yboaron | leyal, #link http://logs.openstack.org/66/529966/17/check/build-openstack-sphinx-docs/bef6873/html/devref/kuryr_kubernetes_ingress_design.html | 08:32 |
| leyal | yboaron, yes i still do not understand why the pools is created by the ingress handler - as they need already created during the service creation in the service handler.. | 08:35 |
| yboaron | leyal, did u manage to see the figure under section 'The L7 router driver¶ | 08:36 |
| yboaron | ' in the link I just sent ? | 08:36 |
| yboaron | leyal, so the service handler will create the entities colored in grey (LB,listeners,pools,members) | 08:37 |
| yboaron | leyal, there are two pools in that case , one that attached to the user service . this is not relevant for ingress , just lbaas handler as it supported in current kuryr version | 08:38 |
| yboaron | leyal, the other pool (colored in green) is relevant for ingress | 08:39 |
| yboaron | leyal, this is the link to the relevant figure #link : http://logs.openstack.org/66/529966/17/check/build-openstack-sphinx-docs/bef6873/html/_images/l7_routing_and_user_lb_neutron_entities.svg | 08:40 |
| leyal | yboaron, tell me if i got it right - when service is create it's create a listener on the service ip and port - this listener is connected to pool , and when the ingress is create you create an http listener and new pool that connected to the same members of the service pools? | 08:42 |
| yboaron | leyal, in general that's correct , with one correction , when the ingress is created we need to create some entity that will run the ingress/L7 rules , | 08:44 |
| yboaron | leyal, this entity is the L7policy/L7rule , L7 policy pointed the filtered traffic towards relevant pool (the one that attached to L7 router LB | 08:45 |
| *** janki has joined #openstack-kuryr | 08:46 | |
| *** janki has quit IRC | 08:46 | |
| *** janki has joined #openstack-kuryr | 08:47 | |
| leyal | yboaron , cool. so my question is why we can't reuse the pools of service in the l7-listener instead of creating new polls | 08:47 |
| leyal | yboaron , i think it's will follow the logic that ingress maped to service and service is mapped to endpoints | 08:48 |
| yboaron | leyal, that's could be a good idea , but unfortunately it's invalid in Lbaas/OCTAVIA | 08:50 |
| openstackgerrit | wanghui proposed openstack/kuryr master: change http to https https://review.openstack.org/534191 | 08:50 |
| leyal | yboaron, when i create l7 policy by "neutron --policy policy1 lbaas-create-l7policy --name "policy1" --description "My Description" --listener "listener1" --action redirect_to_pool --pool "pool1" --position 1" | 08:52 |
| leyal | i am not free to choose my pool ? | 08:52 |
| yboaron | leyal, before directing l7 policy to pool u need to create the pool , every pool must be connected to either LB or listener | 08:54 |
| yboaron | leyal, and it's invalid to point to pool that already used by another LB | 08:56 |
| yboaron | leyal, anyhow , you should consider the case in which ingress is created before service ? | 08:58 |
| leyal | yboaron, yep i saw now that it's 1:1 relationship between listener/LB and pool , it's surprises me . | 08:59 |
| leyal | yboaron, is that scenario is valid? | 09:00 |
| leyal | you will need the members anyway .. | 09:00 |
| yboaron | leyal, the members will be added later , when the service is created | 09:01 |
| leyal | yboaron , any way if you can add 2 listeners to same pools it's not relevant .. | 09:02 |
| leyal | s/can/can't | 09:03 |
| leyal | thanks for all explanations :) | 09:03 |
| yboaron | leyal, you're welcome | 09:04 |
| *** garyloug has joined #openstack-kuryr | 09:07 | |
| *** kiennt26 has quit IRC | 10:11 | |
| *** dougbtv_ has joined #openstack-kuryr | 10:14 | |
| *** dougbtv has quit IRC | 10:16 | |
| *** aojea_ has quit IRC | 10:16 | |
| *** aojea_ has joined #openstack-kuryr | 10:16 | |
| *** dougbtv__ has joined #openstack-kuryr | 10:28 | |
| *** dougbtv_ has quit IRC | 10:31 | |
| *** vkuramshin has joined #openstack-kuryr | 11:01 | |
| *** aojea_ has quit IRC | 11:19 | |
| *** yamamoto_ has quit IRC | 11:27 | |
| *** aojea has joined #openstack-kuryr | 11:39 | |
| *** aojea has quit IRC | 11:40 | |
| *** maysamacedos has joined #openstack-kuryr | 11:40 | |
| *** leyal has quit IRC | 11:48 | |
| *** lihi has quit IRC | 11:48 | |
| *** irenab has quit IRC | 11:48 | |
| *** oanson has quit IRC | 11:48 | |
| *** irenab has joined #openstack-kuryr | 11:48 | |
| *** oanson has joined #openstack-kuryr | 11:50 | |
| *** lihi has joined #openstack-kuryr | 11:52 | |
| openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Use virtualenv to builid kuryr-cni Docker image https://review.openstack.org/534253 | 12:05 |
| dulek | apuimedo: ^ | 12:05 |
| *** pmannidi has quit IRC | 12:09 | |
| *** lihi has quit IRC | 12:10 | |
| *** oanson has quit IRC | 12:10 | |
| *** irenab has quit IRC | 12:10 | |
| *** oanson has joined #openstack-kuryr | 12:12 | |
| *** irenab has joined #openstack-kuryr | 12:14 | |
| *** lihi has joined #openstack-kuryr | 12:18 | |
| *** yamamoto has joined #openstack-kuryr | 12:26 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements https://review.openstack.org/519094 | 12:26 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements https://review.openstack.org/519094 | 12:27 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements https://review.openstack.org/519094 | 12:27 |
| *** leyal has joined #openstack-kuryr | 12:27 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr master: Updated from global requirements https://review.openstack.org/516936 | 12:28 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork master: Updated from global requirements https://review.openstack.org/523740 | 12:28 |
| apuimedo | dulek: to builid? | 12:28 |
| dulek | apuimedo: Oh, you know what I've meant. ;) | 12:29 |
| apuimedo | :-) | 12:30 |
| dulek | apuimedo: https://review.openstack.org/#/c/534206/ | 12:30 |
| dulek | apuimedo: And this broke Nova, I'm not sure if we won't suffer as well: https://github.com/svinota/pyroute2/pull/443 | 12:31 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr master: Updated from global requirements https://review.openstack.org/516936 | 12:31 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork master: Updated from global requirements https://review.openstack.org/523740 | 12:31 |
| *** pmannidi has joined #openstack-kuryr | 12:32 | |
| *** pcaruana has joined #openstack-kuryr | 12:33 | |
| maysamacedos | irenab dulek apuimedo: May I ask you to take a look at https://review.openstack.org/#/c/533292/ ? Thank you. | 12:35 |
| irenab | maysamacedos, asap | 12:35 |
| maysamacedos | irenab: appreciate it | 12:38 |
| *** pmannidi has quit IRC | 13:05 | |
| yboaron | ping irenab | 13:06 |
| irenab | yboaron, pong | 13:10 |
| yboaron | irenab, hi | 13:11 |
| yboaron | irenab, regarding your comment #link https://review.openstack.org/#/c/529966/ | 13:11 |
| yboaron | irenab, could u please elaborate ? | 13:11 |
| yboaron | irenab, do u suggest to annotate the L7 router details in ingress resources ? | 13:12 |
| irenab | yes, somehow annotating the node (which is physical related entity) does not look the way to go | 13:14 |
| irenab | other option is to make sure that Router is 'created' at deploy time and then populate the details in the kuryr.conf | 13:15 |
| *** garyloug has quit IRC | 13:16 | |
| yboaron | irenab, yes that my plan , the router will be install at deploy time , do u suggest to keep all neutron resources used for L7 router in kuryr.conf file ? | 13:18 |
| yboaron | irenab, I think that having L7_ROUTER CRD is the most suitable way to solve this issue , but I prefer to have it in next phase | 13:19 |
| yboaron | irenab, considering all the issues we"ll have with introducing new Paas into kuryr sources, I would prefer to have it in next phase | 13:21 |
| irenab | yboaron, agree on CRD, but I think having it in the kuryr.cof initially is prefered over node annotation | 13:21 |
| irenab | It can be similar to what we have with the service subnet or project id, the neutron details come in kuryr.conf | 13:23 |
| yboaron | irenab, do you mean to keep all neutron resources relevant for L7 router in kuryr.conf ? | 13:23 |
| irenab | my suggestion was for the details you have for the initial router details | 13:23 |
| irenab | the details you planned to keep in node annotation | 13:23 |
| irenab | {"fip": "172.24.4.14", | 13:24 |
| irenab | "router_lb": {"versioned_object.data": {"id": "90732f0a-651a-4b17-a14e-9b0e01fbe774", | 13:24 |
| irenab | "ip": "10.0.0.154", "name": "kuryr-l7-router", "port_id": "5c71a29a-0dc1-461e-81ee-2258a7e3842d", | 13:24 |
| irenab | "project_id": "868307936d384c21824e5eb0425a3f42", "subnet_id": "9f6d8c9f-d22d-480e-80f5-867daa050ff8" | 13:24 |
| yboaron | irenab, don't us think it will be 'dirty' to have all this details in kuryr.conf ? | 13:25 |
| irenab | but if you think node annotation is better, we can start with it. I may miss something | 13:25 |
| irenab | The queston is if you need all the details in the config, or some details can be retrieved given for example the LB-id | 13:26 |
| irenab | or port-id | 13:26 |
| yboaron | irenab, you are not missing anything , it's definitely a point that should be handled in more appropriate way | 13:26 |
| yboaron | irenab, keep in mind that all the LbaaS drivers drivers works with this objects - I don't want to invent the wheel .. | 13:27 |
| yboaron | irenab, so yes I prefer to keep this information to re-use the Lbaas driver code | 13:27 |
| yboaron | irenab, I meant all the functions of lbaas driver .. | 13:28 |
| *** atoth has joined #openstack-kuryr | 13:29 | |
| irenab | yboaron, arewe talking about my other question on the different parts between OCP Router and Ingress ? | 13:30 |
| *** garyloug has joined #openstack-kuryr | 13:30 | |
| yboaron | irenab, no , we discussed the L7 router annotation | 13:31 |
| irenab | ok :-) | 13:31 |
| yboaron | irenab, as per the ingress/route , in openshift deployment we"ll be able to create both ingress and route resources | 13:31 |
| *** yamamoto_ has joined #openstack-kuryr | 13:32 | |
| yboaron | irenab, while in K8s environment only ingress resources will be available | 13:32 |
| irenab | So each may have its own l7 router, right? | 13:32 |
| irenab | so for OCP is having 2 routers is a valid option, better not go with kuryr.conf | 13:33 |
| yboaron | irenab, no we"ll have a single router , | 13:33 |
| irenab | ok, then it should be ok | 13:34 |
| irenab | But probably not annotating Ingress/Route, since it will be data duplication | 13:34 |
| apuimedo | can we just forget about multiple routers for now? | 13:35 |
| *** yamamoto_ has quit IRC | 13:35 | |
| irenab | apuimedo, we did :-) | 13:35 |
| *** yamamoto has quit IRC | 13:35 | |
| irenab | now trying to see the least hacky way to keep L7 Router details | 13:36 |
| yboaron | irenab, single router , annotating node (in first phase) , is it OK with you ? | 13:37 |
| apuimedo | which details? | 13:38 |
| *** danil has joined #openstack-kuryr | 13:38 | |
| yboaron | apuimedo, neutron LB,listeners , floating IP | 13:38 |
| irenab | I prefer not to bind the L7 Router details to node | 13:38 |
| yboaron | apuimedo, all the objects that used for recovering lbaas resources by lbaas driver | 13:39 |
| irenab | yboaron, given LB uuid, can you get the rest of the LB related details? | 13:39 |
| apuimedo | can't we just tag them in lbaasv2 api? | 13:40 |
| irenab | apuimedo, since L7 router will be created at deploy time, we can put the details in the kuryr.conf | 13:40 |
| yboaron | apuimedo, irenab : we need to support two flows , create and recover | 13:41 |
| irenab | tags can work as well | 13:41 |
| irenab | yboaron, create is at deploy time | 13:41 |
| yboaron | irenab, apuimedo and unlike k8s services we don't have the VIP for uniqueness | 13:41 |
| irenab | before kuryr-controller is running | 13:41 |
| irenab | yboaron, what do you mean by non-unique VIP | 13:42 |
| irenab | apuimedo, going forward CRD probably the way to keep the details | 13:43 |
| yboaron | irenab, for k8s service LB , kubernetes allocates the VIP for this LB , this vip is unique ,for the L7 router we should allocate the VIP | 13:44 |
| irenab | yboaron, but we have the same with k8s api service. apuimedo please correct me if I am wrong | 13:44 |
| irenab | yboaron, apuimedo sorry, have some meeting for the next 1.5 hours. Will follow up | 13:45 |
| yboaron | irenab, OK | 13:45 |
| *** yamamoto has joined #openstack-kuryr | 13:47 | |
| *** yamamoto has quit IRC | 13:52 | |
| *** kiennt26 has joined #openstack-kuryr | 13:58 | |
| danil | hello, folks. Could you please help me? Maybe someone faced this error. Error in "...python2.7/site-packages/oslo_versionedobjects/base.py" AttributeError: Subnet object has no attribute 'ips' while creating pod. Maybe problems with version of oslo_versionedobjects ?? | 13:59 |
| danil | thanks | 13:59 |
| apuimedo | xD | 14:02 |
| apuimedo | danil: yes | 14:02 |
| apuimedo | modify requirements.txt | 14:02 |
| apuimedo | where there is os-vif | 14:02 |
| apuimedo | instead of >=1.7.0 | 14:02 |
| apuimedo | put <1.8.0 | 14:02 |
| * apuimedo back to meeting | 14:02 | |
| danil | apuimedo, thanks a lot. This works. But other errors have appeared "NoMatches: No 'kuryr_kubernetes.vif_translators' driver found, looking for u'binding_failed'" | 14:07 |
| *** yamamoto has joined #openstack-kuryr | 14:17 | |
| *** danil has quit IRC | 14:17 | |
| *** yamamoto has quit IRC | 14:22 | |
| *** yboaron has quit IRC | 14:33 | |
| *** hongbin has joined #openstack-kuryr | 14:39 | |
| *** garyloug_ has joined #openstack-kuryr | 14:46 | |
| *** garyloug has quit IRC | 14:49 | |
| *** yamamoto has joined #openstack-kuryr | 15:03 | |
| *** yamamoto has quit IRC | 15:03 | |
| *** vkuramshin has left #openstack-kuryr | 15:11 | |
| *** yamamoto has joined #openstack-kuryr | 15:12 | |
| *** yamamoto has quit IRC | 15:17 | |
| *** maysamacedos has quit IRC | 15:31 | |
| *** gcheresh has quit IRC | 15:32 | |
| *** maysamacedos has joined #openstack-kuryr | 15:36 | |
| *** maysamacedos has quit IRC | 15:39 | |
| *** yamamoto has joined #openstack-kuryr | 15:58 | |
| *** yamamoto has quit IRC | 16:01 | |
| *** yamamoto has joined #openstack-kuryr | 16:01 | |
| *** yamamoto has quit IRC | 16:01 | |
| *** kiennt26 has quit IRC | 16:06 | |
| *** gcheresh has joined #openstack-kuryr | 16:24 | |
| *** janki has quit IRC | 16:29 | |
| *** vikasc has quit IRC | 16:35 | |
| *** maysamacedos has joined #openstack-kuryr | 16:39 | |
| *** yboaron has joined #openstack-kuryr | 16:40 | |
| *** maysamacedos has quit IRC | 16:41 | |
| *** yamamoto has joined #openstack-kuryr | 16:41 | |
| openstackgerrit | Eyal Leshem proposed openstack/kuryr-kubernetes master: Translate k8s policy to SG https://review.openstack.org/526916 | 16:41 |
| *** yboaron has quit IRC | 16:45 | |
| *** yamamoto has quit IRC | 16:46 | |
| *** juriarte has quit IRC | 16:46 | |
| openstackgerrit | Eyal Leshem proposed openstack/kuryr-kubernetes master: Translate k8s policy to SG https://review.openstack.org/526916 | 16:47 |
| *** vikasc has joined #openstack-kuryr | 16:51 | |
| *** atoth has quit IRC | 17:03 | |
| *** maysamacedos has joined #openstack-kuryr | 17:04 | |
| apuimedo | dulek: thanks for the quick replies to the patch review :-) | 17:26 |
| *** openstackgerrit has quit IRC | 17:33 | |
| *** kaliya has quit IRC | 17:39 | |
| *** yamamoto has joined #openstack-kuryr | 17:44 | |
| *** yamamoto has quit IRC | 17:58 | |
| *** gcheresh has quit IRC | 17:59 | |
| *** maysamacedos has quit IRC | 18:23 | |
| *** maysamacedos has joined #openstack-kuryr | 18:30 | |
| *** gcheresh has joined #openstack-kuryr | 19:26 | |
| *** gcheresh has quit IRC | 19:33 | |
| *** maysamacedos has quit IRC | 19:41 | |
| *** maysamacedos has joined #openstack-kuryr | 19:47 | |
| *** pcaruana has quit IRC | 19:50 | |
| *** maysamacedos has quit IRC | 20:07 | |
| *** janonymous has quit IRC | 20:15 | |
| *** dougbtv_ has joined #openstack-kuryr | 20:40 | |
| *** dougbtv__ has quit IRC | 20:43 | |
| *** dougbtv has joined #openstack-kuryr | 20:44 | |
| *** dougbtv_ has quit IRC | 20:46 | |
| *** gcheresh has joined #openstack-kuryr | 20:50 | |
| *** phuoc has quit IRC | 21:01 | |
| *** phuoc has joined #openstack-kuryr | 21:02 | |
| *** gcheresh has quit IRC | 21:08 | |
| *** threestrands_ has joined #openstack-kuryr | 21:18 | |
| *** garyloug_ has quit IRC | 21:40 | |
| *** pmannidi has joined #openstack-kuryr | 23:17 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
