| *** hongbin has quit IRC | 00:02 | |
| *** limao has joined #openstack-kuryr | 00:53 | |
| dougbtv | wicked awesome :) | 00:58 |
|---|---|---|
| *** shashank_hegde has quit IRC | 01:46 | |
| *** yedongcan has joined #openstack-kuryr | 01:50 | |
| *** hongbin has joined #openstack-kuryr | 02:14 | |
| *** yuanying_ has quit IRC | 02:48 | |
| *** yuanying has joined #openstack-kuryr | 02:52 | |
| *** shashank_hegde has joined #openstack-kuryr | 03:03 | |
| janonymous | alraddarla_: yea, base.py have nothing much to do in it, once test_kuryr.py is completed only few changes needs to be made | 03:05 |
| *** yamamoto_ has joined #openstack-kuryr | 03:22 | |
| *** flaviof has left #openstack-kuryr | 03:24 | |
| *** saneax-_-|AFK is now known as saneax | 03:27 | |
| *** yuanying has quit IRC | 03:44 | |
| *** ivc_ has quit IRC | 03:45 | |
| *** saneax is now known as saneax-_-|AFK | 03:58 | |
| *** limao has quit IRC | 04:13 | |
| *** yedongcan has quit IRC | 04:23 | |
| *** hongbin has quit IRC | 04:31 | |
| *** limao has joined #openstack-kuryr | 04:35 | |
| *** ivc_ has joined #openstack-kuryr | 04:38 | |
| *** dougbtv has quit IRC | 04:39 | |
| *** ivc_ has quit IRC | 04:43 | |
| *** shashank_hegde has quit IRC | 04:43 | |
| *** janki has joined #openstack-kuryr | 04:56 | |
| *** openstackgerrit has joined #openstack-kuryr | 05:04 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/373745 | 05:04 |
| openstackgerrit | Jaivish Kothari(janonymous) proposed openstack/kuryr-libnetwork: Removing common mox functions https://review.openstack.org/405855 | 05:10 |
| openstackgerrit | Jaivish Kothari(janonymous) proposed openstack/kuryr-libnetwork: [WIP: to do after parent patch merge] Removing common mox functions https://review.openstack.org/405855 | 05:12 |
| janonymous | alraddarla_: ^^ | 05:13 |
| *** shashank_hegde has joined #openstack-kuryr | 05:18 | |
| *** shashank_hegde has quit IRC | 05:24 | |
| *** yamamoto_ has quit IRC | 05:31 | |
| *** yedongcan has joined #openstack-kuryr | 05:35 | |
| *** shashank_hegde has joined #openstack-kuryr | 05:36 | |
| *** shashank_hegde has quit IRC | 05:42 | |
| *** yamamoto_ has joined #openstack-kuryr | 05:57 | |
| *** yamamoto_ has quit IRC | 06:28 | |
| *** vikasc has quit IRC | 06:28 | |
| *** shashank_hegde has joined #openstack-kuryr | 06:31 | |
| *** vikasc has joined #openstack-kuryr | 06:31 | |
| *** openstackgerrit has quit IRC | 06:33 | |
| *** yedongcan has left #openstack-kuryr | 07:00 | |
| *** yamamoto_ has joined #openstack-kuryr | 07:08 | |
| *** ram___ has quit IRC | 07:18 | |
| *** saneax-_-|AFK is now known as saneax | 07:43 | |
| *** shashank_hegde has quit IRC | 07:51 | |
| *** pmannidi has quit IRC | 07:56 | |
| apuimedo | vikasc: please, review https://review.openstack.org/#/c/405426/ | 08:31 |
| apuimedo | I'd like to merge this fix | 08:31 |
| vikasc | apuimedo, looks good to me, have not tried though | 08:36 |
| apuimedo | vikasc: can you launch a devstack and let me know? | 08:36 |
| vikasc | apuimedo, sure | 08:36 |
| apuimedo | thanks vikasc | 08:36 |
| vikasc | yw | 08:36 |
| *** lmdaly has joined #openstack-kuryr | 08:53 | |
| vikasc | apuimedo, working | 08:56 |
| *** garyloug has joined #openstack-kuryr | 08:58 | |
| vikasc | apuimedo, but i am thinking how come it was working earlier also. i was using hyperkube 1.4.6 and it dint throw any error communicating with api-server | 08:58 |
| *** ivc_ has joined #openstack-kuryr | 09:02 | |
| *** limao has quit IRC | 09:18 | |
| apuimedo | vikasc: for a simple reason | 09:30 |
| apuimedo | that you had run with the same volume a docker 1.3.6 container :-D | 09:30 |
| apuimedo | so your volume got mounted | 09:31 |
| apuimedo | when you ran with 1.4.6 | 09:31 |
| vikasc | ooh | 09:31 |
| *** limao has joined #openstack-kuryr | 09:31 | |
| vikasc | :) | 09:31 |
| vikasc | right | 09:31 |
| *** openstackgerrit has joined #openstack-kuryr | 09:32 | |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-libnetwork: Move interface creation and deletion https://review.openstack.org/394547 | 09:32 |
| apuimedo | vikasc: let's get ^^ in | 09:32 |
| apuimedo | I just rebased it and applied some of dongcan's suggestions | 09:33 |
| vikasc | please allow me some time to take a look at it | 09:34 |
| apuimedo | vikasc: sure. I just would like to get it in today | 09:34 |
| apuimedo | so we can all start working on moving nested to kuryr-k8s | 09:34 |
| vikasc | sure | 09:34 |
| vikasc | cool | 09:35 |
| openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-libnetwork: [WIP]Driver based model for kuryr-libnetwork https://review.openstack.org/400365 | 09:50 |
| ltomasbo | hi apuimedo, vikasc. I follow irenab advice to split the vlan support out of the controller | 10:00 |
| ltomasbo | https://review.openstack.org/#/c/402462/ | 10:00 |
| vikasc | thanks ltomasbo | 10:01 |
| ltomasbo | not sure how that fits with the patch you just rebased... | 10:01 |
| ltomasbo | can you take a look? | 10:01 |
| apuimedo | ltomasbo: heh | 10:01 |
| apuimedo | hopefully it work be too conflictive | 10:02 |
| apuimedo | ltomasbo: you've got a lot of unit test failures | 10:02 |
| ltomasbo | will see :D | 10:02 |
| ltomasbo | yes, I included a new object in app | 10:03 |
| ltomasbo | so I need to update the tests to mock that | 10:03 |
| ltomasbo | also want to include some new ones (as well as fullstack) | 10:03 |
| openstackgerrit | Merged openstack/kuryr-kubernetes: devstack: fix k8s setup-files for hyperkube >=1.4 https://review.openstack.org/405426 | 10:04 |
| ltomasbo | I think all the failures are due to the next: | 10:04 |
| ltomasbo | AttributeError: 'Flask' object has no attribute 'driver' | 10:04 |
| apuimedo | ltomasbo: do you have some fix in mind for that? | 10:05 |
| ltomasbo | didn't have the time yet to dig into it, but for the old tests it should be as simple as mocking an driver object | 10:06 |
| ltomasbo | into app.driver | 10:06 |
| ltomasbo | with the default driver class | 10:07 |
| ltomasbo | (hopefully) | 10:07 |
| apuimedo | ok | 10:07 |
| ltomasbo | I'll give it a try later today | 10:07 |
| ltomasbo | by the way, as you may already discuss this before | 10:07 |
| ltomasbo | in https://review.openstack.org/#/c/394547/14/kuryr_libnetwork/controllers.py | 10:08 |
| *** lmdaly has quit IRC | 10:08 | |
| ltomasbo | I see the plan is to move the binding.port_bind() to CreateEndpoint | 10:08 |
| apuimedo | ltomasbo: yes, that has to be done to follow properly docker libnetwork spec | 10:09 |
| apuimedo | also, on stopping the container, it moves the iface back into the host namespace for us | 10:09 |
| ltomasbo | this will solve that issue? | 10:09 |
| *** lmdaly has joined #openstack-kuryr | 10:10 | |
| ltomasbo | going to try it right way and see if it also solves the vlan issue with that! | 10:11 |
| apuimedo | ltomasbo: it will | 10:12 |
| apuimedo | (I think) | 10:12 |
| ltomasbo | I'll give it a big +1 then! It was so annoying to have to delete manually the interfaces for the vlan testing :D | 10:13 |
| *** limao has quit IRC | 10:27 | |
| ltomasbo | yep, it does solve the problem! | 10:29 |
| apuimedo | ;-) | 10:30 |
| apuimedo | ivc_: I'm curious about l76 https://review.openstack.org/#/c/404038/5/kuryr_kubernetes/cni/handlers.py | 10:42 |
| ivc_ | apuimedo thats because we use 'active' in annotations to signal neutron port availability, but for os-vif we need an active port | 10:48 |
| apuimedo | odd | 10:49 |
| apuimedo | but I'll take it | 10:49 |
| apuimedo | ivc_: any reason for the change s/kuryr/kuryr-cni/ for the executable? | 10:49 |
| ivc_ | apuimedo yup. i expect that eventually we'll have 'kuryr' as cli | 10:50 |
| apuimedo | ivc_: if you spin up another patch set of this change, remove the depends on, since it is not necessary anymore | 10:50 |
| apuimedo | ivc_: cni drivers are usually not in path | 10:50 |
| apuimedo | and we could always have kuryrctl | 10:51 |
| apuimedo | :P | 10:51 |
| ivc_ | apuimedo yes, but i'd like to make it less confusing :P | 10:51 |
| apuimedo | very well | 10:51 |
| apuimedo | ivc_: now you only need to get vikasc or irenab | 10:52 |
| apuimedo | ;-) | 10:52 |
| apuimedo | ivc_: anything I can help you with? | 10:52 |
| ivc_ | apuimedo regarding 'odd' - well it is but we need something for neutron.port.status and it does fit that role. another option would probably be to have our own data model for neutron, but that would require neutron-client -> model translation and then model -> os-vif translation | 10:53 |
| ivc_ | apuimedo i wish we had a common neutron data model in neutron-lib that neutron-client would return instead of those 'dict's | 10:55 |
| *** yamamoto_ has quit IRC | 10:57 | |
| ivc_ | apuimedo e.g. if you look at nova, they have their own models and have to translate both neutron-client and os-vif | 10:57 |
| apuimedo | haven't looked since before os-vif times | 10:58 |
| apuimedo | mmm | 10:58 |
| apuimedo | having to do translation sucks | 10:58 |
| apuimedo | IMHO nova's and our usage should be the intended purpose and hsouldn't need translation | 10:58 |
| ivc_ | yes, but we don't have network model that is shared among projects. not yet at least | 10:59 |
| apuimedo | right | 11:01 |
| ivc_ | apuimedo have you seen https://review.openstack.org/#/c/313056/ ? | 11:05 |
| apuimedo | no, I hadn't, ivc_ | 11:06 |
| apuimedo | but it was coming any day | 11:06 |
| apuimedo | just moving the code around | 11:06 |
| ivc_ | apuimedo i mean, should we target octavia with our services implementation? | 11:12 |
| ivc_ | apuimedo because that would mean that we probably won't be able to use lightweight haproxy (agent-based) and will be limited to haproxy-in-vm, which is slow | 11:13 |
| ivc_ | apuimedo and even tho octavia doc states that "Octavia accomplishes its delivery of load balancing services by managing a fleet of virtual machines, *containers, or bare metal servers*—collectively known as amphorae" - there's no support for containers/bm yet (and from summit feedback it does not seem like its coming in near future) | 11:15 |
| apuimedo | ivc_: afaik they are not dropping the lightweight one just yet | 11:16 |
| apuimedo | they moved the code for now | 11:16 |
| apuimedo | but ajo can confirm | 11:16 |
| * ajo reads | 11:16 | |
| ajo | hmm, last time I talked to the octavia cores, it seemed like containers were near, but | 11:17 |
| ajo | there have been many changes, I can't confirm | 11:17 |
| apuimedo | ajo: but the old neutron code is still there, right? | 11:19 |
| ajo | apuimedo what do you mean about the old code? | 11:19 |
| ajo | ah, sorry, I didn't read it all | 11:20 |
| apuimedo | the one that spawned haproxys anywhere | 11:20 |
| ajo | yes, but how's that related to kuryr ? | 11:20 |
| ajo | we keep it, I know nmagnezi updated some patches to have something in the meanwhile | 11:20 |
| apuimedo | because we have to use lbaasv2 | 11:20 |
| ajo | ahhhhhh, ok | 11:21 |
| ajo | apuimedo talk to nmagnezi | 11:21 |
| apuimedo | and we don't want VMs | 11:21 |
| ajo | he'll be a reliable source of information | 11:21 |
| ajo | ok | 11:21 |
| apuimedo | so the old approach is better than VMs | 11:21 |
| ajo | I understand now | 11:21 |
| ajo | apuimedo of course, less heavy, but in the DVR world, VMs will be better | 11:21 |
| ajo | you spread the load across your whole cloud | 11:21 |
| ajo | instead of a node | 11:21 |
| ajo | or a few | 11:21 |
| apuimedo | nah... distributed LB | 11:21 |
| ajo | it will still be better once they're able to use containers, of course | 11:22 |
| apuimedo | that's the way | 11:22 |
| ajo | apuimedo what do you mean by distributed LB? :) | 11:22 |
| ivc_ | ajo, distributed LB: e.g ovn with their loadbalancer abstraction | 11:23 |
| ivc_ | ajo, i.e. similar to how K8s does loadbalancing in kube-proxy - they have DNAT on hosts that initiate the connection instead of having centralised LB | 11:24 |
| ivc_ | ajo, that enables great scalability for LB and makes it EW instead of NB | 11:26 |
| ivc_ | apuimedo btw it might make sense if octavia could use kuryr/k8s to provide containers | 11:34 |
| ivc_ | apuimedo i mean we are enabling neutron ports for containers and octavia can just run haproxy pod without taking care of network config | 11:35 |
| ivc_ | apuimedo tho that might require some way for pod to be connected to multiple networks | 11:36 |
| ajo | :) ovn support for lbaasv2, that'd be great ;D | 11:40 |
| ivc_ | ajo, for us anything that does not require ~5 minutes to start LB would be great :) | 11:42 |
| *** yamamoto has joined #openstack-kuryr | 11:46 | |
| *** yamamoto has quit IRC | 11:50 | |
| *** yamamoto has joined #openstack-kuryr | 11:53 | |
| *** yamamoto has quit IRC | 11:53 | |
| *** palexster has quit IRC | 11:58 | |
| *** neil_ has quit IRC | 12:00 | |
| *** palexster has joined #openstack-kuryr | 12:03 | |
| *** yamamoto has joined #openstack-kuryr | 12:09 | |
| *** neiljerram has joined #openstack-kuryr | 12:16 | |
| *** limao has joined #openstack-kuryr | 12:18 | |
| *** limao has quit IRC | 12:23 | |
| *** lmdaly has quit IRC | 12:26 | |
| *** limao has joined #openstack-kuryr | 12:32 | |
| *** limao_ has joined #openstack-kuryr | 12:40 | |
| *** limao has quit IRC | 12:40 | |
| *** limao_ has quit IRC | 12:44 | |
| *** limao has joined #openstack-kuryr | 12:44 | |
| *** garyloug has quit IRC | 13:05 | |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: Experimental CNI & VIFBridge binding https://review.openstack.org/404038 | 13:11 |
| ivc_ | vikasc, can you take a look at https://review.openstack.org/404038 ? | 13:13 |
| *** dougbtv has joined #openstack-kuryr | 13:21 | |
| *** saneax is now known as saneax-_-|AFK | 13:38 | |
| alraddarla_ | janonymous, got it :) | 13:47 |
| *** yamamoto has quit IRC | 14:06 | |
| *** yamamoto has joined #openstack-kuryr | 14:10 | |
| *** yamamoto has quit IRC | 14:21 | |
| *** lmdaly has joined #openstack-kuryr | 14:22 | |
| *** yamamoto has joined #openstack-kuryr | 14:27 | |
| *** yamamoto has quit IRC | 14:37 | |
| *** yamamoto has joined #openstack-kuryr | 14:37 | |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-libnetwork: Nested-Containers: trunk subports management https://review.openstack.org/402462 | 14:40 |
| *** yamamoto has quit IRC | 14:41 | |
| *** yamamoto has joined #openstack-kuryr | 14:51 | |
| *** janki has quit IRC | 14:58 | |
| *** garyloug has joined #openstack-kuryr | 15:01 | |
| ltomasbo | hi apuimedo, I quickly update the unit/base.py for my patch and there is no more unit tests failures | 15:10 |
| ltomasbo | just 2 extra lines :D | 15:11 |
| *** openstackgerrit has quit IRC | 15:33 | |
| *** hongbin has joined #openstack-kuryr | 15:39 | |
| *** limao has quit IRC | 15:43 | |
| *** openstackgerrit has joined #openstack-kuryr | 15:52 | |
| openstackgerrit | Darla Ahlert proposed openstack/fuxi: Move from "tenant" to "project" https://review.openstack.org/406189 | 15:52 |
| *** david-lyle_ has joined #openstack-kuryr | 16:17 | |
| *** david-lyle has quit IRC | 16:19 | |
| *** saneax-_-|AFK is now known as saneax | 16:39 | |
| apuimedo | cool | 16:50 |
| *** saneax is now known as saneax-_-|AFK | 16:55 | |
| *** yamamoto has quit IRC | 16:56 | |
| *** yamamoto has joined #openstack-kuryr | 16:58 | |
| *** yamamoto has quit IRC | 16:58 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/373745 | 17:09 |
| *** tonanhngo has joined #openstack-kuryr | 17:10 | |
| *** lmdaly has quit IRC | 17:14 | |
| *** yamamoto has joined #openstack-kuryr | 17:18 | |
| *** yamamoto has quit IRC | 17:19 | |
| *** yamamoto has joined #openstack-kuryr | 17:20 | |
| *** l_jacob has joined #openstack-kuryr | 17:31 | |
| *** l_jacob has quit IRC | 17:33 | |
| *** shashank_hegde has joined #openstack-kuryr | 17:40 | |
| *** yamamoto has quit IRC | 17:49 | |
| *** yamamoto has joined #openstack-kuryr | 17:51 | |
| *** yamamoto has quit IRC | 17:55 | |
| *** david-lyle_ is now known as david-lyle | 18:08 | |
| *** garyloug has quit IRC | 18:19 | |
| *** diogogmt has joined #openstack-kuryr | 18:50 | |
| *** shashank_hegde_ has joined #openstack-kuryr | 19:19 | |
| *** shashank_hegde has quit IRC | 19:23 | |
| *** shashank_hegde_ is now known as shashank_hegde | 19:23 | |
| *** yamamoto has joined #openstack-kuryr | 19:31 | |
| *** yamamoto has quit IRC | 19:37 | |
| *** devlaps has quit IRC | 20:57 | |
| *** dougbtv has quit IRC | 21:46 | |
| *** shashank_hegde has quit IRC | 21:48 | |
| *** shashank_hegde has joined #openstack-kuryr | 21:50 | |
| *** yamamoto has joined #openstack-kuryr | 23:05 | |
| *** hongbin has quit IRC | 23:26 | |
| *** yamamoto has quit IRC | 23:35 | |
| *** yamamoto has joined #openstack-kuryr | 23:41 | |
| *** yamamoto has quit IRC | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!