| *** tonanhngo has joined #openstack-kuryr | 00:11 | |
| *** tonanhngo has quit IRC | 00:13 | |
| *** yuanying_ has joined #openstack-kuryr | 00:19 | |
| *** yuanying has quit IRC | 00:19 | |
| *** yuanying_ has quit IRC | 00:19 | |
| *** yuanying has joined #openstack-kuryr | 00:19 | |
| *** yuanying_ has joined #openstack-kuryr | 00:20 | |
| *** yuanying has quit IRC | 00:20 | |
| *** yuanying has joined #openstack-kuryr | 00:21 | |
| *** yuanying_ has quit IRC | 00:21 | |
| *** yuanying_ has joined #openstack-kuryr | 00:21 | |
| *** yuanying has quit IRC | 00:21 | |
| *** yuanying_ has quit IRC | 00:21 | |
| *** yuanying has joined #openstack-kuryr | 00:22 | |
| *** yuanying has quit IRC | 00:22 | |
| *** yuanying has joined #openstack-kuryr | 00:22 | |
| *** vikasc has quit IRC | 00:39 | |
| *** tonanhngo has joined #openstack-kuryr | 00:43 | |
| *** tonanhngo has quit IRC | 00:46 | |
| *** limao has joined #openstack-kuryr | 00:50 | |
| openstackgerrit | Zhigang Li proposed openstack/kuryr-libnetwork: Replace six.iteritems() with .items() https://review.openstack.org/401056 | 01:10 |
|---|---|---|
| *** tonanhngo has joined #openstack-kuryr | 01:11 | |
| *** tonanhngo has quit IRC | 01:13 | |
| *** shashank_hegde has quit IRC | 01:16 | |
| *** tonanhngo has joined #openstack-kuryr | 01:23 | |
| *** tonanhngo has quit IRC | 01:24 | |
| *** yedongcan has joined #openstack-kuryr | 01:29 | |
| *** shashank_hegde has joined #openstack-kuryr | 01:31 | |
| openstackgerrit | Merged openstack/fuxi: Retrieve volumes from cinder with matched metadata https://review.openstack.org/390448 | 01:40 |
| *** tonanhngo has joined #openstack-kuryr | 02:02 | |
| *** tonanhngo has quit IRC | 02:04 | |
| *** hongbin has joined #openstack-kuryr | 02:10 | |
| *** tonanhngo has joined #openstack-kuryr | 02:21 | |
| *** tonanhngo has quit IRC | 02:23 | |
| *** tonanhngo has joined #openstack-kuryr | 02:35 | |
| *** tonanhngo has quit IRC | 02:37 | |
| openstackgerrit | dengshaolin proposed openstack/fuxi: Raise exception when find more than one matched Cinder volumes https://review.openstack.org/387937 | 02:44 |
| *** vikasc has joined #openstack-kuryr | 02:47 | |
| openstackgerrit | dengshaolin proposed openstack/fuxi: Raise exception when create from existing volume which is being usd by another cluster https://review.openstack.org/387929 | 02:47 |
| *** tonanhngo has joined #openstack-kuryr | 02:48 | |
| *** tonanhngo has quit IRC | 02:49 | |
| *** yuanying has quit IRC | 02:53 | |
| *** yuanying has joined #openstack-kuryr | 03:00 | |
| *** yamamoto_ has joined #openstack-kuryr | 03:02 | |
| *** shashank_hegde has quit IRC | 03:02 | |
| *** tonanhngo has joined #openstack-kuryr | 03:08 | |
| *** tonanhngo has quit IRC | 03:08 | |
| *** shashank_hegde has joined #openstack-kuryr | 03:09 | |
| *** limao has quit IRC | 03:16 | |
| *** limao has joined #openstack-kuryr | 03:17 | |
| *** tonanhngo has joined #openstack-kuryr | 03:21 | |
| *** limao has quit IRC | 03:21 | |
| *** tonanhngo has quit IRC | 03:22 | |
| *** dougbtv has quit IRC | 03:24 | |
| *** tonanhngo has joined #openstack-kuryr | 03:34 | |
| *** tonanhngo has quit IRC | 03:36 | |
| *** yuanying has quit IRC | 03:48 | |
| *** tonanhngo has joined #openstack-kuryr | 03:52 | |
| *** tonanhngo has quit IRC | 03:53 | |
| *** shashank_hegde has quit IRC | 03:53 | |
| *** tonanhngo has joined #openstack-kuryr | 04:05 | |
| *** tonanhngo has quit IRC | 04:07 | |
| *** yedongcan has quit IRC | 04:07 | |
| *** yedongcan has joined #openstack-kuryr | 04:08 | |
| *** tonanhngo has joined #openstack-kuryr | 04:18 | |
| *** tonanhngo has quit IRC | 04:20 | |
| *** tonanhngo has joined #openstack-kuryr | 04:31 | |
| *** tonanhngo has quit IRC | 04:33 | |
| *** tonanhngo has joined #openstack-kuryr | 04:47 | |
| *** limao has joined #openstack-kuryr | 04:48 | |
| *** tonanhngo has quit IRC | 04:48 | |
| *** yamamoto_ has quit IRC | 04:48 | |
| *** yamamoto_ has joined #openstack-kuryr | 04:50 | |
| *** limao_ has joined #openstack-kuryr | 04:51 | |
| *** limao has quit IRC | 04:52 | |
| *** yuanying has joined #openstack-kuryr | 04:54 | |
| *** tonanhngo has joined #openstack-kuryr | 05:00 | |
| *** tonanhngo has quit IRC | 05:01 | |
| *** yedongcan has quit IRC | 05:08 | |
| *** tonanhngo has joined #openstack-kuryr | 05:13 | |
| *** tonanhngo has quit IRC | 05:15 | |
| *** yedongcan has joined #openstack-kuryr | 05:17 | |
| *** saneax-_-|AFK is now known as saneax | 05:26 | |
| *** tonanhngo has joined #openstack-kuryr | 05:26 | |
| *** tonanhngo has quit IRC | 05:27 | |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork: Fix missing subnetpool_id in neutron subnet https://review.openstack.org/387835 | 05:33 |
| *** janki has joined #openstack-kuryr | 05:34 | |
| *** tonanhngo has joined #openstack-kuryr | 05:39 | |
| *** tonanhngo has quit IRC | 05:40 | |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork: Fix missing subnetpool_id in neutron subnet https://review.openstack.org/387835 | 05:44 |
| *** irenab has joined #openstack-kuryr | 05:50 | |
| *** tonanhngo has joined #openstack-kuryr | 05:52 | |
| *** tonanhngo has quit IRC | 05:53 | |
| *** tonanhngo has joined #openstack-kuryr | 06:06 | |
| *** tonanhngo has quit IRC | 06:07 | |
| apuimedo | hongbin: https://review.openstack.org/#/c/402244/4 | 06:08 |
| apuimedo | I posted a review after trying it out with my f24 | 06:09 |
| hongbin | apuimedo: ack, let me check | 06:09 |
| apuimedo | thanks | 06:09 |
| * apuimedo going back to bed | 06:09 | |
| hongbin | apuimedo: i tested it in ubuntu:14.04 | 06:09 |
| hongbin | apuimedo: however, will test it in other distro as well | 06:10 |
| hongbin | apuimedo: thanks for the review | 06:10 |
| *** tonanhngo has joined #openstack-kuryr | 06:19 | |
| *** tonanhngo has quit IRC | 06:20 | |
| *** hongbin has quit IRC | 06:21 | |
| *** tonanhngo has joined #openstack-kuryr | 06:32 | |
| *** yamamoto_ has quit IRC | 06:34 | |
| *** tonanhngo has quit IRC | 06:34 | |
| *** yamamoto_ has joined #openstack-kuryr | 06:36 | |
| *** tonanhngo has joined #openstack-kuryr | 06:45 | |
| *** tonanhngo has quit IRC | 06:46 | |
| *** shashank_hegde has joined #openstack-kuryr | 06:47 | |
| *** yuanying_ has joined #openstack-kuryr | 06:52 | |
| *** yuanying has quit IRC | 06:55 | |
| *** tonanhngo has joined #openstack-kuryr | 06:58 | |
| *** tonanhngo has quit IRC | 06:59 | |
| *** limao_ has quit IRC | 07:08 | |
| *** yedongcan has quit IRC | 07:08 | |
| *** tonanhngo has joined #openstack-kuryr | 07:11 | |
| *** tonanhngo has quit IRC | 07:13 | |
| *** oanson has joined #openstack-kuryr | 07:16 | |
| openstackgerrit | Merged openstack/kuryr: Delete deprecated Hacking in tox.ini https://review.openstack.org/406545 | 07:16 |
| irenab | apuimedo: ivc_ ping | 07:18 |
| *** yedongcan has joined #openstack-kuryr | 07:19 | |
| *** dimak has joined #openstack-kuryr | 07:44 | |
| *** limao has joined #openstack-kuryr | 07:50 | |
| *** limao has quit IRC | 08:01 | |
| *** openstackgerrit has quit IRC | 08:03 | |
| *** yuanying_ has quit IRC | 08:19 | |
| *** dimak has quit IRC | 08:26 | |
| apuimedo | irenab: pong | 08:28 |
| irenab | apuimedo: good morning | 08:29 |
| apuimedo | irenab: boker tov | 08:29 |
| irenab | apuimedo: :-) | 08:29 |
| irenab | apuimedo: shortquestion on devstack with k8s | 08:30 |
| apuimedo | go ahead | 08:30 |
| irenab | I deployed the one with ivc patch (quite sure it isnot related), and stack.sh clains to fail since cannot start docker service. It is already running. In general everything seems to be working, so not sure if its a bug. did you see the same? | 08:31 |
| apuimedo | irenab: that's an unstack bug IIRC | 08:31 |
| apuimedo | sometimes it happens, and the next stacking goes to hell | 08:32 |
| apuimedo | irenab: mind filing a bug. I'll improve it | 08:32 |
| apuimedo | as a workaround. Just unstack and manually uninstall docker | 08:32 |
| irenab | apuimedo: actually it seems that all other services are fine, thats why I am not sure how to describe the bug | 08:32 |
| apuimedo | irenab: ok. I'll investigate a bit more, usually I just work around it :P | 08:34 |
| irenab | apuimedo: just not sure ifsuch setup is expected to work. so to workaround you justkull running docker and run the one from devstack screen? | 08:36 |
| *** dimak has joined #openstack-kuryr | 08:37 | |
| *** shashank_hegde has quit IRC | 08:38 | |
| apuimedo | irenab: I usually stack again | 08:38 |
| apuimedo | because it will need to launch the other containers | 08:38 |
| apuimedo | etcd, k8s, etc | 08:38 |
| irenab | another issue I saw is that kubelet is not started since it requires nsenter | 08:39 |
| irenab | had to add it manually | 08:39 |
| apuimedo | irenab: that's very odd. It started for me | 08:46 |
| apuimedo | I saw it in kubectl get nodes | 08:46 |
| apuimedo | which distro do you use for devstack? | 08:46 |
| irenab | ubuntu14.04 | 08:47 |
| irenab | actually I too the vagrant from the libnetwork and just reduced a bit from it | 08:47 |
| *** palexster has quit IRC | 08:50 | |
| *** lmdaly has joined #openstack-kuryr | 08:51 | |
| *** palexster has joined #openstack-kuryr | 08:55 | |
| *** limao has joined #openstack-kuryr | 08:57 | |
| apuimedo | aha | 08:59 |
| apuimedo | I used a new f24 | 08:59 |
| irenab | apuimedo: so on f24 nsenter was already installed? | 09:00 |
| irenab | I will fill a bug for now | 09:01 |
| *** irenab_ has joined #openstack-kuryr | 09:05 | |
| *** ram_____ has joined #openstack-kuryr | 09:05 | |
| apuimedo | I suppose | 09:05 |
| *** irenab has quit IRC | 09:06 | |
| *** irenab_ is now known as irenab | 09:06 | |
| apuimedo | irenab: yes. It seems so. It was provided by util-linux-2.28.2-1.fc24.x86_64 | 09:06 |
| *** yuanying has joined #openstack-kuryr | 09:08 | |
| ram_____ | apuimedo: Hi. I am trying to deploy basic devstack- kuryr. But I was facing issue. Issue info: http://paste.openstack.org/show/591390/. Can you please help me in this. | 09:09 |
| *** openstackgerrit has joined #openstack-kuryr | 09:11 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/373745 | 09:11 |
| * apuimedo checking, ram_____ | 09:11 | |
| apuimedo | ram_____: I'll make a new deployment to verify | 09:12 |
| yedongcan | limao: ping | 09:13 |
| limao | yedongcan: hi | 09:14 |
| ram_____ | apuimedo : Thanks. I have taken Ubuntu 16.04 | 09:14 |
| yedongcan | limao: I had updated the patch today, you can review it if you have free time. | 09:15 |
| limao | yedongcan: sure, thanks | 09:15 |
| yedongcan | limao: thanks. | 09:15 |
| yedongcan | apuimedo, irenab: Hi , what's your opinions of this patch: https://review.openstack.org/#/c/404591/ | 09:18 |
| irenab | yedongcan: will check in 40 mins, currently at the meeting | 09:18 |
| yedongcan | irenab: thanks | 09:19 |
| *** reedip has quit IRC | 09:22 | |
| *** garyloug has joined #openstack-kuryr | 09:24 | |
| apuimedo | yedongcan: does it happen on 16.04 or only on 14.04? | 09:24 |
| yedongcan | apuimedo: the gate is running on 14.04 | 09:28 |
| yedongcan | apuimedo: http://logs.openstack.org/35/387835/4/check/gate-kuryr-libnetwork-dsvm-rally-nv/0576c1f/ | 09:29 |
| apuimedo | yedongcan: maybe we could try if it happens on 16.04 and move it to that | 09:30 |
| yedongcan | apuimedo: sure, we can try it in local, but we can't update gate machine. | 09:32 |
| apuimedo | yedongcan: but we can change which node is used to one with 16.06 IIRC | 09:33 |
| irenab | apuimedo: is there any ‘official’ distro versions that kuryr is supposed to work on? It should be the same as openstack, right? | 09:36 |
| yedongcan | apuimedo: 14.04 sometimes can run successful, http://logs.openstack.org/65/400365/7/check/gate-kuryr-libnetwork-dsvm-rally-nv/06789a7/console.html | 09:38 |
| apuimedo | 16.04 and f24 are the ones I care the most | 09:38 |
| apuimedo | pre systemd... not so much | 09:39 |
| apuimedo | irenab: ^^ | 09:39 |
| irenab | apuimedo: got it, thanks. Probably we should bump the vagrant versions | 09:40 |
| apuimedo | we should | 09:40 |
| * apuimedo doesn't use vagrant | 09:40 | |
| apuimedo | /q vikasc | 09:43 |
| apuimedo | vikasc: ping | 09:43 |
| vikasc | apuimedo, pong | 09:43 |
| vikasc | apuimedo, u have a very long life | 09:43 |
| apuimedo | ? | 09:43 |
| vikasc | apuimedo, just now was thinking to ping u | 09:43 |
| apuimedo | I replied to one of the comments you put on ivc_'s cni patch | 09:44 |
| vikasc | apuimedo, just saw | 09:44 |
| apuimedo | ;-) | 09:44 |
| openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-libnetwork: Nested-Containers: trunk subports management https://review.openstack.org/402462 | 09:56 |
| *** pmannidi has quit IRC | 09:59 | |
| *** reedip has joined #openstack-kuryr | 10:00 | |
| *** tonanhngo has joined #openstack-kuryr | 10:01 | |
| *** tonanhngo has quit IRC | 10:03 | |
| apuimedo | ram_____: I just did a redeploy | 10:09 |
| apuimedo | and it ended successfully | 10:09 |
| apuimedo | ram_____: you may have been unlucky when pulling neutron-lib and neutron | 10:10 |
| apuimedo | for reference, my local.conf is | 10:10 |
| apuimedo | ram_____: https://paste.fedoraproject.org/499599/14809327/ | 10:12 |
| *** limao has quit IRC | 10:14 | |
| ram_____ | apuimedo: Oh. Thank you very much. I got the point. I used neutron plugin extra. I will try with your conf. | 10:16 |
| apuimedo | thanks | 10:16 |
| apuimedo | irenab: ivc_: The more I think about it, the more I want to use k8s' third party resources for making neutron notifications end up in k8s and just have normal k8s watchers instead of polling neutron | 10:18 |
| *** yedongcan has left #openstack-kuryr | 10:24 | |
| *** irenab has quit IRC | 10:25 | |
| *** yamamoto_ has quit IRC | 10:27 | |
| *** gsagie has joined #openstack-kuryr | 10:27 | |
| *** yamamoto has joined #openstack-kuryr | 10:43 | |
| *** irenab has joined #openstack-kuryr | 10:50 | |
| irenab | apuimedo: do you want neutron notifier to post on k8s-api? | 10:50 |
| apuimedo | irenab: That's what I'm considering | 10:51 |
| ivc_ | apuimedo do you mean that we would replace 'annotations' with ThirdPartyResource with custom fields? or keep 'annotations' but move them to ThirdPartyResource? | 10:51 |
| apuimedo | ivc_: I'm still in that | 10:51 |
| apuimedo | deciding if we do a mixed approach | 10:51 |
| apuimedo | s/if we do/if I propose/ | 10:52 |
| apuimedo | replacing completely would be problematic | 10:52 |
| irenab | apuimedo: for mixed approach, you mean daemon + slim CNI? | 10:52 |
| ivc_ | apuimedo, irenab, i'd prefer if instead of using neutron notifier we make another daemon that will use neutron's MQ and translate events into ThirdPartyResource | 10:52 |
| apuimedo | irenab: ivc_: I mean mixed as in, some things go in normal annotations, some others, like lbaasv2 listeners and pools go into third party resources | 10:53 |
| irenab | ivc_: apuimedo : need to think a bit on this. I am still recovering from illness, so thinking may take a while:-) | 10:53 |
| apuimedo | ivc_: I don't really care all that much, what is the source that ends up putting stuff on third party resources | 10:53 |
| ivc_ | i mean in neutron we have 2 facilities - notifier (same as nova; could be problematic to implement as it affects neutron codebase) and notifications (used by ceilometer and imo designed for our case) | 10:53 |
| apuimedo | irenab: the trick is not recovering | 10:53 |
| * apuimedo still with a big cold | 10:53 | |
| apuimedo | xD | 10:53 |
| apuimedo | ivc_: I know | 10:53 |
| apuimedo | I'm now checking what's on octavia's, now that the neutron lbaas agent moved there | 10:54 |
| apuimedo | depending on what's there, it may not even be necessary to take a decision | 10:54 |
| apuimedo | because maybe it was taken for us already | 10:55 |
| ivc_ | apuimedo what 'decision' are you talking about here? xD | 10:55 |
| apuimedo | ivc_: notifier vs rpc | 10:56 |
| apuimedo | it's likely that octavia only has the rpc | 10:56 |
| ivc_ | notifier is for nova and port activity only | 10:56 |
| apuimedo | so nova notifier most likely just goes out of the window | 10:56 |
| apuimedo | exactly | 10:56 |
| irenab | apuimedo: if lbas is separate project, it should not have RPC channel with neutron, only use its ‘official’ apis | 10:56 |
| apuimedo | irenab: IIUC both neutron and lbaas publish to rpc | 10:56 |
| irenab | REAST + notifications | 10:57 |
| apuimedo | we'll have to subscribe to both | 10:57 |
| apuimedo | irenab: reast? | 10:57 |
| irenab | REST | 10:57 |
| ivc_ | we could have rpc (tho i don't think rpc is the right term here, its just MQ notification) as primary choice and polling as fall-back | 10:57 |
| apuimedo | ah | 10:57 |
| irenab | my understanding is that RPC is internal channel between neutron servers and neutron agents | 10:58 |
| ivc_ | irenab, thats why i say RPC is not the right term here :) | 10:58 |
| apuimedo | ivc_: short term makes sense. Long term I really prefer that the watcher only needs to check resources and third party resources | 10:58 |
| apuimedo | for events | 10:58 |
| irenab | ivc_: by MQ you amly that neutron will have to implement something? | 10:58 |
| ivc_ | we are talking about http://docs.openstack.org/admin-guide/networking-adv-operational-features.html | 10:58 |
| apuimedo | irenab: ivc_: sorry. i meant MQ notifications | 10:58 |
| irenab | the ones that ceilometer use? | 10:59 |
| ivc_ | apuimedo, irenab, the link above is about MQ notifications (used by ceilometer) | 10:59 |
| ivc_ | note that it is not enabled by default and it would be nice to support polling as fallback imo | 11:00 |
| apuimedo | if it's not default, we may very well have to have a fallback | 11:01 |
| irenab | apuimedo: ivc_ :so to catch up on your proposal, you suggest to add 3rd party resource that will be created/updated onceproper notification is received? | 11:02 |
| irenab | or polling done | 11:02 |
| apuimedo | irenab: that's what I'm thinking about, yes | 11:02 |
| apuimedo | that listener and pool notifications will end up in third party resources | 11:02 |
| irenab | apuimedo: what is the advantage versus annotation approach? | 11:03 |
| irenab | apuimedo: is it only for service support oralso for pod? | 11:03 |
| apuimedo | irenab: we could just add more data into the service annotation, of course, I just think that it is probably not relevant | 11:03 |
| apuimedo | but! | 11:04 |
| apuimedo | I may be wrong | 11:04 |
| apuimedo | and we may end up doing it in annotations | 11:04 |
| irenab | apuimedo: ivc_ I probably missed when you started service discussion, feeling a bit in the middle | 11:04 |
| apuimedo | and putting listeners on the services annotation and pool in the endpoints annotation | 11:04 |
| apuimedo | irenab: heh... We didn't start... That was me just having to think instead of type due to holding the baby a while this morning | 11:05 |
| irenab | apuimedo: :-) | 11:05 |
| apuimedo | in any case, the important part here is. I'd really prefer if we can have another component to post/update K8s resources (or thirdparty resources) and the watcher process only needs to 'watch' | 11:06 |
| irenab | apuimedo: so you want the processwide split, right? since from the SW perspective, watcher invokes handlers,so there is sort of separation | 11:07 |
| irenab | ivc_ has incredibly generic model there | 11:07 |
| apuimedo | irenab: well, it's not really splitting the process in that axis | 11:08 |
| apuimedo | it's adding another daemon that receives notifications | 11:08 |
| apuimedo | and patches resources | 11:08 |
| apuimedo | completely orthogonal and independent to the watches/handlers | 11:08 |
| irenab | apuimedo: so daemin for read and daemon for write? | 11:09 |
| irenab | additional daemon only for notifications or polling too? | 11:09 |
| apuimedo | irenab: I would make the same daemon for both approaches | 11:10 |
| apuimedo | think of it as a MQ driver and a polling driver | 11:10 |
| apuimedo | and that in the end it always goes to K8s to update some resource | 11:10 |
| irenab | I am not experienced with Octavia, but probably pool lifecycle has states, so you need to wait till it becomes Active, right? | 11:10 |
| *** reedip has quit IRC | 11:10 | |
| apuimedo | irenab: I'm trying to find out if they send MQ notifications | 11:10 |
| apuimedo | for their resources | 11:10 |
| irenab | and you prefer to keep this async? | 11:11 |
| apuimedo | if they don't, we have to fallback to polling | 11:11 |
| apuimedo | with disappointment | 11:11 |
| irenab | apuimedo: what is the main drive for separate daemon? | 11:12 |
| irenab | keep it simple? | 11:12 |
| apuimedo | irenab: deployment | 11:12 |
| apuimedo | and simplicity | 11:12 |
| apuimedo | I don't want temptations to couple things afterwards | 11:12 |
| irenab | just more to take care for HA | 11:13 |
| irenab | but may be reasonable approach, expecially with all the micro services trend | 11:14 |
| apuimedo | that's ture | 11:14 |
| apuimedo | *true | 11:14 |
| apuimedo | it is a tradeoff indeed | 11:14 |
| *** reedip has joined #openstack-kuryr | 11:24 | |
| apuimedo | irenab: ivc_: Octavia has an Event Streamer that is stevedore based | 11:24 |
| irenab | apuimedo: nice | 11:24 |
| apuimedo | this means that we could probably have our own driver for it | 11:24 |
| apuimedo | that goes straight to k8s | 11:24 |
| irenab | so either need to know what object to patch or create new one | 11:25 |
| apuimedo | irenab: https://github.com/openstack/octavia/blob/master/octavia/controller/queue/endpoint.py | 11:25 |
| apuimedo | irenab: well, my idea of the thirdparty resources was the following | 11:26 |
| apuimedo | from the services handler, when you create, let's say a listener | 11:26 |
| apuimedo | you know its UUID | 11:26 |
| apuimedo | then you can watch for a thirdparty resource with that UUID as name | 11:27 |
| apuimedo | from the stevedore driver | 11:27 |
| apuimedo | (this would not be necessary if octavia objects had tags, obviously | 11:27 |
| *** reedip has quit IRC | 11:28 | |
| apuimedo | it seems it may only be for the health manager | 11:33 |
| apuimedo | I don't know if it always run | 11:33 |
| openstackgerrit | vikas choudhary proposed openstack/kuryr-libnetwork: Delay neutron extension checks to first neutron interaction https://review.openstack.org/403325 | 11:36 |
| apuimedo | devstack seems to run it | 11:36 |
| irenab | apuimedo: wecan do the same as with networkbefore tags, use its name | 11:39 |
| apuimedo | irenab: maybe | 11:39 |
| *** reedip has joined #openstack-kuryr | 11:41 | |
| irenab | need to check the model, hope listener has name attribute | 11:41 |
| apuimedo | irenab: it does | 11:42 |
| irenab | apuimedo: anyway, its just implementation details. with octavia notifier it is possible to make clean separation | 11:43 |
| apuimedo | irenab: it even has description :-) | 11:44 |
| apuimedo | (and l7 policies that may be usable in the future for ingress controller) | 11:45 |
| irenab | apuimedo: was also thinking about it :-) | 11:46 |
| apuimedo | •_•) | 11:47 |
| apuimedo | ( •_•)>⌐■-■, | 11:47 |
| apuimedo | (⌐■_■) | 11:47 |
| gsagie | snow man? | 11:48 |
| irenab | with a gun? | 11:48 |
| *** gsagie has quit IRC | 11:49 | |
| apuimedo | no! | 11:50 |
| apuimedo | irenab: https://33.media.tumblr.com/8d4adac905b4c9155ba64a4a2b8ef6ca/tumblr_nbam09PAtE1spxz7uo1_250.gif | 11:51 |
| irenab | :-) | 11:51 |
| apuimedo | more inspired on https://media.giphy.com/media/v9rfTQBNqdsSA/giphy.gif | 11:51 |
| *** limao has joined #openstack-kuryr | 11:58 | |
| *** limao has quit IRC | 12:03 | |
| ivc_ | apuimedo, irenab, sry was on a meeting for the past hour, so wasn't able to respond. i think using k8s 3rd party resources makes a lot of sense for both lbaas and ports | 12:03 |
| irenab | ivc_: trying out your latest patch | 12:04 |
| ivc_ | that allows us to have full control over them. might also solve the current 'leftover hybrid bridge' problem we have now due to cni being called after pod is deleted from k8s db | 12:05 |
| irenab | ivc_: if I want more detals on k8s service log, what flag do I add? | 12:05 |
| ivc_ | irenab, uhm, what? :) | 12:05 |
| ivc_ | you mean from cni? | 12:06 |
| irenab | now, the controller | 12:06 |
| irenab | I see port being created and active, but kubelet fails | 12:07 |
| irenab | docker_manager.go:357] NetworkPlugin cni failed on the status hook for pod 'my-nginx-379829228-z1at9' - Unexpected command output Device "eth0" does not exist. | 12:07 |
| ivc_ | for the controller just add '-d' after 'run_server.py' | 12:08 |
| ivc_ | for the cni, add "'debug': true" to 10-kuryr.conf | 12:08 |
| ivc_ | you probably want the 'cni' debug | 12:09 |
| irenab | ivc_: probably both. thanks | 12:10 |
| *** limao has joined #openstack-kuryr | 12:11 | |
| *** limao has quit IRC | 12:21 | |
| *** limao has joined #openstack-kuryr | 12:21 | |
| ivc_ | irenab, are you stacking on a clean env? also do you 'kubectl run' with a single replica? | 12:22 |
| irenab | ivc_: now run again stack.sh, previously single pod wasnot running, but saw neutron port created for it with Active status | 12:23 |
| irenab | ivc_: something strange. I just had new stack completed, but see kubelet comlain on not being able to plug pod | 12:25 |
| ivc_ | irenab, kk. thats weird actually, by the time it becomes 'active' veth should already be created. if it happens again, pls try 'kubectl exec ... ip a' and check if 'eth0' is there | 12:26 |
| ivc_ | irenab, apuimedo, vikasc, i can confirm that https://review.openstack.org/#/c/404038/6/devstack/plugin.sh@79 is indeed broken | 12:27 |
| apuimedo | ivc_: so that means that it should the link should be done on postconfig | 12:30 |
| ivc_ | apuimedo yup, or we can hardcode '/usr/local/bin/kuryr-cni' instead of 'which' | 12:31 |
| apuimedo | ivc_: nope. It's distro dependant | 12:31 |
| ivc_ | true | 12:31 |
| apuimedo | and the path you put is a result of debian's/Ubuntu's screwed patching of distutils | 12:31 |
| apuimedo | it always bothered me that they change distutils to point to local | 12:32 |
| ivc_ | :) | 12:32 |
| apuimedo | I have some python code to detect it, if you want | 12:32 |
| apuimedo | so instead of which, this oneliner prints the path that it will have | 12:32 |
| ivc_ | i actually like it that they use /usr/local | 12:32 |
| apuimedo | when I'm root and I install with pip it pisses me off | 12:33 |
| ivc_ | well, /usr/bin is for .deb and /usr/local for custom stuff. makes sense to me :) | 12:33 |
| apuimedo | I'm not saying it doesn't make sense, I say I don't like it | 12:34 |
| apuimedo | maybe if pip supported parameters like --prefix | 12:34 |
| apuimedo | :P | 12:34 |
| *** limao has quit IRC | 12:36 | |
| irenab | ivc_: http://paste.openstack.org/show/591412/ | 12:38 |
| irenab | ivc_: log message on kubelet is terrible, very difficult to find the issue with tons of content | 12:40 |
| irenab | ivc_: intersting that the same logging blockis on pod describe | 12:45 |
| *** limao has joined #openstack-kuryr | 12:51 | |
| ivc_ | irenab, i think i've seen those '\n' as real linebreaks in my logs. i'm restacking now | 12:51 |
| ivc_ | irenab, do you use ubuntu or centos for devstack? | 12:52 |
| irenab | ubuntu 14.04 | 12:52 |
| irenab | shall I move to 16.04? | 12:52 |
| ivc_ | i'm using 16.04; it should not matter tbh | 12:53 |
| *** garyloug has quit IRC | 12:54 | |
| ivc_ | i'm a bit worried about 'describe' those log messages should go to stderr and for 'describe' i would expect the CNI-formatted stdout errors | 12:54 |
| *** dougbtv has joined #openstack-kuryr | 12:54 | |
| irenab | as for the kubelet failure, it seems to be at https://review.openstack.org/#/c/404038/6/kuryr_kubernetes/cni/binding/bridge.py line 36-37 | 12:55 |
| irenab | ivc_: yes, in ‘pod describe’ it looks very unexpeted , and appears few times | 12:56 |
| irenab | ivc_: looks like this http://paste.openstack.org/show/591414/ | 12:58 |
| ivc_ | irenab, yeah. it looks like we got oslo_log in stdout instead of stderr | 12:59 |
| ivc_ | irenab, now i wonder how did it work for me http://docs.openstack.org/developer/oslo.log/opts.html#DEFAULT.use_stderr | 13:01 |
| ivc_ | as it is stdout by default >.< | 13:02 |
| ivc_ | apuimedo btw, symlink creation is done in 'post-config', does that mean that 'post-config' is called before 'install'? | 13:03 |
| ivc_ | apuimedo i'll just move symlink to 'install' anyway | 13:04 |
| *** reedip has quit IRC | 13:11 | |
| ivc_ | apuimedo in fact, 'ln -s' fails not due to 'which', but due to the lack of 'CNI_BIN_DIR' | 13:11 |
| *** lmdaly has quit IRC | 13:12 | |
| apuimedo | ivc_: shouldn't | 13:12 |
| *** limao has quit IRC | 13:12 | |
| apuimedo | heh | 13:12 |
| ivc_ | CNI_BIN_DIR is created in your 'extract_hyperkube' | 13:12 |
| apuimedo | ivc_: Maybe we should use install instead of symlink | 13:12 |
| *** limao has joined #openstack-kuryr | 13:12 | |
| *** irenab has quit IRC | 13:13 | |
| ivc_ | apuimedo but use 'which'? | 13:13 |
| apuimedo | I don't mind ending with two copies | 13:13 |
| apuimedo | in the end they will only call the library in /opt/stack/kuryr-kubernetes | 13:13 |
| ivc_ | apuimedo, irenab, we need fullstack gates ... | 13:23 |
| *** reedip has joined #openstack-kuryr | 13:24 | |
| *** yamamoto has quit IRC | 13:26 | |
| *** janki has quit IRC | 13:31 | |
| *** limao_ has joined #openstack-kuryr | 13:34 | |
| *** irenab has joined #openstack-kuryr | 13:35 | |
| *** limao has quit IRC | 13:36 | |
| apuimedo | ivc_: I have to fix my patch :/ | 13:37 |
| ivc_ | apuimedo what? | 13:37 |
| ivc_ | apuimedo oh, the one for gate you mean? | 13:38 |
| apuimedo | ivc_: yes. The one that runs the installation | 13:39 |
| apuimedo | and then, after your patch | 13:39 |
| apuimedo | I'll make one that tries starting to containers | 13:39 |
| ivc_ | apuimedo, irenab, i've fixed issues with logs and ln. gonna push for review now. just need to verify once more | 13:40 |
| *** yamamoto has joined #openstack-kuryr | 13:44 | |
| apuimedo | ivc_: I'll check it too when you push | 13:45 |
| apuimedo | ivc_: irenab: I'd like to merge it today | 13:45 |
| *** gsagie has joined #openstack-kuryr | 13:46 | |
| *** lmdaly has joined #openstack-kuryr | 13:49 | |
| apuimedo | ltomasbo: do the vlan devices remain still or are they properly removed? | 13:51 |
| apuimedo | with https://review.openstack.org/#/c/402462/ | 13:51 |
| ltomasbo | hi apuimedo | 13:52 |
| ltomasbo | in meeting right now, so I may answer with delay | 13:52 |
| ltomasbo | but yes, they work fine right now | 13:52 |
| ltomasbo | I included the driver detach in the patch as irenab suggested | 13:53 |
| ltomasbo | but now it seems it is colliding with the other patch which also splits the port binding drivers out | 13:53 |
| ivc_ | irenab, apuimedo: here's what i get in kubelet.log: http://paste.openstack.org/show/591429/ | 13:53 |
| ltomasbo | not sure about the proper order of both patches | 13:53 |
| *** limao_ has quit IRC | 13:54 | |
| ivc_ | irenab, apuimedo seems to work, but not sure how to interpret: E1205 05:49:27.553765 125394 docker_manager.go:357] NetworkPlugin cni failed on the status hook for pod 'nginx-701339712-aj9pg' - Unexpected command output Device "eth0" does not exist. | 13:54 |
| ivc_ | with error: exit status 1 | 13:54 |
| *** limao has joined #openstack-kuryr | 13:55 | |
| apuimedo | ivc_: and the device is there? | 13:56 |
| ivc_ | yup | 13:56 |
| apuimedo | ivc_: you got nsenter? | 13:56 |
| ivc_ | yes | 13:57 |
| ivc_ | apuimedo, irenab check L111-112 in http://paste.openstack.org/show/591429/ | 13:57 |
| apuimedo | aha! | 13:58 |
| apuimedo | even worse | 13:58 |
| apuimedo | I wonder WTH they use ethtool | 13:58 |
| ivc_ | what bothers me is that 'cni error' is BEFORE our cni is called | 13:58 |
| apuimedo | you mean line 83? | 13:59 |
| ivc_ | yes | 13:59 |
| ivc_ | check timestamps in L83 and L86 | 13:59 |
| * apuimedo checking | 13:59 | |
| ivc_ | L86 is stderr of our cni | 13:59 |
| ivc_ | so i wonder WTF that L83 is | 14:00 |
| * apuimedo checking k8s code | 14:00 | |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: Experimental CNI & VIFBridge binding https://review.openstack.org/404038 | 14:01 |
| ivc_ | apuimedo, irenab, done: https://review.openstack.org/#/c/404038/6..7 | 14:02 |
| irenab | ivc_: checking | 14:03 |
| ivc_ | also updated 10-kuryr.conf to include 'debug': true by default | 14:03 |
| *** yedongcan has joined #openstack-kuryr | 14:03 | |
| mchiappero | meeting time? | 14:03 |
| ivc_ | imo makes sense while we are in 'development' phase | 14:03 |
| mchiappero | :P | 14:03 |
| irenab | ivc_: need to restack or restart k8s-kuryr and kubelet is enough? | 14:03 |
| ivc_ | yup | 14:03 |
| *** tonanhngo has joined #openstack-kuryr | 14:03 | |
| ivc_ | irenab, restart should do it | 14:03 |
| *** tonanhngo has quit IRC | 14:04 | |
| ivc_ | but make sure you got /opt/stack/cni/bin/kuryr-cni | 14:04 |
| *** garyloug has joined #openstack-kuryr | 14:04 | |
| *** oanson has quit IRC | 14:04 | |
| ivc_ | and run 'ln -s $(which kuryr-cni) /opt/stack/cni/bin/kuryr-cni' if you dont | 14:04 |
| irenab | I have | 14:05 |
| *** hongbin has joined #openstack-kuryr | 14:09 | |
| irenab | ivc_: still get the trace in the ‘describe pod’ | 14:12 |
| ivc_ | irenab, oh you also need '[DEFAULT]use_stderr=true' in /etc/kuryr/kuryr.conf | 14:13 |
| irenab | ivc_: got it, thanks | 14:13 |
| ivc_ | plugin has it so restacking would have updated it | 14:13 |
| irenab | just restarted the services | 14:13 |
| ivc_ | that's not for controller (tho it affects it) but just for CNI | 14:14 |
| *** limao has quit IRC | 14:16 | |
| *** limao has joined #openstack-kuryr | 14:16 | |
| ivc_ | lol i've run '--replicas=40' and it killed OVS :) another reason to have daemon to limit the DoP :) | 14:18 |
| ivc_ | (thats unrelated to kuryr-k8s) | 14:19 |
| apuimedo | ivc_: what do you mean 'it killed ovs' ? | 14:19 |
| apuimedo | which part went to hell? | 14:19 |
| apuimedo | the daemon? | 14:19 |
| apuimedo | ! | 14:19 |
| ivc_ | my vm is 99% in systime :) | 14:19 |
| apuimedo | :O | 14:19 |
| apuimedo | that doesn't sound good | 14:19 |
| ivc_ | ovs-vsctl show, kubectl get pods, neutron port-list | 14:19 |
| ivc_ | neither of them work :) | 14:19 |
| ivc_ | i'm pretty sure i could do the same with nova | 14:19 |
| janonymous | :D sounds interesting :P | 14:20 |
| ivc_ | well maybe i cant do it with nova as nova is slow :) | 14:20 |
| apuimedo | xD | 14:21 |
| apuimedo | but come on, ovs must have tests that storm it much more | 14:22 |
| ivc_ | i love finding weak points and hammering them xD | 14:22 |
| apuimedo | something must have happened | 14:22 |
| ivc_ | apuimedo, neutron-agent has quite a bit of ERRORs too :) | 14:24 |
| apuimedo | that will be relevant to ajo's interests :-) | 14:25 |
| apuimedo | ivc_: some example? | 14:25 |
| ivc_ | 2016-12-05 06:24:33.919 119247 ERROR oslo.messaging._drivers.impl_rabbit [-] [31ad0a5f-d6b3-4aeb-9e0b-698a1f3f8f7b] AMQP server 172.16.240.46:5672 closed the connection. Check login credentials: Socket closed | 14:25 |
| ivc_ | 2016-12-05 06:24:42.324 119247 INFO neutron.plugins.ml2.drivers.openvswitch.agent.openflow.native.ovs_bridge [req-86cfb37b-5bb6-4714-a7a1-63c5afe1de11 - -] Bridge br-int changed its datapath-ID from e668938d8946 to 0000e668938d8946 | 14:25 |
| ivc_ | 2016-12-05 06:23:42.074 119247 ERROR neutron.plugins.ml2.drivers.openvswitch.agent.openflow.native.ofswitch [req-86cfb37b-5bb6-4714-a7a1-63c5afe1de11 - -] Switch connection timeout | 14:26 |
| ivc_ | first q-agt error: | 14:28 |
| ivc_ | 2016-12-05 06:09:23.420 119247 ERROR neutron.common.rpc [req-86cfb37b-5bb6-4714-a7a1-63c5afe1de11 - -] Ti | 14:28 |
| ivc_ | meout in RPC method update_device_list. Waiting for 1 seconds before next attempt. If the server is not d | 14:28 |
| ivc_ | own, consider increasing the rpc_response_timeout option as Neutron server(s) may be overloaded and unabl | 14:28 |
| ivc_ | e to respond quickly enough. | 14:28 |
| ivc_ | 2016-12-05 06:09:24.059 119247 ERROR neutron.plugins.ml2.drivers.openvswitch.agent.ovs_neutron_agent [req | 14:28 |
| ivc_ | -86cfb37b-5bb6-4714-a7a1-63c5afe1de11 - -] Error while processing VIF ports | 14:28 |
| apuimedo | ivc_: must be a timeout trying to get things from ovs | 14:28 |
| ivc_ | apuimedo could be | 14:28 |
| ivc_ | apuimedo but hey we only created 40 ports, right? :) | 14:29 |
| apuimedo | right | 14:29 |
| apuimedo | what I mean is, I think the ovs agent errors are a symptom | 14:29 |
| apuimedo | not the root cause | 14:29 |
| ivc_ | i think what causing this is the the storm on NETLINK from 40 CNI's IPDBs | 14:30 |
| apuimedo | 40 doesn't sound like much | 14:31 |
| apuimedo | for netlink | 14:31 |
| ivc_ | apuimedo 2016-12-05 06:08:23.340 128463 DEBUG pyroute2.netlink.nlsocket [-] Packet burst: delta=13 qsize=13 delay=0.01 get /usr/local/lib/python2.7/dist-packages/pyroute2/netlink/nlsocket.py:716 | 14:33 |
| ivc_ | :) | 14:33 |
| ivc_ | got a bunch of those | 14:34 |
| *** limao has quit IRC | 14:37 | |
| *** limao has joined #openstack-kuryr | 14:37 | |
| apuimedo | ivc_: maybe svinota can give us some feedback on that | 14:38 |
| irenab | ivc_: still having issues to run single pod. did you try on ubuntu 14.04? | 14:38 |
| apuimedo | irenab: you are doing archeology with that distro :P | 14:39 |
| ivc_ | irenab, nope, only on 16.04. ok i'll give it a try on a fresh 14.04 | 14:39 |
| apuimedo | xD | 14:39 |
| apuimedo | irenab: which errors do you get? | 14:39 |
| apuimedo | I'll try reproducing later | 14:39 |
| irenab | apuimedo: just wanted to run devstack and try the patch | 14:39 |
| apuimedo | irenab: I know | 14:40 |
| apuimedo | I was just making fun of trusty | 14:40 |
| apuimedo | irenab: did you post some pastebin above? | 14:40 |
| irenab | yes | 14:40 |
| irenab | ivc_: will try to run 16.04 later today/tomorrow morning and post +2 onceworks | 14:44 |
| ivc_ | kk. i'll see if it works on clean 14.04; maybe you got some issues with some leftovers from initial stack/unstack | 14:45 |
| apuimedo | ivc_: could be. I need to improve devstack to be a bit more corrosive | 14:47 |
| ivc_ | irenab, apuimedo, those UML, BPMN, MDA, SysML in https://www.modelio.org feel so enterprise-y | 14:47 |
| apuimedo | ivc_: Enterprise ready™ | 14:47 |
| apuimedo | xD | 14:47 |
| * ivc_ looks for apuimedo 'suicide' smiley | 14:51 | |
| * ivc_ y=ｰ( ﾟдﾟ)･∵. | 14:53 | |
| apuimedo | :D | 14:54 |
| apuimedo | you found it! | 14:54 |
| ivc_ | :P | 14:54 |
| apuimedo | (屮ﾟДﾟ)屮 | 14:54 |
| apuimedo | I have a few aliases in weechat :D | 14:54 |
| *** limao has quit IRC | 14:58 | |
| *** yedongcan has left #openstack-kuryr | 14:58 | |
| *** limao has joined #openstack-kuryr | 14:59 | |
| ivc_ | apuimedo, irenab btw, note L86-L110 in http://paste.openstack.org/show/591429/ - its just 4 seconde!!! | 15:01 |
| *** hongbin has quit IRC | 15:02 | |
| apuimedo | wth | 15:02 |
| apuimedo | that's terrible | 15:02 |
| apuimedo | xD | 15:02 |
| ivc_ | apuimedo, yeah, last week it was taking anything from 25 to 60 seconds | 15:12 |
| ivc_ | i wonder if i pulled some q-agt patch :) | 15:12 |
| ivc_ | or maybe i just got lucky with some some well aligned timings | 15:13 |
| apuimedo | ivc_: did we cause any trouble on k8s with the 40 watch cni processes? | 15:14 |
| ivc_ | apuimedo the whole vm went nuts, so yes :) | 15:15 |
| apuimedo | :P | 15:16 |
| apuimedo | how big was it? | 15:16 |
| ivc_ | 8g/2cpu | 15:16 |
| ivc_ | only 3g used | 15:17 |
| *** saneax is now known as saneax-_-|AFK | 15:17 | |
| ivc_ | but the fun thing is that its working :) it was 1/40 at start, now 1 hour later its 4/40 :) | 15:18 |
| *** limao has quit IRC | 15:19 | |
| *** reedip has quit IRC | 15:20 | |
| *** limao has joined #openstack-kuryr | 15:20 | |
| *** gsagie has quit IRC | 15:21 | |
| apuimedo | :O | 15:22 |
| apuimedo | that's ludicrous speed | 15:23 |
| ivc_ | its ok :) you can hardly expect anything when OVS is not responding well | 15:24 |
| ivc_ | but i'm puzzled as last week it was slow, but no issues with OVS, but this week its fast but has issues with OVS. gonna do another clean restack on both 14.04 and 16.04 to see if it is consistent | 15:25 |
| apuimedo | https://www.youtube.com/watch?v=ygE01sOhzz0 | 15:25 |
| ivc_ | xD | 15:26 |
| *** hongbin has joined #openstack-kuryr | 15:31 | |
| *** reedip has joined #openstack-kuryr | 15:32 | |
| *** oanson has joined #openstack-kuryr | 15:35 | |
| *** limao has quit IRC | 15:41 | |
| *** limao has joined #openstack-kuryr | 15:41 | |
| *** dougbtv_ has joined #openstack-kuryr | 15:59 | |
| *** dougbtv has quit IRC | 15:59 | |
| *** dougbtv_ is now known as dougbtv | 15:59 | |
| *** dougbtv_ has joined #openstack-kuryr | 16:00 | |
| *** limao has quit IRC | 16:02 | |
| *** limao has joined #openstack-kuryr | 16:02 | |
| *** yamamoto has quit IRC | 16:19 | |
| *** limao has quit IRC | 16:23 | |
| *** limao has joined #openstack-kuryr | 16:24 | |
| mchiappero | I just had a very quick look at the vlan driver, I would say there is a lot in common, I'll have another more in depth look later but I think we need to combine those changes somehow | 16:25 |
| mchiappero | I would use classes as you ltomasbo did, but I would primarily use our interface and see how to extend that so that can fit your driver too | 16:26 |
| mchiappero | But I guess we would need to either combine them and abandon one, or simply start with one and drop some pieces from the other | 16:27 |
| *** yuanying has quit IRC | 16:29 | |
| *** dimak has quit IRC | 16:33 | |
| *** janonymous has quit IRC | 16:34 | |
| ltomasbo | mchiappero, I agree they have a lot in common | 16:39 |
| ltomasbo | the main different I see is that yours in just focused in the port binding, while mine has also to do with some extra actions that needs to be done depending on the driver used | 16:40 |
| ltomasbo | but I agree they can be merge to try to cover both in just one | 16:41 |
| ltomasbo | mchiappero, let me know when you take a deeper look if you think it is easy to merge both driver models (perhaps inside my patch as it also cover the vlan kuryr-libnetwork functionality) | 16:42 |
| *** reedip has quit IRC | 16:43 | |
| mchiappero | one approach, that is disruptive for both is to have the base system first and then the drivers added | 16:45 |
| mchiappero | later | 16:45 |
| mchiappero | or some drivers added later | 16:45 |
| mchiappero | but we need the interface we have (I agree with moving to classes) | 16:45 |
| mchiappero | I guess you have less requirements there, I'll have a look later | 16:45 |
| ltomasbo | ok great | 16:46 |
| *** limao has quit IRC | 16:47 | |
| *** irenab has quit IRC | 16:50 | |
| *** irenab has joined #openstack-kuryr | 16:54 | |
| openstackgerrit | Darla Ahlert proposed openstack/fuxi: Move from "tenant" to "project" https://review.openstack.org/406189 | 16:55 |
| *** reedip has joined #openstack-kuryr | 16:56 | |
| *** irenab_ has joined #openstack-kuryr | 17:00 | |
| *** irenab has quit IRC | 17:01 | |
| *** irenab_ is now known as irenab | 17:01 | |
| *** reedip has quit IRC | 17:02 | |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: Experimental CNI & VIFBridge binding https://review.openstack.org/404038 | 17:02 |
| *** dougbtv_ has quit IRC | 17:07 | |
| *** reedip has joined #openstack-kuryr | 17:15 | |
| *** yamamoto has joined #openstack-kuryr | 17:20 | |
| *** yamamoto has quit IRC | 17:26 | |
| *** lmdaly_ has joined #openstack-kuryr | 17:31 | |
| *** lmdaly has quit IRC | 17:31 | |
| *** yuanying has joined #openstack-kuryr | 17:38 | |
| *** shashank_hegde has joined #openstack-kuryr | 18:02 | |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: devstack: add nsenter and fix setup-files race https://review.openstack.org/407119 | 18:03 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: Experimental CNI & VIFBridge binding https://review.openstack.org/404038 | 18:05 |
| *** reedip has quit IRC | 18:08 | |
| *** shashank_hegde has quit IRC | 18:10 | |
| *** lmdaly_ has quit IRC | 18:19 | |
| *** reedip has joined #openstack-kuryr | 18:21 | |
| *** garyloug has quit IRC | 18:22 | |
| *** reedip has quit IRC | 18:30 | |
| ivc_ | apuimedo updated, rebased and verified: https://review.openstack.org/#/c/404038/ | 18:35 |
| ivc_ | apuimedo also added https://review.openstack.org/#/c/407119/ | 18:35 |
| ivc_ | irenab, https://review.openstack.org/#/c/404038/ tested with both trusty and xenial (used local.conf tied to that specific patchset) | 18:36 |
| ivc_ | irenab, local.conf: http://paste.openstack.org/show/591462/ | 18:37 |
| *** reedip has joined #openstack-kuryr | 18:43 | |
| *** shashank_hegde has joined #openstack-kuryr | 18:47 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Add devstack plugin for Fuxi https://review.openstack.org/402244 | 19:00 |
| *** dougbtv has quit IRC | 19:08 | |
| *** dougbtv has joined #openstack-kuryr | 19:11 | |
| *** diogogmt has joined #openstack-kuryr | 19:15 | |
| *** dougbtv has quit IRC | 19:16 | |
| *** tonanhngo has joined #openstack-kuryr | 19:17 | |
| *** tonanhngo_ has joined #openstack-kuryr | 19:20 | |
| *** tonanhngo has quit IRC | 19:21 | |
| *** madgin has joined #openstack-kuryr | 19:22 | |
| *** tonanhngo_ has quit IRC | 19:25 | |
| *** mattmceuen has quit IRC | 19:28 | |
| *** tonanhngo has joined #openstack-kuryr | 19:31 | |
| *** tonanhngo has quit IRC | 19:36 | |
| *** dougbtv has joined #openstack-kuryr | 19:36 | |
| *** tonanhngo has joined #openstack-kuryr | 20:01 | |
| *** tonanhngo has quit IRC | 20:01 | |
| *** tonanhngo has joined #openstack-kuryr | 20:02 | |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-libnetwork: Fix the check of docker installation https://review.openstack.org/407172 | 20:04 |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-libnetwork: Fix the check of docker installation https://review.openstack.org/407172 | 20:05 |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-kubernetes: Fix the check of docker installation https://review.openstack.org/407179 | 20:07 |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Separate unit tests from fullstack tests https://review.openstack.org/403931 | 20:26 |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Add basic fullstack volume tests https://review.openstack.org/403941 | 20:26 |
| *** madgin is now known as mattmceuen | 20:40 | |
| *** oanson has quit IRC | 20:42 | |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittest with mock https://review.openstack.org/405203 | 21:30 |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittest with mock https://review.openstack.org/405203 | 21:37 |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: [WIP]Unittest with mock https://review.openstack.org/405203 | 22:13 |
| openstackgerrit | Marco Chiappero proposed openstack/kuryr-libnetwork: Fully support MAC address change on CreateEndpoint https://review.openstack.org/406636 | 22:30 |
| openstackgerrit | Marco Chiappero proposed openstack/kuryr-libnetwork: [WIP] Driver based model for kuryr-libnetwork https://review.openstack.org/400365 | 22:37 |
| *** pmannidi has joined #openstack-kuryr | 23:46 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
