| *** salv-orlando has quit IRC | 00:09 | |
| *** saneax is now known as saneax-_-|AFK | 00:54 | |
| *** limao has joined #openstack-kuryr | 00:55 | |
| *** yedongcan has joined #openstack-kuryr | 00:55 | |
| *** saneax-_-|AFK is now known as saneax | 01:42 | |
| *** yamamoto has joined #openstack-kuryr | 01:43 | |
| *** hongbin has joined #openstack-kuryr | 02:00 | |
| *** saneax is now known as saneax-_-|AFK | 02:46 | |
| *** hongbin has quit IRC | 02:46 | |
| *** hongbin has joined #openstack-kuryr | 02:48 | |
| *** salv-orlando has joined #openstack-kuryr | 03:10 | |
| *** salv-orlando has quit IRC | 03:15 | |
| *** aojea has joined #openstack-kuryr | 03:31 | |
| *** aojea has quit IRC | 03:35 | |
| *** janonymous has joined #openstack-kuryr | 03:37 | |
| *** hongbin has quit IRC | 03:59 | |
| *** janki has joined #openstack-kuryr | 05:01 | |
| *** salv-orlando has joined #openstack-kuryr | 05:11 | |
| *** salv-orlando has quit IRC | 05:15 | |
| irenab | mchiappero, lets meet there! | 05:42 |
|---|---|---|
| *** janonymous has quit IRC | 05:44 | |
| *** salv-orlando has joined #openstack-kuryr | 06:12 | |
| *** salv-orlando has quit IRC | 06:17 | |
| *** salv-orlando has joined #openstack-kuryr | 06:26 | |
| *** saneax-_-|AFK is now known as saneax | 06:34 | |
| *** aojea has joined #openstack-kuryr | 06:38 | |
| *** salv-orl_ has joined #openstack-kuryr | 07:23 | |
| *** salv-orlando has quit IRC | 07:26 | |
| irenab | ltomasbo|away, hi | 07:28 |
| *** pcaruana has joined #openstack-kuryr | 07:50 | |
| *** salv-orl_ has quit IRC | 08:10 | |
| *** ltomasbo|away is now known as ltomasbo | 08:14 | |
| ltomasbo | irenab, hi! | 08:14 |
| *** daidv has joined #openstack-kuryr | 08:17 | |
| *** pmannidi has quit IRC | 08:29 | |
| *** salv-orlando has joined #openstack-kuryr | 08:54 | |
| *** janonymous has joined #openstack-kuryr | 08:56 | |
| *** garyloug has joined #openstack-kuryr | 09:02 | |
| *** yedongcan has quit IRC | 09:19 | |
| *** yedongcan has joined #openstack-kuryr | 09:20 | |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork master: Add fullstack test for IPv6 subnet https://review.openstack.org/445376 | 09:24 |
| *** limao has quit IRC | 09:27 | |
| irenab | ltomasbo, I wanted to ask you if there is some order to review the pools patches? | 09:28 |
| ltomasbo | hi irenab | 09:39 |
| ltomasbo | there is an initial one: | 09:39 |
| ltomasbo | https://review.openstack.org/#/c/436875 | 09:39 |
| ltomasbo | then a follow up for initial baremetal support | 09:39 |
| ltomasbo | https://review.openstack.org/#/c/436876/ | 09:40 |
| ltomasbo | and then, on the one hand there is the extension for the baremetal: https://review.openstack.org/#/c/436877/ | 09:40 |
| ltomasbo | and on the other hand the 2 related to the nested: | 09:40 |
| ltomasbo | https://review.openstack.org/#/c/436893/, https://review.openstack.org/#/c/436894/ | 09:40 |
| *** aojea has quit IRC | 10:00 | |
| *** salv-orlando has quit IRC | 10:18 | |
| *** aojea has joined #openstack-kuryr | 10:21 | |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork master: Add docs for IPv6 address pool https://review.openstack.org/445405 | 10:21 |
| *** daidv has quit IRC | 10:21 | |
| openstackgerrit | Merged openstack/kuryr-libnetwork master: Split the test of requesting specific address https://review.openstack.org/441689 | 10:23 |
| *** aojea has quit IRC | 10:26 | |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork master: Add fullstack test for IPv6 subnet https://review.openstack.org/445376 | 10:27 |
| openstackgerrit | Merged openstack/kuryr-libnetwork master: Refactor the logic of updating existing port https://review.openstack.org/444601 | 10:30 |
| openstackgerrit | Merged openstack/kuryr-libnetwork master: Handle containers with both ipv4 & ipv6 https://review.openstack.org/439932 | 10:30 |
| apuimedo | vikasc: ping | 10:34 |
| openstackgerrit | Dongcan Ye proposed openstack/kuryr-libnetwork master: Make port bind on dual-stack port https://review.openstack.org/442525 | 10:38 |
| *** salv-orlando has joined #openstack-kuryr | 10:45 | |
| *** salv-orl_ has joined #openstack-kuryr | 10:46 | |
| *** salv-orlando has quit IRC | 10:49 | |
| *** salv-orl_ has quit IRC | 10:51 | |
| *** yedongcan has quit IRC | 10:58 | |
| *** yedongcan has joined #openstack-kuryr | 10:58 | |
| mchiappero | anyone familiar with GenericPodVIFDriver, is this driver really supposed to be generic? I guess it's more bridge oriented | 11:06 |
| mchiappero | as I'm considering the possibility to actually move all the really generic code in a base abstract class | 11:07 |
| mchiappero | a skeleton one | 11:07 |
| mchiappero | the Generic could then be renamed "bridge" or something like that | 11:08 |
| mchiappero | any thoughts? | 11:08 |
| vikasc | apuimedo, pong | 11:09 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork master: Updated from global requirements https://review.openstack.org/431966 | 11:20 |
| *** yedongcan has left #openstack-kuryr | 11:22 | |
| apuimedo | vikasc: any news on the ssl patch? | 11:23 |
| apuimedo | https://review.openstack.org/#/c/440979/ | 11:23 |
| vikasc | apuimedo, working on adding unit test to address ltomasbo comment | 11:24 |
| apuimedo | ok | 11:24 |
| ltomasbo | thanks vikasc | 11:24 |
| vikasc | yw ltomasbo :) | 11:25 |
| openstackgerrit | Merged openstack/kuryr-kubernetes master: Add ddt to test-requirements https://review.openstack.org/436509 | 12:03 |
| *** yuval has joined #openstack-kuryr | 12:04 | |
| *** yuval has quit IRC | 12:06 | |
| *** yuval has joined #openstack-kuryr | 12:07 | |
| *** saneax is now known as saneax-_-|AFK | 12:10 | |
| *** janonymous has quit IRC | 12:14 | |
| apuimedo | ltomasbo: ping | 12:16 |
| *** yuval has quit IRC | 12:31 | |
| *** yuval has joined #openstack-kuryr | 12:31 | |
| *** aojea has joined #openstack-kuryr | 12:45 | |
| *** garyloug has quit IRC | 12:46 | |
| openstackgerrit | vikas choudhary proposed openstack/kuryr-kubernetes master: Add support for HTTPS client https://review.openstack.org/440979 | 13:12 |
| vikasc | apuimedo, ltomasbo, irenab ivc_ ^ | 13:12 |
| *** aojea has quit IRC | 13:15 | |
| apuimedo | thanks vikasc. Will look into it in a little while | 13:18 |
| vikasc | apuimedo, thank you! | 13:20 |
| apuimedo | ltomasbo: regarding https://review.openstack.org/#/c/436877/7/kuryr_kubernetes/controller/drivers/generic_ports_pool.py . Please look at the paste http://paste.openstack.org/show/602683/ | 13:21 |
| apuimedo | I rewrote how you generate the bulk request | 13:21 |
| apuimedo | as you can see in the %timeit run, it's almost 20x faster | 13:22 |
| ltomasbo | apuimedo, pong | 13:22 |
| ltomasbo | just came back from lunch | 13:22 |
| apuimedo | :-) | 13:23 |
| ltomasbo | going to check what did you change! | 13:23 |
| ltomasbo | thanks! 20x is great! | 13:23 |
| apuimedo | ltomasbo: the crux of it, is that you were generating the same dictionary as many times as ports you wanted | 13:24 |
| *** garyloug has joined #openstack-kuryr | 13:24 | |
| apuimedo | when you could just have a single dict be included as N elements | 13:24 |
| apuimedo | other small improvements as well | 13:24 |
| ltomasbo | great thanks! | 13:25 |
| *** mattmceuen has joined #openstack-kuryr | 13:32 | |
| apuimedo | ltomasbo: also, I'd like to go over the patches with you | 13:43 |
| apuimedo | maybe we could do a bluejeans | 13:43 |
| apuimedo | ivc_: irenab and others, you'd be welcome to it if you want | 13:43 |
| vikasc | apuimedo, i am interested | 13:44 |
| apuimedo | good | 13:45 |
| ltomasbo | sure, I have a meeting from 4-5, other than that I'm available | 13:45 |
| vikasc | apuimedo, so does dict.append recreates a dict ? | 13:45 |
| apuimedo | vikasc: where? | 13:46 |
| vikasc | apuimedo, ltomasbo code was calling append() number of ports time. you said that ltomasbo is regenerating dict | 13:47 |
| vikasc | apuimedo, performance improvement that you did on paste output | 13:47 |
| apuimedo | vikasc: he was doing append on a dict | 13:48 |
| apuimedo | append on a list | 13:48 |
| apuimedo | I mean | 13:48 |
| apuimedo | that then he was adding to the dict | 13:48 |
| vikasc | apuimedo, yeah and u said "generating the same dictionary as many times as ports", so does that mean append() generates the same dict? | 13:49 |
| vikasc | apuimedo, ahh got it | 13:49 |
| vikasc | apuimedo, nevermind | 13:49 |
| vikasc | :) thank you! | 13:50 |
| apuimedo | ;-) | 13:51 |
| *** yedongcan has joined #openstack-kuryr | 13:51 | |
| apuimedo | ltomasbo: so, what do you think about the paste? | 13:52 |
| ltomasbo | sure, I will update the code accordingly! | 13:53 |
| ltomasbo | can't say 'no' to 20x speed up! | 13:53 |
| ltomasbo | :D | 13:53 |
| vikasc | :D | 13:53 |
| vikasc | but just by looking at both logics, its hard to infer that it would be 20x improvement. atleast to me | 13:54 |
| * vikasc noted down that list.append is too slow | 13:55 | |
| *** mattmceuen has quit IRC | 13:55 | |
| *** mattmceuen__ has joined #openstack-kuryr | 13:55 | |
| apuimedo | vikasc: there's the numbers at the bottom of the paste | 13:55 |
| vikasc | apuimedo, yeah i noticed those | 13:55 |
| vikasc | apuimedo, i am not denying :D | 13:56 |
| apuimedo | creating dicts is cheap, but it is not free | 13:56 |
| apuimedo | ltomasbo: so let's have the call in 20 minutes | 13:56 |
| apuimedo | I want to understand why we need to watch pod creation | 13:56 |
| apuimedo | why we need the config option instead of just having a noop pool driver | 13:57 |
| apuimedo | etc | 13:57 |
| dmellado | ltomasbo: 20x? | 13:57 |
| ltomasbo | apuimedo, sure! | 13:58 |
| apuimedo | dmellado: I'm disapointed though. I wanted https://i.ytimg.com/vi/LqSg9yVfzV0/maxresdefault.jpg | 14:00 |
| dmellado | apuimedo: LOL | 14:01 |
| dmellado | why don't you just have kuryr go into super-saiyan mode? xD | 14:01 |
| apuimedo | that should be a plugin | 14:04 |
| apuimedo | some kind of easter egg | 14:04 |
| *** IRCFrEAK has joined #openstack-kuryr | 14:13 | |
| *** IRCFrEAK has quit IRC | 14:14 | |
| *** saneax-_-|AFK is now known as saneax | 14:17 | |
| dmellado | apuimedo: note that I won't be speaking about further transformations | 14:18 |
| apuimedo | like? | 14:21 |
| apuimedo | Xd | 14:21 |
| dmellado | oh, those weird stuff that they added later | 14:22 |
| dmellado | like super-saiyan-god and so xD | 14:22 |
| apuimedo | you don't like the blue hair? | 14:22 |
| dmellado | that was just going overkill | 14:23 |
| dmellado | xD | 14:23 |
| vikasc | apuimedo, ltomasbo , bj meeting? | 14:24 |
| ltomasbo | yep | 14:24 |
| apuimedo | dmellado: and no eyebrows third level wasn't? | 14:25 |
| dmellado | apuimedo: heh, that kinda reminded me about whoopie goldberg | 14:25 |
| apuimedo | https://bluejeans.com/5508709975 | 14:25 |
| * apuimedo facepalms | 14:26 | |
| dmellado | apuimedo: is that for joining the whoopie-saiyan discussion? | 14:26 |
| *** hongbin has joined #openstack-kuryr | 14:26 | |
| apuimedo | dmellado: it's for going over luis' patches | 14:26 |
| dmellado | heh, just joking this time | 14:26 |
| dmellado | I'll join, but if you mention the saiyan topic I promise I'll go away | 14:27 |
| *** yedongcan has left #openstack-kuryr | 14:28 | |
| apuimedo | :-) | 15:08 |
| apuimedo | I managed not to mention it | 15:08 |
| apuimedo | alraddarla: mattmceuen__: I marked your contrail bp for pike-2 | 15:19 |
| apuimedo | next monday we can vote on the weekly IRC meeting for direction approval | 15:20 |
| apuimedo | so let's see if the spec already made it through reviews then | 15:20 |
| alraddarla | apuimedo, for a minute i thought you meant -2 ...not pike2 and I was so sad | 15:20 |
| *** danil has joined #openstack-kuryr | 15:28 | |
| apuimedo | xD | 15:33 |
| apuimedo | alraddarla: I'm really looking forward to it | 15:33 |
| dmellado | alraddarla: if apuimedo is being mean to you, just let us know xD | 15:35 |
| alraddarla | dmellado, thank you. I appreciate the back up | 15:36 |
| apuimedo | dmellado: you are risking your fuet supply | 15:36 |
| apuimedo | :B | 15:36 |
| dmellado | xD | 15:36 |
| dmellado | apuimedo: btw, you should give me your home address | 15:37 |
| dmellado | I'll send you some stuff next time I go to Galicia | 15:37 |
| apuimedo | I'm not sure I want the sort of stuff that arrives hidden in the Galicia ports | 15:37 |
| apuimedo | xD | 15:37 |
| apuimedo | I'm not from Ciudadanos | 15:37 |
| apuimedo | xD | 15:37 |
| dmellado | LOL | 15:38 |
| dmellado | apuimedo: Ciutadans, don't forget where does he come from | 15:39 |
| dmellado | xD | 15:39 |
| apuimedo | from La Caixa | 15:39 |
| apuimedo | xD | 15:39 |
| dmellado | apart from that xD | 15:39 |
| *** danil has quit IRC | 15:39 | |
| *** tonanhngo has joined #openstack-kuryr | 15:48 | |
| *** janki has quit IRC | 15:54 | |
| mchiappero | pod['status']['hostIP'] | 15:55 |
| mchiappero | do you know whether this info comes from the kubelet or the api-server? I would assume the latter, but it would help to get a confirmation | 15:56 |
| mchiappero | currently for nested deployments the 'link_iface' option is provided, but actually is not fully utilized, it could be avoided give the current detection logic based on the pod information | 15:57 |
| mchiappero | do you have any preference? | 15:57 |
| apuimedo | mchiappero: I've never checked. I'd hope the latter | 15:59 |
| apuimedo | Using the hostIP could be good | 16:00 |
| apuimedo | the link_iface was always a stopgap | 16:00 |
| mchiappero | but what if you run a single apiserver with many minions within VMs? | 16:00 |
| apuimedo | and an option to force the iface if multiple are there | 16:00 |
| apuimedo | mchiappero: not sure I get the point | 16:00 |
| mchiappero | if the hostIP comes from the host the apiserver runs on, you would get the host/BM IP address | 16:01 |
| mchiappero | say you have multiple minions running in different VMs | 16:01 |
| mchiappero | well, I'm not too sure it's a valid usecase | 16:01 |
| mchiappero | but if it is it would not work | 16:02 |
| mchiappero | I was considering leveraging the link_iface if present, leverage 'hostIP' otherwise | 16:02 |
| mchiappero | the important thing is to be able to retrieve the port ID of the master, once you have that I guess it can be cached | 16:03 |
| *** salv-orlando has joined #openstack-kuryr | 16:03 | |
| mchiappero | any feedback? | 16:03 |
| *** salv-orlando has quit IRC | 16:05 | |
| *** salv-orlando has joined #openstack-kuryr | 16:05 | |
| apuimedo | sorry, I got on something else | 16:06 |
| * apuimedo reading | 16:06 | |
| apuimedo | mchiappero: with "coming from the apiserver" I meant that the api server checks the API object Node of the node where the scheduler put the pod | 16:06 |
| apuimedo | so it would be the IP of the worker node VM | 16:07 |
| mchiappero | ok, perfect | 16:08 |
| mchiappero | I guess we still want to cache the name of the master interface too, instead of using netlink every time | 16:08 |
| apuimedo | mchiappero: yup | 16:09 |
| mchiappero | do we want to eliminate "link_iface" entirely or leave it optional? | 16:09 |
| apuimedo | that's a good point too | 16:09 |
| apuimedo | I'd prefer to have it optional | 16:09 |
| mchiappero | ok, cool | 16:10 |
| mchiappero | apuimedo: thanks for the feedback :) | 16:11 |
| *** salv-orlando has quit IRC | 16:13 | |
| apuimedo | you're most welcome :-) | 16:16 |
| apuimedo | irenab: vikasc: ivc_: ltomasbo: dmellado: https://blueprints.launchpad.net/kuryr-kubernetes/+spec/driver-handler-opt-loading | 16:25 |
| *** ltomasbo is now known as ltomasbo|away | 16:45 | |
| mchiappero | is there a way to get that IP without having to wait for a pod? | 16:51 |
| *** ltomasbo|away is now known as ltomasbo | 17:05 | |
| fkautz | mchiappero: you may be able to get it from the kuryr agent updating the pod's annotations which may land before the pod spins up | 17:11 |
| fkautz | not sure which component adds the pod ip, but the pod ip is definitely updated before the pod's state is set to running | 17:13 |
| apuimedo | mchiappero: you could do a get on Nodes | 17:17 |
| apuimedo | sorry, a watch on nodes | 17:17 |
| apuimedo | and keep it updated on memory | 17:18 |
| *** ltomasbo is now known as ltomasbo|away | 17:19 | |
| ivc_ | apuimedo about the opt loading bp, afair we discussed something like that a couple months ago, so i agree with the idea in general. though i'm not sure if i like it as a 'register_opts' method or maybe the cfg object can be declared as an attribute (similar to handler's OBJECT_KIND) | 17:19 |
| ivc_ | apuimedo btw, where/how do we discuss blueprints? is there some patchset in gerrit for it? | 17:21 |
| apuimedo | ivc_: you mean to have oslo opts as subgroup | 17:21 |
| apuimedo | ivc_: the idea was to use irc or ml | 17:21 |
| apuimedo | let me rephrase the subgroup thing | 17:21 |
| apuimedo | you mean to have an oslo opt subgroup in the handler/driver similar to the object_kind? | 17:22 |
| ivc_ | apuimedo yes. opts are declarative | 17:24 |
| apuimedo | I'm fine with that | 17:24 |
| apuimedo | ivc_: I never understood why blueprints do not have a comments section | 17:25 |
| apuimedo | :/ | 17:25 |
| ivc_ | apuimedo i thought the idea was to have an .rst doc/spec in tree that can be discussed under gerrit and once w+1 linked to the bp | 17:26 |
| apuimedo | for small things a single bp should be enough | 17:27 |
| ivc_ | apuimedo also regarding group name for a driver, we could probably use Driver's ALIAS + DriverImplClassName as the group name or smth like that | 17:32 |
| apuimedo | ivc_: can't we just let the driver/handler choose its own name? | 17:34 |
| apuimedo | ivc_: like, driver, handler is supposed to have a class object called | 17:35 |
| ivc_ | apuimedo sure, i mean i'd like the subgroup name to be something like '%s:%s' % (ALIAS, something) | 17:35 |
| apuimedo | OPT | 17:35 |
| apuimedo | OPTGROUP | 17:35 |
| apuimedo | and it is an oslo OptGroup | 17:35 |
| apuimedo | which already includes the name | 17:35 |
| apuimedo | ivc_: well, that would help avoid collisions | 17:36 |
| ivc_ | apuimedo more like VARIANT for its more generic and could be used not only for OPT, but something else too | 17:36 |
| apuimedo | like? | 17:36 |
| ivc_ | apuimedo dunno. just intuition. | 17:37 |
| ivc_ | apuimedo we could as well rename ALIAS to TYPE and have NAME for OPTGROUP/VARIANT | 17:37 |
| apuimedo | I'm a bit confused | 17:38 |
| apuimedo | let me make a pastebin | 17:38 |
| apuimedo | ivc_: http://paste.openstack.org/show/602731/ | 17:45 |
| apuimedo | and then, the loading code in kuryr-controller would make the option group name from the class name | 17:46 |
| apuimedo | it would be handler_classname | 17:46 |
| apuimedo | or driver_classname | 17:46 |
| apuimedo | in this case, the latter | 17:46 |
| ivc_ | apuimedo thats what i dont quite like. class names are not that pretty from user's perspective | 17:47 |
| ivc_ | apuimedo for the pastebin you linked i'd like the optgroup to be 'pod_vif:nested_vlan' | 17:48 |
| ivc_ | apuimedo where pod_vif comes from PodVifDriver.ALIAS and nested_vlan is NestedVlanPodVIFDriver.NAME | 17:49 |
| ivc_ | apuimedo s/NAME/VARIANT/ or something. i cant figure a proper name for it now | 17:50 |
| apuimedo | ivc_: I actually had played with the idea to convert the class name | 17:50 |
| apuimedo | like this | 17:50 |
| apuimedo | NestedVlanPodVIFDriver -> nested_vlan_pod_vif_driver | 17:50 |
| apuimedo | for the opt group | 17:50 |
| ivc_ | apuimedo i thought you hate implicit things :) | 17:50 |
| apuimedo | I do | 17:50 |
| apuimedo | but I also like avoiding naming conflicts | 17:51 |
| apuimedo | ivc_: why not NAME? | 17:52 |
| apuimedo | (I didn't get it above) | 17:52 |
| ivc_ | apuimedo mostly because i expect NAME to be unique, but in that case it should be unique within the Driver.ALIAS | 17:53 |
| ivc_ | apuimedo but if we rename Driver.ALIAS to Driver.TYPE, it makes sense as NAME (so the opt group is 'TYPE:NAME') | 17:54 |
| ivc_ | apuimedo also as part of renaming ALIAS to TYPE we'd get rid of package.module...ALIAS notation for driver/handler's names in setup.cfg | 17:56 |
| ivc_ | apuimedo since TYPE:NAME would have to be unique for kuryr.conf | 17:57 |
| ivc_ | apuimedo also imho Driver.TYPE + DriverImpl.NAME would be quite easy to explain in docs and for other devs to understand | 17:59 |
| *** aojea has joined #openstack-kuryr | 18:02 | |
| *** pcaruana has quit IRC | 18:03 | |
| *** garyloug has quit IRC | 18:07 | |
| mchiappero | I think we should start merging with or moving some code to kuryr.lib | 18:08 |
| ivc_ | mchiappero too early. APIs are not stable yet | 18:20 |
| apuimedo | agreed | 18:22 |
| ivc_ | apuimedo mchiappero i'm also considering making Handlers and Drivers share a common base. a) to allow loading Handlers similar to Drivers and b) to allow lifecycle management for Drivers similar to Retry/Log from Handlers (e.g. the LBaaS driver has own 'retry' now) | 18:23 |
| ivc_ | apuimedo so probably both Handlers and Drivers will become Actors :) | 18:24 |
| apuimedo | that's radical | 18:24 |
| apuimedo | xD | 18:24 |
| ivc_ | cmon we discussed it many times already :) | 18:25 |
| apuimedo | I know | 18:25 |
| apuimedo | but not for Pike | 18:25 |
| apuimedo | that's the radical part | 18:25 |
| ivc_ | not if we can pull it off by the end of p-2 milestone | 18:26 |
| apuimedo | that's ambitious | 18:27 |
| apuimedo | ivc_: http://paste.openstack.org/show/602736/ | 18:30 |
| ivc_ | apuimedo aye. and [pod_vif/nested_vlan] or [pod_vif:nested_vlan] in kuryr.conf, right? | 18:32 |
| apuimedo | and then, in base.py http://paste.openstack.org/show/602737/ | 18:32 |
| apuimedo | the rename from ALIAS to TYPE | 18:32 |
| apuimedo | ivc_: either '/' or '|' | 18:32 |
| apuimedo | I have to check what oslo ini reader tolerates | 18:33 |
| ivc_ | apuimedo and also get rid of '.controller.drivers' in setup.cfg at some point later | 18:34 |
| apuimedo | that's for a separate blueprint though, isn't it? | 18:37 |
| ivc_ | apuimedo yup | 18:38 |
| ivc_ | apuimedo btw, we still have that 'Please fill here a long description which must be at least 3 lines' in our README.rst :) | 18:55 |
| alraddarla | I noticed that when I was digging around....should have mentioned it. | 18:58 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes master: Add project description to README.rst https://review.openstack.org/445621 | 19:01 |
| ivc_ | apuimedo ^^ | 19:02 |
| ivc_ | apuimedo btw, maybe we should also reference http://superuser.openstack.org/articles/networking-kubernetes-kuryr/ in our README.rst? | 19:07 |
| apuimedo | lol | 19:07 |
| apuimedo | I didn't notice it | 19:07 |
| apuimedo | ivc_: feel free to mention it | 19:07 |
| apuimedo | ivc_: what's that character that you used instead of an apostrophe? | 19:08 |
| apuimedo | in "it's" | 19:09 |
| ivc_ | apuimedo i've copied it from our article :) | 19:09 |
| ivc_ | you know i'm lazy | 19:10 |
| apuimedo | xD | 19:11 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes master: Add project description to README.rst https://review.openstack.org/445621 | 19:15 |
| ivc_ | apuimedo updated ^^ | 19:21 |
| apuimedo | fkautz: I reviewed your patch | 19:27 |
| *** mattmceuen__ has quit IRC | 19:40 | |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes master: K8s Services support: LBaaSv2Driver https://review.openstack.org/433359 | 20:05 |
| apuimedo | :O | 20:10 |
| apuimedo | I'm excited to look at it! | 20:10 |
| apuimedo | ivc_: +2 :-) | 20:11 |
| ivc_ | apuimedo :) | 20:12 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes master: [WIP] K8s Services support: LoadBalancerHandler https://review.openstack.org/376045 | 20:19 |
| portdirect | apuimedo (or anyone!): you have a moment to talk about tests? | 20:19 |
| ivc_ | portdirect sure | 20:20 |
| portdirect | I'd like to get the ball rolling on creating a similar setup to my old harbor project (been talking about it long enough...) | 20:20 |
| portdirect | what testing is being run in the gate atm? looks just like a smoke test from the logs? | 20:20 |
| portdirect | (this in in the context of kuryr-kubernetes) | 20:22 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes master: [WIP] K8s Services support: LoadBalancerHandler https://review.openstack.org/376045 | 20:22 |
| ivc_ | portdirect are you referring to the python27/python35 tests or dsvm ones? the former only do unit testing and afaik the latter only test the success of ./stack.sh. | 20:28 |
| ivc_ | portdirect fullstack&functional tests are yet to be implemented | 20:29 |
| portdirect | ivc_: cool - thanks for the info, that was what it looked like from 1st parse of the logs | 20:30 |
| portdirect | ivc_: I'm currently knocking together a PoC of running kuryr-kubernetes in openstack-helm | 20:30 |
| portdirect | ivc_: and we should be moving to openstack governance in a couple of weeks | 20:31 |
| portdirect | ivc_: would be great if we could colab on providing functional/fullstack testing? | 20:31 |
| ivc_ | portdirect thats interesting | 20:32 |
| ivc_ | apuimedo ^^ | 20:32 |
| *** alraddarla_ has joined #openstack-kuryr | 20:37 | |
| *** alraddarla_ has quit IRC | 20:37 | |
| *** tonanhngo has quit IRC | 21:04 | |
| fkautz | apuimedo: thanks, i'll update soon | 21:39 |
| *** mchiappero is now known as mchiappe|vacatio | 22:04 | |
| *** mchiappe|vacatio is now known as mchiappe|timeoff | 22:05 | |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-libnetwork master: Add fullstack test for containers with dual-net https://review.openstack.org/444661 | 22:18 |
| *** aojea has quit IRC | 22:41 | |
| *** aojea has joined #openstack-kuryr | 22:42 | |
| *** aojea has quit IRC | 22:46 | |
| *** saneax is now known as saneax-_-|AFK | 23:05 | |
| *** tonanhngo has joined #openstack-kuryr | 23:35 | |
| *** pmannidi has joined #openstack-kuryr | 23:41 | |
| openstackgerrit | Hongbin Lu proposed openstack/kuryr-libnetwork master: [WIP] Reset neutron existing port on release_address https://review.openstack.org/445683 | 23:44 |
| *** hongbin has quit IRC | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
