openstackgerrit | Merged openstack-infra/shade: Remove pointless and fragile unittest https://review.openstack.org/397345 | 00:54 |
---|---|---|
openstackgerrit | Merged openstack-infra/shade: Be specific about protected being bool https://review.openstack.org/397323 | 01:31 |
*** abregman has joined #openstack-shade | 06:15 | |
*** abregman has quit IRC | 08:17 | |
*** abregman has joined #openstack-shade | 08:53 | |
*** abregman has quit IRC | 09:04 | |
*** abregman has joined #openstack-shade | 09:12 | |
*** openstackgerrit has quit IRC | 09:18 | |
*** openstackgerrit has joined #openstack-shade | 09:18 | |
openstackgerrit | Arie Bregman proposed openstack-infra/shade: Add compute usage support https://review.openstack.org/393873 | 12:50 |
*** abregman is now known as abregman|mtg | 13:06 | |
*** abregman|mtg is now known as abregman | 13:23 | |
*** yolanda has quit IRC | 13:39 | |
openstackgerrit | Monty Taylor proposed openstack/oaktree: Expand on the nodepool API todo in the docs https://review.openstack.org/397747 | 13:44 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Expose visibility on images https://review.openstack.org/397750 | 13:47 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Collapse list/search for networks, users and keypairs https://review.openstack.org/397208 | 13:48 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Normalize volumes https://review.openstack.org/396953 | 13:48 |
*** yolanda has joined #openstack-shade | 13:55 | |
*** yolanda has quit IRC | 14:07 | |
*** yolanda has joined #openstack-shade | 14:23 | |
openstackgerrit | Merged openstack/oaktree: Expand on the nodepool API todo in the docs https://review.openstack.org/397747 | 14:48 |
*** gouthamr has joined #openstack-shade | 14:59 | |
*** sigmavirus has joined #openstack-shade | 15:01 | |
*** stevemar has joined #openstack-shade | 15:02 | |
stevemar | mordred: i'd like to help with oaktree, but realistically only after atlanta | 15:02 |
mordred | stevemar: yay! happy to have you any time you've got time | 15:02 |
stevemar | mordred: but mark me as interested -- "You Had My Curiosity But Now You Have My Attention" | 15:03 |
*** edtubill has joined #openstack-shade | 15:03 | |
mordred | \o/ | 15:03 |
*** cdent has joined #openstack-shade | 15:04 | |
*** sc68cal has joined #openstack-shade | 15:07 | |
*** pabelanger has joined #openstack-shade | 15:12 | |
*** abregman is now known as abregman|brb | 15:13 | |
*** gouthamr has quit IRC | 15:16 | |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Create and use a Adapter wrapper for REST in TaskManager https://review.openstack.org/384206 | 15:29 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Have OpenStackHTTPError inherit from HTTPError https://review.openstack.org/390136 | 15:29 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Use REST for listing images https://review.openstack.org/384208 | 15:29 |
Shrews | i wish i had time for oaktree | 15:33 |
cdent | ditto | 15:33 |
Shrews | ugh, atlanta. when is that? feb? | 15:38 |
mordred | yah - I think so | 15:38 |
pabelanger | I joined to learn more about oaktree :) | 15:42 |
*** openstackgerrit has quit IRC | 15:48 | |
*** openstackgerrit has joined #openstack-shade | 15:48 | |
*** gouthamr has joined #openstack-shade | 15:51 | |
*** jroll has joined #openstack-shade | 15:52 | |
*** dtroyer has joined #openstack-shade | 15:53 | |
* mordred waves to all the nice people | 15:57 | |
Shrews | wait, we have nice people here? | 15:58 |
*** gouthamr has quit IRC | 16:00 | |
mordred | Shrews: ssssh | 16:01 |
* dtroyer looks around to see where they are hiding | 16:02 | |
*** gouthamr has joined #openstack-shade | 16:07 | |
*** gouthamr_ has joined #openstack-shade | 16:10 | |
*** gouthamr_ has quit IRC | 16:11 | |
*** gouthamr has quit IRC | 16:11 | |
*** gouthamr_ has joined #openstack-shade | 16:12 | |
openstackgerrit | Sean M. Collins proposed openstack/oaktreemodel: [WIP] Add Neutron network resources https://review.openstack.org/397806 | 16:12 |
cdent | so, if oaktree is all gRPC, that means that random curl user is unable to walk up and just use it, right? That's kind of a shame, but understandable. | 16:15 |
mordred | cdent: yes - that is the tradeoff for sure | 16:16 |
* cdent weeps for the web that was | 16:17 | |
openstackgerrit | Sean M. Collins proposed openstack/oaktreemodel: [WIP] Add Neutron network resources https://review.openstack.org/397806 | 16:17 |
* cdent pours one out for his dead urls | 16:17 | |
cdent | etc | 16:17 |
mordred | cdent: well, I mean, I say that - it's protobuf over http/2 - so in _theory_ one could still use curl | 16:17 |
mordred | it would just be way more awkward | 16:17 |
cdent | indeed | 16:17 |
cdent | at this stage it seems a fair price to pay | 16:18 |
cdent | but it still makes me all maudlin | 16:18 |
mordred | yah - and I think if we wanted ot accept json payloads instead of protobuf (or return them if requested) it wouldn't be the world's hardest thing to add later | 16:18 |
mordred | cdent: in the go gRPC server impl there is a REST/gRPC gateway - we can't use it directly, but we could implement a similar construct | 16:19 |
mordred | (I tried to come up with ways to use the go gateway with the python grpc service, but reality folded in on itself and I got lost in the nether reaches of time and space) | 16:19 |
cdent | anybody mind if I add purplerbot in here? https://anticdent.org/purple-irc-bot.html | 16:20 |
cdent | (mostly for the transclusion, not so much the loggin) | 16:20 |
mordred | cdent: works for me! | 16:21 |
cdent | rad | 16:21 |
*** purplerbot has joined #openstack-shade | 16:22 | |
cdent | boo | 16:22 |
cdent | [t 1LuT] | 16:22 |
purplerbot | <cdent> boo [2016-11-15 16:22:28] [n 1LuT] | 16:22 |
cdent | \o/ | 16:22 |
* mordred wave to purplerbot | 16:23 | |
mordred | sc68cal: added some comments inline on your WIP patch there | 16:27 |
*** abregman|brb is now known as abregman | 16:27 | |
sc68cal | mordred: thanks. | 16:35 |
sc68cal | mordred: Do we want to paper over some of the oddities of the networking API? For example the whole subnet object, with the ipv6 attributes | 16:36 |
sc68cal | do we want to break subnet into two subobjects, a v4 one and a v6 one? | 16:36 |
sc68cal | just so we don't have this subnet object with tons of mutually exclusive fields | 16:36 |
mordred | sc68cal: yah. that's a great question and where my brain had been leaning - the network/subnet split is TERRIBLE from an end user perspective | 16:37 |
mordred | like, you hand a network to create_server, but what you're really doing is attaching a vm to a subnet | 16:37 |
mordred | and the qualities of what you can do with the vm networking-wise are all tied to the subnet - even though it's not the unit that's reported | 16:38 |
sc68cal | s/subnet/subnets/g | 16:38 |
mordred | ++ | 16:38 |
mordred | sc68cal: so - in short - yeah- same as we've invented our own api in shade for creating servers that totally hides much of the complexity in ips behind the scenes, doing that for networks seems like a good idea | 16:39 |
mordred | sc68cal: I'll warn you - you may want to gouge your eyes out by the time you're done though | 16:39 |
*** abregman has quit IRC | 16:40 | |
sc68cal | mordred: I'll start wearing a welding mask | 16:40 |
Shrews | that's not sufficient enough | 16:41 |
*** calebb has quit IRC | 17:27 | |
thingee | mordred finally get around to https://review.openstack.org/#/c/384206/12 today ... catching up on the task manager code atm | 17:39 |
mordred | thingee: it's some fun code! | 17:39 |
*** calebb has joined #openstack-shade | 17:46 | |
* Shrews loves seeing more people in the shade codez | 17:50 | |
*** edtubill has quit IRC | 17:59 | |
*** edtubill has joined #openstack-shade | 19:16 | |
*** cdent has quit IRC | 19:30 | |
thingee | mordred this seems to break the deprecated dreamhost profile https://review.openstack.org/#/c/384208/12 http://paste.openstack.org/show/589322/ | 19:41 |
thingee | dreamcompute is working though | 19:41 |
mordred | thingee: oh! thanks! that's an excellent test case | 19:42 |
mordred | thingee: there are evil evil things that happen in glanceclient to deal with urls with and without versions prepended | 19:42 |
mordred | clearly we need to do more of them :) | 19:42 |
thingee | other clouds seem to be fine from my testing. I guess dreamhost's old cluster was already including a version? | 19:43 |
thingee | in the endpoint | 19:43 |
thingee | I can't even see the endpoints. doing a cloud_config.get_endpoint('dreamhost') just gives None. I get None for all other clouds too. | 19:44 |
thingee | errr | 19:44 |
thingee | cloud_config.get_endpoint('image') ... where cloud_config was initialized with dreamhost profile. | 19:45 |
thingee | mordred other than that though, everything else looks good to me | 19:46 |
mordred | thingee: hrm. poo. my old dreamhost creds don't seem to work anymore. I should go check my accout | 19:50 |
mordred | thingee: try doing "cloud=shade.openstack_cloud(cloud='dreamhost') ; print cloud.service_catalog" and it should print you the whole catalog for dreamhost (assuming that's the right cloud name) | 19:50 |
thingee | mordred boom http://paste.openstack.org/show/589326/ | 19:54 |
*** cdent has joined #openstack-shade | 19:55 | |
*** edtubill has left #openstack-shade | 19:56 | |
mordred | thingee: cool. thanks! so - my comment in https://review.openstack.org/#/c/384208/12/shade/openstackcloud.py about figuring out discovery later ... it seems later may be now :) | 19:57 |
thingee | mordred I'd be fine with a follow-up. Just as long as we have this fixed before a release. | 19:58 |
* thingee is not aware of who is actively using non-released builds though | 19:59 | |
thingee | jlk is it possible for us to get an openstack label in the github repo ansible/ansible-modules-core. Right now I've been filtering issues/pr with the cloud label and just vague 'openstack' search | 20:04 |
jlk | That could work. There is also tooling that sivel wrote to be made aware of things matching code paths. | 20:06 |
jlk | one sec. | 20:06 |
jlk | https://ansible.sivel.net/pr/byfile.html you can find open things there | 20:07 |
jlk | by file | 20:07 |
jlk | https://ansible.sivel.net has a link where you can sign up for notifications, by path I believe | 20:08 |
jlk | I think the bot is what applies labels, https://github.com/ansible/ansibullbot A PR to that might work. | 20:10 |
thingee | thanks | 20:10 |
jlk | perhaps https://github.com/ansible/ansibullbot/blob/master/lib/triagers/ansible_ansible_pulltriager.py#L50 | 20:11 |
jlk | thingee: confirmed, a PR to the bot is the best way to go about that. | 20:13 |
*** gouthamr_ has quit IRC | 20:14 | |
mordred | thingee: we can also add you in there to the list of people who should get pinged about openstack PRs | 20:26 |
mordred | thingee: https://github.com/ansible/ansibullbot/blob/master/MAINTAINERS-EXTRAS.txt#L77 and https://github.com/ansible/ansibullbot/blob/master/MAINTAINERS-CORE.txt#L53 | 20:27 |
thingee | Shrews fyi https://github.com/ansible/ansible-modules-core/pull/5601#issuecomment-260746794 | 20:48 |
mordred | Shrews, jlk, thingee: I wonder if we should try to get a GH PR bot here that announces openstack module PRs | 20:49 |
*** cdent has quit IRC | 21:06 | |
mordred | thingee: did dreamcompute fall off the internet? | 21:09 |
mordred | thingee: nevermind - I seem to be having internet issues locally | 21:12 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Use REST for listing images https://review.openstack.org/384208 | 21:30 |
mordred | thingee: warning TERRIBLE HACK but may still be the best bet (/me cries) ^^ | 21:31 |
jlk | There is a github to IRC integration available on github itself, I'm not sure if you can target it to specific paths though. | 21:56 |
*** gouthamr_ has joined #openstack-shade | 22:19 | |
*** gouthamr_ has quit IRC | 22:26 | |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Use REST for listing images https://review.openstack.org/384208 | 22:26 |
mordred | jlk: looks like not :( | 23:02 |
mordred | thingee: ^^ that version of the patch MAY actually do version discovery right and also fallback properly for glances configured with the version url | 23:02 |
mordred | thingee: the version discovery is not tested in that patch ... I have another patch I need to ressurect to submit to do that | 23:18 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!