| *** jamielennox is now known as jamielennox|away | 01:01 | |
| *** jamielennox|away is now known as jamielennox | 01:19 | |
| *** gouthamr has quit IRC | 02:16 | |
| mordred | jamielennox: we do something special with the interface value if someone is requesting a client for the 'identity' service | 02:58 |
|---|---|---|
| mordred | jamielennox: so instead of 'public' or 'admin' - we pass in the singleton you have defined in ksa ... plugin.AUTH_INTERFACE | 02:59 |
| mordred | jamielennox: I remember putting that in for ksc a while back - but when I was looking at the ksc code the other day it looked much cleaner than I remmber, so I think you might have fixed whatever it was that made us need to do a workaround | 03:00 |
| jamielennox | mordred: yea - i'm surprised you need to do that | 03:02 |
| jamielennox | mordred: AUTH_INTERFACE is the value passed to the plugin as 'auth_url' - so bypasses the catalog completely | 03:02 |
| jamielennox | it's basically something that the plugins themselves need access to so they can do things like scoping | 03:03 |
| jamielennox | and operations on unscoped token | 03:03 |
| jamielennox | s | 03:03 |
| jamielennox | there's no reason that keystone shouldn't lookup the catalog there just like everything else | 03:03 |
| jamielennox | Shrews, mordred: for the os_network and related ansible modules you do: https://github.com/ansible/ansible-modules-core/blob/devel/cloud/openstack/os_router.py#L326-L335 | 03:53 |
| jamielennox | so if no project is provided then don't scope the neutron request | 03:53 |
| jamielennox | i feel like the way more logical behaviour their is that if no project is provided filter based on the current auth project | 03:54 |
| jamielennox | ksa can obviously give you this and i'm assuming you could fetch it from os-c-c, but is it possible to even suggest making this change? | 03:54 |
| jamielennox | ansible still marks the module in preview so it can change, but would you be alright with such a break? | 03:55 |
| *** jamielennox is now known as jamielennox|away | 04:14 | |
| *** jamielennox|away is now known as jamielennox | 04:22 | |
| *** yfried has joined #openstack-shade | 04:40 | |
| *** yfried has quit IRC | 04:57 | |
| *** yfried has joined #openstack-shade | 06:05 | |
| *** rcarrillocruz has joined #openstack-shade | 07:45 | |
| *** ioggstream has joined #openstack-shade | 07:58 | |
| *** openstackgerrit has quit IRC | 08:33 | |
| *** rcarrillocruz has quit IRC | 09:57 | |
| *** ioggstream has quit IRC | 11:44 | |
| *** ioggstream has joined #openstack-shade | 12:08 | |
| *** gouthamr has joined #openstack-shade | 12:21 | |
| *** ioggstream has quit IRC | 12:31 | |
| Shrews | jamielennox: been a long time since i did anything with that module, but isn't that how it already works if no project is specified? maybe i need a more concrete example so see what you are seeing. | 12:33 |
| Shrews | but i think changing behavior now would make folks upset, regardless of the status of the modules | 12:34 |
| *** ioggstream has joined #openstack-shade | 13:16 | |
| *** gouthamr_ has joined #openstack-shade | 13:36 | |
| *** gouthamr has quit IRC | 13:36 | |
| mordred | jamielennox: I dislike pretty much all combinations related to how cross-project things work in neutron | 13:54 |
| *** openstackgerrit has joined #openstack-shade | 14:17 | |
| openstackgerrit | Monty Taylor proposed openstack/os-client-config master: Stop special-casing idenity catalog lookups https://review.openstack.org/450259 | 14:17 |
| mordred | jamielennox: ^^ let's see if that works | 14:17 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Add normalization for heat stacks https://review.openstack.org/450260 | 14:18 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Replace heatclient testing with requests_mock https://review.openstack.org/450261 | 14:18 |
| mordred | Shrews: ^^ woot! heatclient testing mocks replaced (I had to learn a few things that I didn't really want to know) | 14:22 |
| Shrews | sweet | 14:29 |
| pabelanger | mordred: is the dependency removal of ironic difficult? | 14:31 |
| mordred | pabelanger: it shouldn't be terribly difficult, no- ironicclient already requires users todo jsonpatch things, so that part shouldn't change | 14:39 |
| mordred | pabelanger: unfortunately (or fortunately, depending on your POV) ironic functions have a bunch of tests - so the test conversion patch will not be quick :) | 14:40 |
| mordred | pabelanger: if you have the urge to work on that one, I'd be happy to give you pointers :) | 15:19 |
| pabelanger | mordred: Ya, I might take a go at it. I'll see if I can get some free time this week | 15:20 |
| mordred | pabelanger: sweet! I believe once ironicclient is gone we'll drop a _bunch_ of depends, because ironicclient still depends on python-openstackclient so I think it's the source of most of shade's transitive depends | 15:24 |
| pabelanger | mordred: Ya, that's my motivation :) RPM still pulling it in | 15:25 |
| notmorgan | ironicclient depends on OSC?! | 15:28 |
| notmorgan | what kind of crazy is that? | 15:28 |
| dtroyer | That is what happened before osc-lib existed when they wanted to use some stuffs, one reason osc-lib now exists… | 15:30 |
| *** yfried has quit IRC | 15:36 | |
| notmorgan | dtroyer: ah. but wow. seems very circular even knowing that. | 15:39 |
| dtroyer | Yeah, it was never a good idea, and I think there may have only ever been one other client lib to try that, so it fortunately didn't cat ch on :) | 15:45 |
| mordred | notmorgan: GAH | 15:52 |
| mordred | notmorgan: testtools.matchers._impl.MismatchError: !=: | 15:52 |
| mordred | reference = ('GET', | 15:52 |
| mordred | 'https://orchestration.example.com/v1/1c36b64c840a42cd9e9b931a369337f0/stacks/name-3/events?limit=1&sort_dir=desc') | 15:52 |
| mordred | actual = ('GET', | 15:52 |
| mordred | 'https://orchestration.example.com/v1/1c36b64c840a42cd9e9b931a369337f0/stacks/name-3/events?sort_dir=desc&limit=1') | 15:52 |
| notmorgan | mordred: yep | 15:53 |
| notmorgan | I am going to fix that by exploding the qs | 15:53 |
| notmorgan | and parsing it like complete_qs does in requests mock | 15:53 |
| mordred | notmorgan: that sounds lke a completely sane thing to do | 15:53 |
| notmorgan | I hit that exact issue with the idea tity role tests | 15:55 |
| notmorgan | it's not fun. | 15:55 |
| mordred | notmorgan: I'm going to make a bad hack around it in this heatclient patch - but am putting in a TODO to fix the bad hack once your qs thing is available | 15:58 |
| notmorgan | truncate the qs from your match as a bad hack | 15:58 |
| notmorgan | in assert_calls | 15:58 |
| notmorgan | easiest | 15:58 |
| mordred | well - I actually have to match on different qs's - the poll_for_events code depends heavily on qs differences | 16:02 |
| mordred | notmorgan: but - passing an OrderedDict in to get(params= is working :) | 16:03 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Remove python-heatclient and replace with REST https://review.openstack.org/450317 | 16:11 |
| mordred | woot! | 16:11 |
| mordred | another one bits the dust | 16:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Add list_availability_zone_names method https://review.openstack.org/450332 | 16:39 |
| *** ioggstream has quit IRC | 17:10 | |
| *** gouthamr_ is now known as gouthamr | 17:32 | |
| *** yfried has joined #openstack-shade | 17:43 | |
| *** ioggstream has joined #openstack-shade | 18:19 | |
| openstackgerrit | Merged openstack-infra/shade master: Add list_availability_zone_names method https://review.openstack.org/450332 | 18:20 |
| *** ioggstream has quit IRC | 18:51 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Remove python-heatclient and replace with REST https://review.openstack.org/450317 | 18:56 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Replace heatclient testing with requests_mock https://review.openstack.org/450261 | 18:56 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Add normalization for heat stacks https://review.openstack.org/450260 | 18:56 |
| *** samueldmq has quit IRC | 19:23 | |
| *** samueldmq has joined #openstack-shade | 19:24 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Remove python-heatclient and replace with REST https://review.openstack.org/450317 | 19:29 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Replace heatclient testing with requests_mock https://review.openstack.org/450261 | 19:29 |
| *** ioggstream has joined #openstack-shade | 19:29 | |
| *** yfried has quit IRC | 19:49 | |
| *** yfried has joined #openstack-shade | 20:06 | |
| *** jamielennox is now known as jamielennox|away | 20:13 | |
| *** jamielennox|away is now known as jamielennox | 20:24 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Remove python-heatclient and replace with REST https://review.openstack.org/450317 | 20:28 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade master: Replace heatclient testing with requests_mock https://review.openstack.org/450261 | 20:28 |
| *** gouthamr has quit IRC | 20:29 | |
| openstackgerrit | Merged openstack-infra/shade master: Add normalization for heat stacks https://review.openstack.org/450260 | 20:34 |
| *** yfried has quit IRC | 20:37 | |
| *** Matias has joined #openstack-shade | 21:00 | |
| *** gouthamr has joined #openstack-shade | 21:01 | |
| openstackgerrit | Merged openstack-infra/shade master: Replace heatclient testing with requests_mock https://review.openstack.org/450261 | 21:52 |
| Shrews | yeah, take THAT, heatclient | 21:54 |
| Shrews | mordred: i can't WAIT for you to take on neutronclient | 21:55 |
| mordred | Shrews: soon ... | 21:57 |
| mordred | Shrews: actually, I think neutronclient will be the easiest | 21:57 |
| mordred | because it doesn't really do anything | 21:57 |
| mordred | the interface in shade is a bunch of dicts | 21:57 |
| Shrews | mordred: that's what it wants you to believe.... | 21:57 |
| mordred | :) | 21:57 |
| mordred | Shrews: heatclient was much harder because they put a giant pile of logic into the client lib | 21:57 |
| *** ioggstream has quit IRC | 22:06 | |
| openstackgerrit | Merged openstack-infra/shade master: Remove python-heatclient and replace with REST https://review.openstack.org/450317 | 22:12 |
| mordred | BOOM | 22:23 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!