| *** chlong has quit IRC | 00:01 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 00:01 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 00:01 | |
| *** chlong has joined #openstack-sdks | 00:02 | |
| *** chlong has quit IRC | 00:06 | |
| *** chlong has joined #openstack-sdks | 00:29 | |
| *** pm90__ has quit IRC | 00:55 | |
| *** etoews has quit IRC | 01:33 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:34 | |
| *** terrylhowe has quit IRC | 01:35 | |
| *** britthou_ has joined #openstack-sdks | 01:38 | |
| *** britthouser has quit IRC | 01:42 | |
| *** chlong has quit IRC | 02:11 | |
| *** stevemar has quit IRC | 02:26 | |
| *** etoews has joined #openstack-sdks | 03:07 | |
| *** etoews has quit IRC | 03:12 | |
| *** briancurtin has joined #openstack-sdks | 03:46 | |
| *** chlong has joined #openstack-sdks | 03:57 | |
| *** chlong has quit IRC | 04:02 | |
| openstackgerrit | Dean Troyer proposed openstack/python-openstackclient: Add version url config hack https://review.openstack.org/145681 | 04:05 |
|---|---|---|
| dtroyer | jamielennox: ^^^^ we talked about that concept quite a while ago and I couldn't find it in the current ksc. I don't expect to merge that review, but it demonstrates what I need to do to work with the broken default Keystone configs I have the joy of dealing with. | 04:08 |
| dtroyer | thoughts on the actual right implementation? should it just go directly into BaseGenericPlugin? | 04:08 |
| jamielennox | dtroyer: so i have part of that in ksc | 04:08 |
| jamielennox | where are you trying to do it from though? | 04:09 |
| jamielennox | is this the initial OS_AUTH_URL url? | 04:09 |
| dtroyer | no. the URL returned in the default Keystone version response is http://localhost:xxxx/v2.0 and if that is not set then version discovery falls over | 04:09 |
| jamielennox | oh | 04:10 |
| jamielennox | sorry was thinking something else | 04:10 |
| jamielennox | so i know i fixed that server side | 04:10 |
| dtroyer | right, different from the v2.0 hack that I see you do have (nice solution btw) | 04:10 |
| jamielennox | at least for keystone | 04:10 |
| jamielennox | heh, not sure if i'm proud of it or not but it's working | 04:11 |
| dtroyer | I couldn't find a better way to do this in OSC because of the timing of the delayed auth, I didn't see any place to hook it | 04:12 |
| jamielennox | so that's just going to fix keystone | 04:12 |
| dtroyer | right, but I don't think there is a single solution to do both auth and other services too | 04:12 |
| dtroyer | same process, different locations | 04:13 |
| jamielennox | well i guess the place to hook the other services is when discovery happens, that if you return a localhost then use the url that you sent discovery to instead | 04:13 |
| jamielennox | i don't know how you'd hook that one either | 04:14 |
| dtroyer | anyway, no hurry, I've sat on this for a couple of months already… just wanted some ideas of where it should really go | 04:15 |
| jamielennox | dtroyer: are you seeing it much in reality? | 04:16 |
| dtroyer | not on real production clouds, yet, but on a few POC and internal things | 04:17 |
| jamielennox | i could handle it as part of all identity plugins - but i'm not really sure we should | 04:17 |
| *** chlong has joined #openstack-sdks | 04:18 | |
| dtroyer | right, which is why I looked for an OSC solution first | 04:18 |
| jamielennox | i'd be inclined to override the get_endpoint function rather than the builder | 04:19 |
| jamielennox | oh - but that will only help with once you have a client - it won't help with creating a new plugin | 04:19 |
| dtroyer | right, the auth path is separate | 04:19 |
| dtroyer | and frankly, that's the only one I've had a problem with so far, but I don't think anything else is doing discovery from the URL retrieved from the service catalog (yet) | 04:20 |
| jamielennox | keystone will | 04:20 |
| jamielennox | there is a patch up for cinder that makes it do ti | 04:21 |
| jamielennox | they're the only two i'm aware of | 04:21 |
| dtroyer | I didn't know about cinder, nice | 04:21 |
| jamielennox | yea - that wasn't even me - thingee did that | 04:21 |
| jamielennox | because v1 is deprecated as of juno they want to have the version less endpoint in the catalog for k | 04:22 |
| jamielennox | dtroyer: so looking back through you should probably override create_plugin rather than get_auth_ref | 04:24 |
| jamielennox | signature is (self, session, version, url, ...) | 04:24 |
| jamielennox | if localhost in 'url' replace it with what you got from self.auth_url | 04:25 |
| jamielennox | then just chain up | 04:25 |
| dtroyer | ok, thanks, I'll give that a shot in the morning | 04:26 |
| jamielennox | it'll get called a couple of times but it's not a big deal i think | 04:26 |
| dtroyer | ah, I see, it's earlier in the sequence…called for each URL in the server data, right? | 04:27 |
| jamielennox | yea, it gets called with every possible version | 04:29 |
| jamielennox | and the first to respond wins | 04:29 |
| jamielennox | so it'll ask for a plugin to match v3 | 04:29 |
| jamielennox | then it'll ask for a plugin to match v2 | 04:29 |
| jamielennox | or whatever combination actually happens | 04:29 |
| jamielennox | realisticly it'll probably never happen more than a couple of times as keystone only specifies 2 versions | 04:30 |
| jamielennox | in short yes - once for each URL found in discovery | 04:30 |
| *** subscope has joined #openstack-sdks | 04:52 | |
| *** briancurtin has quit IRC | 05:02 | |
| *** stevemar has joined #openstack-sdks | 05:22 | |
| *** briancurtin has joined #openstack-sdks | 06:03 | |
| *** briancurtin has quit IRC | 06:07 | |
| *** subscope has quit IRC | 06:08 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-openstackclient: Imported Translations from Transifex https://review.openstack.org/144829 | 06:11 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: consumer https://review.openstack.org/145694 | 06:29 |
| *** jamielennox is now known as jamielennox|away | 06:33 | |
| *** briancurtin has joined #openstack-sdks | 07:03 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token creation docs + tweaks https://review.openstack.org/145705 | 07:04 |
| *** briancurtin has quit IRC | 07:08 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token authorize https://review.openstack.org/145706 | 07:14 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: access token https://review.openstack.org/145707 | 07:22 |
| *** k4n0 has joined #openstack-sdks | 07:38 | |
| *** k4n0 has quit IRC | 07:42 | |
| *** chlong has quit IRC | 07:47 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: image https://review.openstack.org/145711 | 07:54 |
| stevemar | dtroyer, finally done. you can assume i won't be doing anything doc related for a while | 07:55 |
| *** stevemar has quit IRC | 08:02 | |
| openstackgerrit | Zhiyuan Cai proposed openstack/python-openstackclient: Rename title to show name and description for backup list https://review.openstack.org/145718 | 08:44 |
| *** briancurtin has joined #openstack-sdks | 08:52 | |
| *** briancurtin has quit IRC | 08:57 | |
| *** chlong has joined #openstack-sdks | 11:43 | |
| *** Miouge has quit IRC | 12:02 | |
| *** Miouge has joined #openstack-sdks | 12:03 | |
| *** Miouge has quit IRC | 12:08 | |
| *** briancurtin has joined #openstack-sdks | 12:30 | |
| *** briancurtin has quit IRC | 12:34 | |
| *** Miouge has joined #openstack-sdks | 13:03 | |
| *** briancurtin has joined #openstack-sdks | 13:46 | |
| *** briancurtin has quit IRC | 13:50 | |
| *** britthou_ has quit IRC | 14:14 | |
| *** britthouser has joined #openstack-sdks | 14:14 | |
| *** terrylhowe has joined #openstack-sdks | 14:27 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:27 | |
| *** pm90_ has joined #openstack-sdks | 14:34 | |
| *** pm90_ has quit IRC | 14:34 | |
| *** pm90_ has joined #openstack-sdks | 14:34 | |
| *** mattfarina has joined #openstack-sdks | 14:44 | |
| *** mattfarina has quit IRC | 14:57 | |
| *** mattfarina has joined #openstack-sdks | 14:57 | |
| *** Miouge has quit IRC | 15:09 | |
| *** briancurtin has joined #openstack-sdks | 15:21 | |
| *** etoews has joined #openstack-sdks | 15:22 | |
| *** pm90_ has quit IRC | 15:24 | |
| *** stevemar has joined #openstack-sdks | 15:29 | |
| *** pm90_ has joined #openstack-sdks | 15:30 | |
| *** Miouge has joined #openstack-sdks | 15:33 | |
| *** pm90_ has quit IRC | 15:44 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token creation docs + tweaks https://review.openstack.org/145705 | 16:12 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token authorize https://review.openstack.org/145706 | 16:12 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: access token https://review.openstack.org/145707 | 16:12 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token authorize https://review.openstack.org/145706 | 16:16 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: access token https://review.openstack.org/145707 | 16:16 |
| *** chlong has quit IRC | 16:34 | |
| *** pm90_ has joined #openstack-sdks | 17:10 | |
| *** pm90_ has quit IRC | 17:19 | |
| *** pm90_ has joined #openstack-sdks | 17:19 | |
| *** stevemar2 has joined #openstack-sdks | 17:50 | |
| *** stevemar has quit IRC | 17:50 | |
| *** stevemar2 has quit IRC | 17:56 | |
| *** stevemar2 has joined #openstack-sdks | 17:57 | |
| *** briancurtin has quit IRC | 18:38 | |
| *** stevemar2 has quit IRC | 18:50 | |
| *** etoews has quit IRC | 18:59 | |
| *** etoews has joined #openstack-sdks | 19:00 | |
| *** stevemar has joined #openstack-sdks | 19:07 | |
| *** etoews has quit IRC | 19:26 | |
| *** etoews has joined #openstack-sdks | 19:26 | |
| *** briancurtin has joined #openstack-sdks | 19:29 | |
| *** briancurtin has joined #openstack-sdks | 19:29 | |
| *** stevemar has quit IRC | 19:37 | |
| *** stevemar has joined #openstack-sdks | 19:38 | |
| *** etoews has quit IRC | 20:14 | |
| *** subscope_ has joined #openstack-sdks | 20:16 | |
| *** etoews has joined #openstack-sdks | 20:19 | |
| *** subscope_ has quit IRC | 20:28 | |
| *** briancurtin has quit IRC | 20:29 | |
| stevemar | dtroyer, thanks for reviewing :) | 20:30 |
| *** fifieldt_ has quit IRC | 20:32 | |
| *** fifieldt_ has joined #openstack-sdks | 20:33 | |
| dtroyer | stevemar: I'll get to them all eventually… | 20:37 |
| stevemar | dtroyer, i think we'll just have a handful of commands left undoc'ed by the end of what i submitted | 20:37 |
| stevemar | i think just host/hypervisor/sec group/compute agent/compute server | 20:38 |
| openstackgerrit | Merged openstack/python-openstackclient: tweak the server command docs https://review.openstack.org/144937 | 20:50 |
| openstackgerrit | Merged openstack/python-openstackclient: Command doc: volume type https://review.openstack.org/144936 | 20:55 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fixup backup list output https://review.openstack.org/145105 | 21:03 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: backup https://review.openstack.org/145108 | 21:05 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Fix up snapshot command https://review.openstack.org/145114 | 21:05 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: snapshot https://review.openstack.org/145117 | 21:05 |
| *** jamielennox|away is now known as jamielennox | 21:24 | |
| *** pm90_ has quit IRC | 21:28 | |
| stevemar | dtroyer, you're going to hate me for the oauth related changes | 21:35 |
| stevemar | also, if you have any idea on whats going on here, that would be great: https://bugs.launchpad.net/python-openstackclient/+bug/1407805 getting to different outcomes | 21:35 |
| dtroyer | stevemar: so we're going back to RADIUS and RSA tokens? ;) | 21:35 |
| stevemar | hue hue | 21:36 |
| dtroyer | I've not used cinder backups in a while so nothing right off…it'll be interesting to se who's results I manage to replicate | 21:37 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: mapping https://review.openstack.org/145144 | 21:39 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: identity provider https://review.openstack.org/145150 | 21:40 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: identity provider https://review.openstack.org/145150 | 21:41 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: federation protocol https://review.openstack.org/145156 | 21:41 |
| stevemar | dtroyer, http://paste.openstack.org/show/156023/ | 21:46 |
| dtroyer | yeah, I think that's clear enough. also, very similar to how python docs are done ;) | 21:48 |
| dtroyer | well, between minor revs anyway | 21:49 |
| stevemar | excellent | 21:49 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Allow user list to filter by project https://review.openstack.org/142210 | 21:49 |
| stevemar | i'm surprised i haven't had any merge conflicts yet | 21:49 |
| *** ycombinator_ has joined #openstack-sdks | 21:52 | |
| dtroyer | stevemar: are you ready to merge the translations? I don't know the custom on those | 21:54 |
| dtroyer | stevemar: maybe I just found the bit you warned me about in request token authorize… | 21:56 |
| stevemar | dtroyer, i'm sure you did | 21:56 |
| stevemar | dtroyer, in keystone we just merge the translations pretty willy nilly | 21:56 |
| stevemar | dtroyer, whaa... you +2'ed this guy: https://review.openstack.org/#/c/145705/2/openstackclient/identity/v3/token.py i changed a parameter! from --project-id to --project | 21:57 |
| stevemar | i figured you were going to make me keep the old one as a deprecated arg | 21:57 |
| dtroyer | I would have had I noticed… | 21:58 |
| stevemar | and i would have had to plead and claim no one is using this stuff yet | 21:58 |
| dtroyer | what I get for doing reviews during mettings | 21:58 |
| dtroyer | I'll buy that argument... | 21:58 |
| stevemar | pretty sure i'm the only person using oauth stuff :) | 21:59 |
| stevemar | and the intersection of ppl using oauth stuff + osc -> pretty much only me | 21:59 |
| dtroyer | so really, multiple —role options isn't a thing? | 21:59 |
| stevemar | hmm? i don't follow? | 22:00 |
| dtroyer | I'm on the fence about —roles in request token authorize | 22:00 |
| stevemar | it would be --roles admin member | 22:00 |
| stevemar | you saying it should be --role admin --role member | 22:00 |
| stevemar | or just 1 role? | 22:00 |
| stevemar | we handle the delete commands the same way | 22:01 |
| dtroyer | we do multiple properites with —property a=1 —property b=2 etc | 22:01 |
| dtroyer | but for the deletes, thats a positional | 22:01 |
| stevemar | ah | 22:01 |
| stevemar | i see what ya mean | 22:01 |
| dtroyer | in this case there is umlikely to be any ambiguity, | 22:01 |
| dtroyer | so it might be ok, but I'm concerned about opening the nargs door | 22:02 |
| stevemar | normally i do the following: --request-key 123 --roles admin member | 22:02 |
| stevemar | would cliff be smart enough if it were: --roles admin member --request-key 123 | 22:02 |
| stevemar | or would --request-key become part of nargs... | 22:03 |
| *** pm90_ has joined #openstack-sdks | 22:03 | |
| dtroyer | its an argparse issue, and it will most of the time, except when it doesn't | 22:03 |
| openstackgerrit | Merged openstack/python-openstackclient: Command docs: volume https://review.openstack.org/144947 | 22:04 |
| openstackgerrit | Ian Cordasco proposed stackforge/python-openstacksdk: Increase Resource test coverage https://review.openstack.org/136606 | 22:04 |
| dtroyer | the third alternative is to do —role admin,member and parse the argument | 22:05 |
| dtroyer | I don't think we do that a tall so probably shouldn't start now | 22:05 |
| dtroyer | s/a tall/at all/ | 22:06 |
| stevemar | agreed | 22:06 |
| stevemar | so i tried: authorize --roles admin --request-key fc8b71cd4f1e452bb50d6471c3469e01 | 22:06 |
| stevemar | and it works | 22:06 |
| stevemar | i think argparse is smart enough to see the dashes | 22:06 |
| stevemar | sauce: http://paste.openstack.org/show/156030/ | 22:07 |
| dtroyer | it gets it right most of the time…its when something looks like an option, like -1 or something | 22:07 |
| stevemar | fwiw, i think this is a better UX, rather than --role x --role y | 22:08 |
| dtroyer | the other hard part is when a positional foloows —roles, how do you know? in this command there isn't one, but we can't do it if there is | 22:08 |
| stevemar | agreed 100% | 22:08 |
| dtroyer | I agree repeating the option name is sub-optimal, but everything else seems to be worse ;) so any complaints about this get sent to you | 22:09 |
| stevemar | fair enough! | 22:12 |
| stevemar | at which point i release a non-backwards compatible change! | 22:13 |
| stevemar | buahahaha | 22:13 |
| stevemar | the oauth api hasn't changed in a while, i don't expect it to | 22:13 |
| *** ycombinator_ has quit IRC | 22:23 | |
| *** ycombinator_ has joined #openstack-sdks | 22:26 | |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: consumer https://review.openstack.org/145694 | 22:28 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token creation docs + tweaks https://review.openstack.org/145705 | 22:28 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Request token authorize https://review.openstack.org/145706 | 22:28 |
| openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: Command doc: access token https://review.openstack.org/145707 | 22:28 |
| stevemar | dtroyer, you're almost done! just snapshot and image left :P | 22:29 |
| stevemar | kudos sir | 22:29 |
| stevemar | dtroyer, i just +A'ed a bunch of patches that you had initially -1'ed (i fixed up the issues), but let me know if i went on too much of a spree... | 22:39 |
| dtroyer | that's fine, thanks…juggling this many (during the qa meeting no less) gets tricky ;) | 22:40 |
| stevemar | dtroyer, it's all good. | 22:43 |
| dtroyer | I want to make 'neutroned' a thing! | 22:45 |
| dtroyer | then again, I also want to remain friends with MArk and Kyle and co… ;) | 22:46 |
| stevemar | lol | 22:48 |
| stevemar | neutroned is a thing, AFAIC | 22:48 |
| stevemar | alright, i'm outta here, thanks dtroyer | 22:50 |
| dtroyer | that you sir, that's a truckload of work… | 22:50 |
| dtroyer | s/that/thank/ | 22:50 |
| dtroyer | not-exciting work at that | 22:50 |
| *** stevemar has quit IRC | 22:55 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 23:01 | |
| openstackgerrit | Merged openstack/python-openstackclient: Fixup backup list output https://review.openstack.org/145105 | 23:04 |
| openstackgerrit | Merged openstack/python-openstackclient: Command doc: backup https://review.openstack.org/145108 | 23:05 |
| openstackgerrit | Merged openstack/python-openstackclient: Command doc: mapping https://review.openstack.org/145144 | 23:06 |
| openstackgerrit | Merged openstack/python-openstackclient: Command doc: identity provider https://review.openstack.org/145150 | 23:07 |
| openstackgerrit | Merged openstack/python-openstackclient: Command doc: federation protocol https://review.openstack.org/145156 | 23:07 |
| openstackgerrit | Merged openstack/python-openstackclient: Allow user list to filter by project https://review.openstack.org/142210 | 23:07 |
| *** etoews has quit IRC | 23:11 | |
| *** pm90_ has quit IRC | 23:13 | |
| *** pm90_ has joined #openstack-sdks | 23:16 | |
| *** mattfarina has quit IRC | 23:22 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!