Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2019-02-05
15:45:28 johnsom flavor_data is technically a string, so *maybe*?
15:45:52 mordred johnsom: yeah. I mean - it would *physically* work :)
15:47:52 mordred johnsom: dare I ask why flavor_data is a json string and not just a dict?
15:47:58 gtema anyway - removing my "-1". Sorry and thanks for clarification
15:48:24 mordred gtema: thanks for reviewing deeply!
15:48:25 gtema because it is stated in docs: flavor_databodystringThe JSON string containing the flavor metadata.
15:48:35 johnsom It's a string we share with third party drivers, so wanted to make it as neutral as possible.
15:48:48 gtema mordred: welcome
15:49:46 johnsom Basically the whole path for that string is json, including the schema validation, so it seemed silly to keep flipping back and forth with it.
15:50:04 johnsom Plus it makes the CLI super easy
15:53:28 gtema mordred: seen some changes with hints to upcoming R1.0. Any concrete plans on that?
15:58:34 mordred gtema: I think there's three things I'd like to see done first ...
15:58:40 mordred gtema: https://review.openstack.org/#/c/604926/ - remove task manager
15:59:05 mordred gtema: https://review.openstack.org/#/c/630912/ align shade/sdk server objects and use server in shade layer
15:59:45 mordred gtema: https://review.openstack.org/#/c/609684/ - same thing with image
16:00:28 mordred (mostly because at least for image, I'm 100% certain it's going to break the existing image api in the sdk layer - even though I'm also 100% sure that api isn't terribly usable right now)
16:00:46 mordred I think most of the rest of the objects are actually in pretty good shape
16:00:47 gtema ok. Bad question: any "estimation timeline"?
16:00:50 mordred hehe
16:01:24 mordred I don't think those are too far off - the hardest one is finishing landing the semaphore patch to ksa
16:01:38 mordred so "soon"? :)
16:01:45 gtema agree, looks "stalled"
16:02:38 mordred yah - I keep meaning to poke kmalloc in the face to see if we can do the final push and get it in there
16:02:43 gtema ok. I wanted to probably bring DNS in time and potentially add "create container from directory" for object
16:03:02 kmalloc the fair semaphore?
16:03:04 mordred ++
16:03:21 kmalloc because i think we just need to have a real functional test for it.
16:03:23 mordred kmalloc: yah. you went and took vacation, which is totally my excuse for that patch
16:03:25 kmalloc otherwise it's ready
16:03:39 kmalloc the few -1 nits can be cleaned up after.
16:04:35 mordred gtema: oh - we should do the same thing for object as those patches for server and image - we've got comprehensive object upload support in the shade layer, including handling SLO/DLO for large objects
16:04:35 kmalloc i can try an spin up a test for it, but i admit i'm a bit out of my element for that (mostly because my brain is still recovering from vacation, almost all caught up finally)
16:04:53 kmalloc it's amazing how much time a vacation takes to fully recover from :P
16:04:58 mordred kmalloc: yeah - I think we each have almost half of the story going on
16:05:03 gtema mordred: oh, yeah
16:05:04 kmalloc ++
16:05:34 gtema mordred: ok, will have a look on that
16:05:35 mordred kmalloc: I'm a little sad we haven't managed to annoy cmurphy enough with the incomplete code that she just finished it for us
16:06:03 mordred gtema: sweet.
16:06:06 kmalloc i make an effort to not annoy cmurphy for those things :P she's busy enough (she always seems to have multiple meetings going on)
16:06:48 cmurphy i make an effort not to look at ksa when i don't have to
16:08:35 kmalloc haha
16:10:14 kmalloc mordred: you're not in -meeting-alt, just asked for some eyes to make sure that interface is something folks are ok with
16:10:20 kmalloc mordred: just FYI
16:12:30 mordred kmalloc: woot thanks
19:03:40 openstackgerrit Artem Goncharov proposed openstack/openstacksdk master: WIP Add DNS support https://review.openstack.org/623009
19:57:58 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Make all resource locations process project_id https://review.openstack.org/634572
19:57:58 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Rename compute.service.zone to availability_zone https://review.openstack.org/634573
19:57:59 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Stop mocking method in fwaas test https://review.openstack.org/634574
21:01:33 openstackgerrit Artem Goncharov proposed openstack/openstacksdk master: WIP Add DNS support https://review.openstack.org/623009
21:15:24 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: More state handling in volume transfer requests functional tests https://review.openstack.org/634776
21:23:26 openstackgerrit Artem Goncharov proposed openstack/openstacksdk master: WIP Add DNS support https://review.openstack.org/623009
#openstack-sdks - 2019-02-06
09:04:19 openstackgerrit Vlad Gusev proposed openstack/python-openstackclient master: Add support for get details of Quota https://review.openstack.org/515401
12:21:33 openstackgerrit Lajos Katona proposed openstack/openstacksdk master: Add agent property: resources-synced https://review.openstack.org/635143
15:08:58 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: More state handling in volume transfer requests functional tests https://review.openstack.org/634776
19:09:53 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: More state handling in volume transfer requests functional tests https://review.openstack.org/634776
21:41:05 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: More state handling in volume transfer requests functional tests https://review.openstack.org/634776
#openstack-sdks - 2019-02-07
02:52:57 dtroyer mordred, amotoki: ^^^ should finally ease some of the pain with the volume transfer functional tests, I think there is something going on in the back-end between parallel tests and using names rather than IDs, but that may be a Cinder/cinderclient lib issue to follow-up on.
06:48:16 openstackgerrit Adrian Turjak proposed openstack/service-types-authority master: Add Adjutant as operator-logic https://review.openstack.org/633854
09:32:50 mordred dtroyer: I'm a fan of more stableness
09:33:47 gtema mordred: same for me ;)
09:35:07 mordred gtema: :)
10:15:18 openstackgerrit Christian Schneemann proposed openstack/python-openstackclient master: Typo fix https://review.openstack.org/635471
15:56:10 edleafe API-SIG Office Hour: the fun begins in 4 minutes!!!!
15:59:27 elmiko lol
16:00:55 edleafe Having fun yet, elmiko?
16:01:11 elmiko it just started, i'm still ramping up my fun scalar
16:01:37 edleafe Good idea - pace yourself
16:01:41 elmiko haha, ++
16:01:44 elmiko how about you edleafe ?
16:03:57 edleafe I'm at peak fun level right now
16:04:40 elmiko haha
16:27:46 cdent I keep wishing I had an api canard to bring
16:27:52 elmiko lol
16:27:59 elmiko bring it!
16:28:08 elmiko json home anyone?
16:28:10 edleafe What, and make us work??
17:02:14 elmiko catch ya later folks, have a good weekend =)
17:02:24 edleafe c ya, elmiko!
#openstack-sdks - 2019-02-08
14:02:06 openstackgerrit Lajos Katona proposed openstack/openstacksdk master: Add agent property: resources-synced https://review.openstack.org/635143
#openstack-sdks - 2019-02-09
20:38:27 openstackgerrit Colleen Murphy proposed openstack/keystoneauth master: [WIP] Expose app creds in token and fixtures https://review.openstack.org/636030
#openstack-sdks - 2019-02-10
23:51:49 openstackgerrit Colleen Murphy proposed openstack/keystoneauth master: Expose app creds and new attrs in fixtures https://review.openstack.org/636030
23:51:50 openstackgerrit Colleen Murphy proposed openstack/keystoneauth master: Expose application credentials in AccessInfoV3 https://review.openstack.org/636074
#openstack-sdks - 2019-02-11
17:39:11 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: This fix removes an erroneous underscore found within the name of a function named test_snapshot_delete within the test test_snapshot.py found in both volume v1 and v2 of python-openstackclient. https://review.openstack.org/636186
19:42:43 openstackgerrit Jeremy Houser proposed openstack/python-openstackclient master: This fix removes an erroneous underscore found within the function named test_snapshot_delete within test_snapshot.py found in both volume v1 and v2 of python-openstackclient. https://review.openstack.org/636186
20:45:33 openstackgerrit Merged openstack/python-openstackclient master: More state handling in volume transfer requests functional tests https://review.openstack.org/634776
21:43:55 openstackgerrit Colleen Murphy proposed openstack/keystoneauth master: Expose app creds and new attrs in fixtures https://review.openstack.org/636030
23:01:39 openstackgerrit Merged openstack/osc-lib master: Allow use of user-id as well as username for authentication https://review.openstack.org/633964
#openstack-sdks - 2019-02-12
01:02:41 johnsom FYI, the dependent patch for the Octavia SDK updates has now merged. so the chain starting at https://review.openstack.org/#/c/634010 just need a second +2
01:32:50 openstackgerrit Merged openstack/osc-lib master: Change openstack-dev to openstack-discuss https://review.openstack.org/622433
10:01:33 frickler why does that location look so munched? is that a known issue? seen on a fresh devstack deployment http://paste.openstack.org/show/744909/
11:31:16 frickler seems related to openstacksdk==0.24.0, don't see this with 0.23.0
11:37:59 frickler mordred: Shrews: git bisect blames https://review.openstack.org/626380 ^^ it seems to show up for all objects from the network api at least
12:28:45 openstackgerrit Vlad Gusev proposed openstack/python-openstackclient master: Add support for get details of Quota https://review.openstack.org/515401
13:56:14 mordred frickler: it's a field that we added to all objects in shade, and that patch pushed it down into the sdk layer
13:56:30 mordred frickler: we should potentially teach osc about rendering it prettier?
14:00:31 frickler mordred: I'm wondering whether it is useful to a CLI user at all, maybe osc should just discard it?
14:01:04 mordred frickler: yeah - maybe so?
14:02:31 frickler mordred: certainly for stuff that isn't related to a project at all like this http://paste.openstack.org/show/744939/
14:03:26 mordred ++

Earlier   Later