Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2017-04-05
15:47:37 ankur-gupta-f4 reedip: i don't think so, since I've had to do it a couple times already
15:48:05 reedip dtroyer : ^^ shouldnt the proposal bot pick it up ?
15:48:52 openstackgerrit Stephen Finucane proposed openstack/cliff master: Add 'cliff-commands' Sphinx directive https://review.openstack.org/453383
15:48:52 openstackgerrit Stephen Finucane proposed openstack/cliff master: Add 'cliff-command' Sphinx directive https://review.openstack.org/450322
15:49:03 ankur-gupta-f4 once merged to global-req, then proposal bot bumps it for all the individual projects as well... i believe
15:49:42 openstackgerrit David Rabel proposed openstack/python-openstackclient master: Add --network and --port to server create https://review.openstack.org/444924
15:50:25 rabel_b1 new proposal for you ::)
15:50:51 reedip ankur-gupta-f4 : yeah it should, and you dont need to do it yourself, in the Ideal scenario
15:50:55 reedip rabel_b1 : thanks :)
15:53:13 rabel_b1 reedip: it has some weeknesses that should be fixed when migrating to MultiKey...andSoOn. but i think for now it should be fine.
15:54:29 reedip rabel_b1 : ok , will put it in the review tab :)
15:55:23 sindhu ankur-gupta-f4: but we don't need SDK now for 383025 to work right?
15:56:53 reedip we dont ... I agree
15:57:10 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: Add "qos-policy" option to "port create" & "port set" https://review.openstack.org/419236
15:57:51 openstackgerrit Reedip proposed openstack/python-openstackclient master: "floating ip set/unset port" for OSC https://review.openstack.org/383025
15:59:29 dtroyer reedip, ankur-gupta-f1: the proposal bot creates reviews for projects based on what gets merged into the requirements repo. We have to propose the reviews to requirements.
15:59:44 dtroyer ankur-gupta-f1: what did you want input on?
16:00:53 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Policy for Octavia (load balancing) https://review.openstack.org/451980
16:01:29 ankur-gupta-f4 dtroyer: https://review.openstack.org/#/c/446223/
16:01:38 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Rule for Octavia (load balancing) https://review.openstack.org/452832
16:01:42 ankur-gupta-f4 to see if its in the fashion that looks okay/works for you.
16:02:23 ankur-gupta-f4 https://review.openstack.org/#/c/447068/ is piece in OSC to initialize it.
16:02:50 ankur-gupta-f4 Bout to head out to airport so will be on/off for the next couple hours but will log back in on the plane.
16:02:56 mordred cdent: the example iin https://specs.openstack.org/openstack/api-wg/guidelines/microversion_specification.html#version-discovery says it was taken from nova, but it does not match what nova emits
16:03:19 cdent a) /me shakes tiny fist
16:03:24 cdent b) predates me
16:03:35 mordred k. I mena- I like what's in the spec better
16:03:36 cdent c) give me a few minutes and I'll look into it more closely and fix it
16:03:47 openstackgerrit David Rabel proposed openstack/python-openstackclient master: Add --network and --port to server create https://review.openstack.org/444924
16:03:50 mordred but I'm wondering if we should align the spec to nova or try to get nova to add an additional key to theirs?
16:03:55 cdent it is quite likely that what's in the spec is desirable
16:04:03 mordred cdent: (the tl;dr is nova uses "version" to mean max_version)
16:04:27 cdent oh yeah, that's an intentional change
16:04:30 mordred ok. cool
16:04:32 cdent the surrounding tesxt is wrong
16:04:42 mordred I'll make the project-navigator stuff align to the api-wg thing
16:05:00 cdent yes please
16:06:19 mordred cdent: also - for major versions without microversions, min_version and max_version can be omitted or contain empty strings? or only be omitted? or do you care?
16:08:24 cdent mordred: as far as I'm aware we haven't yet provided guidance for discovery in a non microversioned world, but my interpretation is that 'id' takes on the meaning of "the actual version you're getting, majorly" and min and max ought to be omitted
16:08:48 reedip sindhu : I updated the latest PS with the changes mentioned by dtroyer
16:09:09 cdent mordred: so, basically, keep id, put the version in there, leave min and max out
16:09:43 sindhu reedip: great! thanks :)
16:10:04 dtroyer is the presence/absense of min_version and max_version how we would detect microversion support?
16:10:31 cdent dtroyer: that seems to be the implication
16:10:38 cdent but I'm not sure if that's intentional or not
16:10:40 openstackgerrit Reedip proposed openstack/python-openstackclient master: "floating ip set/unset port" for OSC https://review.openstack.org/383025
16:10:51 dtroyer I'm generalizing the client-side negotiation code to put into ksa…
16:12:52 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: "floating ip set/unset port" for OSC https://review.openstack.org/383025
16:13:21 mordred cdent: https://review.openstack.org/453361
16:13:49 mordred dtroyer: that is my hunch, yes
16:14:27 cdent mordred: cool, will look properly soon
16:14:29 mordred dtroyer: and ooh - you adding version discovery and microversion-per-call suport? or just one or just theother?
16:14:50 dtroyer I've started with the negotiation, will be doing per-call since OSC needs it
16:15:04 mordred woot
16:15:08 dtroyer mordred: https://review.openstack.org/451618
16:15:15 mordred I've been avoiding bothering with that so far
16:15:18 dtroyer it's all in OSC for development ATM
16:19:21 mordred dtroyer: ah - so cool - I _also_ want major version discovery and have a minor portion of that somewhere - but can add that in when you're happy with what you've got there for microversions
16:19:25 mordred (and I like what you've got there)
16:29:28 openstackgerrit Chris Dent proposed openstack/api-wg master: Remove reference to nova on version discovery https://review.openstack.org/453744
16:35:11 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Policy for Octavia (load balancing) https://review.openstack.org/451980
16:38:09 mordred dtroyer: also, I finally got around to writing a patch to horizon to add suport for downloading clouds.yaml files
16:38:20 mordred (fixing tests right now)
16:38:49 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Policy for Octavia (load balancing) https://review.openstack.org/451980
16:39:04 dtroyer mordred: \o/
16:44:04 mordred dtroyer: remote: https://review.openstack.org/453669 Add support for downloading clouds.yaml files
16:50:49 dtroyer mordred: Nice. Maybe OSC should grow this too, and maybe both need to have a way to pull the update for the vendors file?
16:52:03 mordred dtroyer: yah - I have some patches I'll push up later today about vendors files too
16:52:28 dtroyer coolio
16:52:29 mordred dtroyer: and I've been thinking of making a few osc commands like "add cloud" that could injest a clouds.yaml from a vendor into an existing one
16:53:20 dtroyer yeah, thats a natural place to manage that file...
16:53:34 mordred ++
16:57:17 reedip stevemar: there ?
17:00:49 openstackgerrit Merged openstack/python-openstackclient master: Enable to create legacy router https://review.openstack.org/449176
18:10:40 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Policy for Octavia (load balancing) https://review.openstack.org/451980
18:46:19 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: Develop api.versions for microversion discovery https://review.openstack.org/451618
18:51:39 openstackgerrit Nakul Dahiwade proposed openstack/python-openstacksdk master: Introduce L7Rule for Octavia (load balancing) https://review.openstack.org/452832
18:54:46 openstackgerrit Dean Troyer proposed openstack/python-openstackclient master: Clean up password prompt work-arounds https://review.openstack.org/453832
19:06:04 malloc83 hi there :-) just a tad curious why this isnt covered or just not documented.. i just started using the openstack proposed sdk.. and realised after authenticating and using authorize to get a token.. it pretty much doesnt mean much,, because after the user browses around on my gui, i have zero way of using the token to authenticate again on the sdk ;-)
19:06:52 briancurtin You can auth again with that token
19:06:57 malloc83 is this just not possible at this time? I noticed there is a plugin ? password? but nothing stating i could use plugin token if it even exists..
19:07:16 malloc83 bradjones: yup i can,, but not using the sdk.. i cant find any intercace or something via the connection class that allows that.
19:07:25 briancurtin huh?
19:07:54 malloc83 so openstack has a connection class that i am using to connect in python, i use authorise() method to get a token...
19:08:00 malloc83 after that i can use the token again against the api...
19:08:09 malloc83 but how do i use it again with the connection handler?
19:08:41 malloc83 since the connection handler has been re-initilized, i cant afford to store a username and password in memory to auth again :) it must have some way for me to provide the token to continue the session
19:09:03 briancurtin as in, you want to create another connection instance elsewhere, perhaps in another instance of your application, using the token that was already received?
19:09:20 malloc83 bradjones: spot on.. this is for wsgi app..
19:09:46 malloc83 briancurtin: i am sorry spot on yes.. :) getting used to this new mac keyboard.
19:10:08 briancurtin malloc83: that has definitely worked in the past. let me write something up and send it to you in a min
19:10:11 malloc83 perhaps i missed this in the documentation, looked through all the examples too.
19:10:16 openstackgerrit Chris Dent proposed openstack/api-wg master: Remove reference to nova on version discovery https://review.openstack.org/453744
19:11:52 malloc83 briancurtin: it works great, just no idea how to re-use the token with the connection handler.. other than that, everything seems to be as expected.. the documentation however has no mention of using token for auth, just v2,v3 keystone username and password, unscoped etc
19:15:39 briancurtin malloc83: is this what you’re looking for https://gist.github.com/briancurtin/a10f4ed985d51a64f9c5de4c63b685fc
19:15:54 malloc83 having a look quick
19:16:22 briancurtin malloc83: yeah, we use keystoneauth for all of the auth handling, so we need to do a better job either explaining some of what can be done, or link out to keystoneauth’s docs on it
19:16:28 malloc83 i guessed it exactly like that, didnt work.. :) but that would be worth putting in the docs.
19:16:50 malloc83 ill have a look now why, perhaps i broke it :)
19:16:59 briancurtin malloc83: what version are you using?

Earlier   Later