Earlier  
Later
New Search
Posted Nick Remark
#openstack-sdks - 2018-11-26
03:12:07 openstackgerrit Lars Kellogg-Stedman proposed openstack/openstacksdk master: do not force interface=admin for identity api v3 https://review.openstack.org/619949
04:23:42 gmann larsks: you need to add task# in cmt message which gets linked to storyboard task which is linked to story
04:24:20 larsks gmann: thanks, I figured that out. Do I need both Story: and Task:, or just the latter?
04:25:22 gmann larsks: both i think but i am not sure exact use case of mentioning story# along with task#.
04:25:34 larsks Okay. Thanks! :)
15:42:32 larsks Hey folks. Can someone take a look at https://review.openstack.org/#/c/619949/ and let me know if that looks sane?
20:55:17 openstackgerrit Slawek Kaplonski proposed openstack/openstacksdk master: Add CRUD methods for Neutron Port Forwarding https://review.openstack.org/616710
#openstack-sdks - 2018-11-27
03:22:19 wjiang hi guys, I am using shade 1.24.0, and always got "Error authenticating to keystone: Problem with auth parameters"
03:23:40 wjiang I put all necessary things in ~/.config/openstack/clouds.yaml, but just got project_name in shade.openstack_cloud().auth without cloud key arg
03:24:30 wjiang if I use specific cloud name, shade.openstack_cloud().auth will got things with u'xxxx' as key and values, so anyone have experience about this?
15:03:40 larsks Hey folks. Can someone take a look at https://review.openstack.org/#/c/619949/ and let me know if that looks sane?
15:11:02 larsks dtantsur|mtg: oh, thanks! I guess I hadn't refreshed my gerrit overview.
15:11:11 dtantsur|mtg :)
15:31:55 openstackgerrit Eric Fried proposed openstack/os-service-types master: Updated from OpenStack Service Type Authority https://review.openstack.org/614786
15:47:45 mordred larsks: +A
15:56:47 larsks mordred: thanks, and also for https://github.com/ansible/ansible/pull/48833 :)
15:58:05 mordred larsks: I just left a shipit on that one actually :)
15:58:43 larsks mordred: yeah, that's okay, now people will stop saying, "have emonty look at it".
15:59:10 mordred \o/
15:59:15 mordred thanks for that, btw
16:05:53 tobias-urdin aren't you suppose to be able to pass paginated yourself?
16:05:54 tobias-urdin https://github.com/openstack/openstacksdk/blob/master/openstack/load_balancer/v2/_proxy.py#L57
16:07:31 mordred tobias-urdin: you *should* never need to - pagination should be autodetected - the parameters are largely leftover when that was all manual
16:08:04 mordred tobias-urdin: is there an issue you're running in to that's making you want to pass a different value there?
16:08:20 tobias-urdin how can i disable pagination? just increase limit somehow to an very large one?
16:09:40 mordred well, that's the thing - you can't really, because there is also a server side max page size setting for most things (Which is opaque to the user)
16:09:51 tobias-urdin oh that was internal to sdk to tell if resource supported pagination
16:10:00 mordred so you could set limit high for a large page size, but the server still may send you only as many as the server config
16:10:46 mordred yah - and we've since added support for telling if the server sent paginated results automatically, so the parameter isn't super useful anymore - we just haven't gone through and gotten rid of it yet
16:12:40 tobias-urdin ack, thanks
16:40:11 larsks What is the typical latency between "fix merging" and "fix shows up on pypi"?
16:41:23 mordred larsks: depends ... we need to tag a release for it to show up on pypi - which can usually happen pretty quickly. I'm fighting a broken laptop at the moment though and have been out for US Thanksgiving, so I'm a little behind on grokking the current state of things
16:41:43 mordred larsks: I'd imagine we could get something cut by the end of the week at least though
16:43:08 larsks No worries. I've got a trio of problems that have conspired to give me a set of "unsharable" ansible playbooks because they require fixes to openstacksdk, the ansible os_server plugin, and the openstack inventory plugin. I'm hoping to get to the point that I don't need to provide people with manual patching instructions, but it's not critical.
16:44:27 mordred larsks: I would also love for you to not need to give people manual patching instructions
16:44:28 mordred larsks: :)
16:44:43 mordred larsks: I think that inventory plugin patch looks backportable
16:45:10 larsks Hopefully also https://github.com/ansible/ansible/pull/48798
16:48:22 mordred larsks: yah. that one too (and I have shipit'd that one too)
16:48:29 larsks \o/
17:06:07 openstackgerrit Merged openstack/openstacksdk master: do not force interface=admin for identity api v3 https://review.openstack.org/619949
17:26:14 larsks mordred: another openstacksdk review, but not critical (rarely used code path, I think; only tripped over it due to a misconfiguration): https://review.openstack.org/#/c/618519/
17:36:18 openstackgerrit Merged openstack/os-service-types master: Updated from OpenStack Service Type Authority https://review.openstack.org/614786
17:37:21 mordred larsks: sweet. thanks
20:27:05 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Register proxy directly in add_service https://review.openstack.org/620398
#openstack-sdks - 2018-11-28
07:47:26 openstackgerrit Vieri proposed openstack/python-openstackclient master: add python 3.6 unit test job https://review.openstack.org/620511
08:15:50 dtantsur|mtg mordred: morning! could you check https://review.openstack.org/#/c/619594/ as well? it may be a bit controversial, so ping me if you need to discuss :)
14:06:54 mordred dtantsur|mtg: haven't read the patch yet, but I agree with the commit message
14:07:01 dtantsur|mtg good start :)
14:07:21 mordred dtantsur|mtg: in the shade layer we've been only working with detail records as well
14:07:29 dtantsur|mtg yeah
14:13:25 mordred dtantsur|mtg: if we go ahead and apply it as a pattern across the board, perhaps we merge your ListMixin into the base Resource and add a flag like has_detail or something?
14:14:02 gtema I would also vote for that, since it is quite broadly spreaded
14:14:32 dtantsur|mtg mordred: sounds good. I can do it, thought I'd prefer in a follow-up :) this thing blocks migrating metalsmith to openstacksdk.
14:15:14 gtema dtantsur|msg: I also think a follow-up would be better
14:15:14 dtantsur|mtg or I can do it here, but only for ironic for now (to avoid bloating the patch)
14:15:15 mordred oh - yeah - I just +Ad your patch ... more thinking out loud
14:15:27 mordred so yeah, follow up sounds great
14:15:40 dtantsur|mtg good :)
14:43:36 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Support remote vendor profiles https://review.openstack.org/616228
14:43:36 openstackgerrit Monty Taylor proposed openstack/openstacksdk master: Slightly refactor vendor profile loading https://review.openstack.org/616227
14:48:48 openstackgerrit Merged openstack/openstacksdk master: Make timeouts in functional tests configurable https://review.openstack.org/619008
14:57:02 mordred dtantsur|mtg: ^^ I updated 616227 from your comment about preferring a global variable
15:01:55 dtantsur|mtg cool, thnx
15:09:24 mordred gtema: comments and questions on https://review.openstack.org/#/c/618993
15:10:00 gtema mordred: ok, thanks. Will look at it
15:11:47 cdent mordred: I recall you had a script that would parse the max (micro)versions out of the services in a catalog and produce a little report. Is that around somewhere and consumable?
15:15:12 mordred cdent: it's definitely around somewhere ...
15:15:18 mordred cdent: the question is ... WHERE
15:15:40 cdent i'm certain you can find it because you used all caps. that will give extra power.
15:16:28 mordred cdent: oh - right. I added it as a command to osc
15:16:32 mordred cdent: openstack versions show
15:16:43 cdent oh, nice
15:17:21 mordred even better than a script!
15:18:32 cdent totes
15:18:44 cdent those all caps were super powerful today
15:23:20 mordred SO POWERFUL
15:29:02 frickler oh, that's a nice one indeed. makes it pretty obvious that I should tackle moving neutron and designate away from using specific ports in their devstack deployment ;)
15:30:27 mordred frickler: our obsession with running services on specific ports is so strange to me
15:33:20 frickler mordred: well, the origin of that to me always was that you cannot bind two different processes to the same port. and after that inertia takes over
15:34:54 mordred ++
15:37:38 frickler mordred: I just noticed that for the image service, it lists eight different versions instead of microversions. is that a bug or is glance doing weird stuff (again)?
15:39:26 dtantsur|mtg frickler: glance has many major versions looking like microversions IIRC
15:43:23 openstackgerrit Merged openstack/openstacksdk master: Change approach to detailed listings of baremetal resources https://review.openstack.org/619594
15:44:10 mordred frickler: glance doing weird things
15:44:35 mordred frickler: they are not selectable - so you can't use them like microversions and say "hey, I see you have 2.1 and 2.2, I want to talk to 2.1"
15:44:50 dtantsur|mtg mordred: do you think you could release the SDK with the just merged patch? I'd give the metalsmith migration another try.
15:45:04 mordred dtantsur|mtg: sure thing
15:46:26 mordred dtantsur|mtg: remote: https://review.openstack.org/620639 Release 0.20.0 of openstacksdk
15:47:32 dtantsur|mtg thnx!
15:47:47 mordred larsks: ^^ release request I think interests you too
16:05:06 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add support for client-side rate limiting https://review.openstack.org/605043
16:07:28 openstackgerrit Monty Taylor proposed openstack/keystoneauth master: Add support for client-side rate limiting https://review.openstack.org/605043
16:38:28 openstackgerrit Merged openstack/osc-lib master: Remove pypy job https://review.openstack.org/617336
16:46:11 larsks mordred: ooo, nifty. Thanks!
17:39:54 openstackgerrit Merged openstack/openstacksdk master: Update link address for vendor support https://review.openstack.org/619830
18:01:24 nicolasochem__ Hello, I would like to use the sdk to validate a heat stack update. It's possible with the CLI (--dry-run command).
18:01:32 nicolasochem__ This is now the CLI does it : https://github.com/openstack/python-heatclient/blob/master/heatclient/v1/stacks.py#L185-L189
18:02:08 nicolasochem__ But I don't see this call anywhere in the openstacksdk source code. Am I missing anything, or is it missing from sdk altogether ?
18:02:38 nicolasochem__ Note that you can do a dry-run of stack creation with the sdk. My question is about stack update.

Earlier   Later