| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2020-03-09 | |||
| 14:55:41 | rm_work | thanks :) | |
| 14:56:39 | brtknr | mordred: yay | |
| 15:09:21 | slaweq | mordred: I just commented in https://review.opendev.org/#/c/711513/ | |
| 15:17:20 | mordred | slaweq: oh good point on the request-side | |
| 15:17:51 | mordred | that seems like something we should add a more general feature to the neutron proxy / resources for | |
| 15:18:17 | mordred | "query_parameter, requires_extension='blah'" ... or something. | |
| 15:18:44 | dtantsur | something similar to microversions, but with extensions? | |
| 15:19:27 | slaweq | mordred: dtantsur yes, we need something like that | |
| 15:19:48 | slaweq | otherwise we will break sdk for clouds without this extension enabled | |
| 15:22:16 | mordred | yah | |
| 15:23:46 | mordred | dtantsur: although I think the difference in this case is for neutron we'd want to be able to tie individual properties of a resource to a needed extension so that attempting to set one if the given extesion isn't there can return a decent error | |
| 15:23:48 | mordred | hrm | |
| 15:23:54 | mordred | actually ... | |
| 15:25:13 | mordred | slaweq: do we really need to do that? if someone tries to set stateful on a cloud that doens't have it - neutron will just throw the error, right? so an sdk-side indication that it needs an extension would be an optimization? | |
| 15:27:33 | slaweq | mordred: actually You're right, if user will not pass "stateful" parameter during e.g. creation of SG then it will be fine | |
| 15:28:03 | slaweq | so maybe we don't need any additional discovery | |
| 15:31:51 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/python-openstackclient master: Add "fields" parameter to ListSecurityGroup query https://review.opendev.org/710825 | |
| 15:43:07 | openstackgerrit | Merged openstack/openstacksdk master: Extract check temp_url_key logic https://review.opendev.org/710973 | |
| 16:05:15 | openstackgerrit | Merged openstack/openstacksdk master: Add retired and retired_reason fields to baremetal node https://review.opendev.org/711032 | |
| 16:14:33 | mgoddard | Hi, seeing some issues with os_ironic in Ansible 2.9.6 and openstacksdk 0.41.0 | |
| 16:14:58 | mgoddard | If a node already exists, we get KeyError: 'uuid' | |
| 16:15:45 | mgoddard | it seems to be because the node create response contains a 'uuid' field, but node patch returns 'id'. Possibly something Munch is doing? | |
| 16:15:50 | mgoddard | dtantsur: seen that ^? | |
| 16:17:22 | TomStappaerts | slaweq could you specify what you would like to see added at https://review.opendev.org/#/c/711513/ ? Would just a stateful=True check be sufficient or do you want to create two SG and check on those? | |
| 16:17:37 | TomStappaerts | And am I correct in saying that would be a follow-up patchset? | |
| 16:23:59 | dtantsur | mgoddard: first time I hear about it | |
| 16:24:14 | dtantsur | have a traceback somewhere? | |
| 16:24:50 | mgoddard | dtantsur: as much of a traceback as you'll get from ansible | |
| 16:24:57 | dtantsur | heh | |
| 16:25:05 | mgoddard | os_ironic.py L355 | |
| 16:25:13 | mgoddard | os_ironic.py L331 | |
| 16:25:27 | mgoddard | they line up with the ansible stable-2.9 branch | |
| 16:25:33 | dtantsur | sdk is supposed to return id in all cases | |
| 16:25:45 | mgoddard | os_ironic always uses uuid :) | |
| 16:25:49 | mgoddard | and previously worked | |
| 16:27:53 | dtantsur | why on earth would it work previously, I wonder... | |
| 16:28:41 | mgoddard | because sdk returned uuid :) | |
| 16:28:59 | dtantsur | it wasn't supposed to | |
| 16:29:08 | dtantsur | and the last change touching this code landed a while ago | |
| 16:29:22 | dtantsur | mgoddard: do you have an idea what the latest working version is? | |
| 16:30:20 | mgoddard | dtantsur: I haven't dug into it yet. Worked around and plan to invetigate later | |
| 16:31:55 | dtantsur | mgoddard: I can see how it could get broken.. but it was back in train | |
| 16:32:29 | mgoddard | dtantsur: possible we just haven't hit it yet | |
| 16:33:09 | dtantsur | sigh | |
| 16:33:12 | dtantsur | lemme try something | |
| 16:35:54 | openstackgerrit | Merged openstack/openstacksdk stable/stein: Add "name" filter in "list" call when retrieving a single register https://review.opendev.org/711314 | |
| 16:45:21 | dtantsur | mgoddard: I think I've figured it, patch incoming | |
| 16:45:51 | mgoddard | dtantsur: nice work! | |
| 16:45:56 | mgoddard | I can test locally | |
| 16:54:38 | openstackgerrit | Dmitry Tantsur proposed openstack/openstacksdk master: Consistent normalization of Machine objects in the cloud layer https://review.opendev.org/711952 | |
| 16:54:41 | dtantsur | mgoddard, mordred ^^^ | |
| 17:00:31 | mgoddard | dtantsur: testing | |
| 17:11:38 | mordred | dtantsur: lgtm | |
| 17:12:01 | mordred | dtantsur: one of these days I'm going to finish getting the cloud layer to just use resource objects ... | |
| 17:12:07 | dtantsur | yeah | |
| 17:12:13 | mordred | too many things to do | |
| 17:12:21 | dtantsur | so true | |
| 17:37:14 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Import generate_temp_url from swiftclient https://review.opendev.org/711640 | |
| 17:42:13 | openstackgerrit | Riccardo Pittau proposed openstack/openstacksdk master: Set min version to test node retirement https://review.opendev.org/711958 | |
| #openstack-sdks - 2020-03-10 | |||
| 01:30:48 | openstackgerrit | Merged openstack/openstacksdk stable/train: Include "fields" to "SecurityGroup" query parameters https://review.opendev.org/711622 | |
| 01:46:09 | openstackgerrit | Sean McGinnis proposed openstack/python-openstackclient master: Change 'Volume' to 'Block Storage' https://review.opendev.org/532439 | |
| 02:03:23 | openstackgerrit | Sean McGinnis proposed openstack/python-openstackclient master: Disallow setting default on internal network https://review.opendev.org/539303 | |
| 02:16:49 | openstackgerrit | Sean McGinnis proposed openstack/python-openstackclient master: Update http links in docs https://review.opendev.org/567729 | |
| 02:34:07 | openstackgerrit | Sean McGinnis proposed openstack/python-openstackclient master: Compute: Add tag support for server add port https://review.opendev.org/580977 | |
| 09:59:20 | frickler | slaweq: could you take a look at https://review.opendev.org/539303 please? I'm not sure this really makes sense in all situations | |
| 10:00:03 | slaweq | frickler: sure, looking | |
| 10:03:51 | slaweq | frickler: IMO it makes sense | |
| 10:04:05 | slaweq | frickler: according to the https://docs.openstack.org/neutron/latest/admin/config-auto-allocation.html at least | |
| 11:02:44 | dtantsur | morning folks, looking for a 2nd +2 on https://review.opendev.org/#/c/711952/ it fixes os_ironic with newer SDK | |
| 11:29:48 | frickler | slaweq: hmm, o.k., thx for checking | |
| 11:56:33 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: Add option for modules_default groups https://review.opendev.org/712057 | |
| 11:59:30 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: Add option for modules_default groups https://review.opendev.org/712057 | |
| 12:12:09 | openstackgerrit | Sagi Shnaidman proposed openstack/ansible-collections-openstack master: Add option for modules_default groups https://review.opendev.org/712057 | |
| 13:04:08 | smcginnis | Anyone aware of a change in the SDK that would cause this: https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_f9e/711930/2/check/openstacksdk-functional-devstack/f9e911a/testr_results.html | |
| 13:04:23 | smcginnis | Failing on a requirements update patch that doesn't seem to be related. | |
| 13:06:10 | mordred | smcginnis: hrm - yes, possibly - but why would it not have failed on that patch that did it | |
| 13:06:12 | mordred | looking | |
| 13:06:55 | smcginnis | I always ask myself that with these kinds of things. :) | |
| 13:07:11 | mordred | smcginnis: I take it back - I don't see any reason for that | |
| 13:07:23 | rpittau | I'm also seeing some weird errors in testing https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_97c/711640/3/check/openstack-tox-py37/97cfb62/testr_results.html | |
| 13:08:43 | dtantsur | Oo | |
| 13:09:54 | mordred | smcginnis: we might want to check to see if nova recently landed any changes disallowing deleting aggregates by name | |
| 13:13:22 | mordred | rpittau: https://zuul.opendev.org/t/openstack/build/97cfb620a44b4c91b886e02318306ea2/log/job-output.txt#4355 - and https://zuul.opendev.org/t/openstack/build/97cfb620a44b4c91b886e02318306ea2/log/job-output.txt#4166 | |
| 13:13:35 | mordred | rpittau: I thnik something maybe timed out? | |
| 13:13:40 | mordred | and it hit a hard-kill | |
| 13:13:52 | mordred | doesn't look like it was that test - looks like it was another one? | |
| 13:14:07 | rpittau | mordred: yeah, the odd part is that it's happening in other jobs too | |
| 13:14:17 | rpittau | I triggered another recheck | |
| 13:14:44 | mordred | *excellent* | |
| 13:18:57 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Fix aggregate functional test for id restriction https://review.opendev.org/712068 | |
| 13:19:39 | mordred | smcginnis, dtantsur: ^^ if you don't mind - nova just landed a change that has broken our functional tests | |
| 13:19:59 | mordred | I'd complain at nova - but it's actually just causing a broken test to be shown to be broken | |
| 13:21:15 | mordred | hrm | |
| 13:21:26 | mordred | actually - no - I think I'm going to go complain | |
| 13:23:41 | rpittau | :D | |
| 13:28:11 | smcginnis | :) | |
| 13:29:44 | mordred | ok. I complained - and I'm back to thinking it's a bug on our side | |
| 13:30:16 | mordred | our delete method has a name_or_id argument, but it just passes it on directly without attempting to get an id | |
| 13:30:18 | smcginnis | Followed there... that makes sense. | |
| 13:30:21 | mordred | the REST call _only_ accepts ids | |