| *** __ministry is now known as Guest3028 | 01:41 | |
| *** ralonsoh_ is now known as ralonsoh | 05:48 | |
| *** sergio is now known as Guest3042 | 06:21 | |
| opendevreview | Artem Goncharov proposed openstack/codegenerator master: Start building Ironic OpenAPI https://review.opendev.org/c/openstack/codegenerator/+/928664 | 07:16 |
|---|---|---|
| opendevreview | Artem Goncharov proposed openstack/codegenerator master: Escape commit message https://review.opendev.org/c/openstack/codegenerator/+/928761 | 09:00 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Update betamax extra deps https://review.opendev.org/c/openstack/keystoneauth/+/928801 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Add pre-commit https://review.opendev.org/c/openstack/keystoneauth/+/928802 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: docs: Remove cruft from configuration files https://review.opendev.org/c/openstack/keystoneauth/+/928803 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: trivial: Remove errant trailing commas https://review.opendev.org/c/openstack/keystoneauth/+/928804 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Apply ruff, ruff-format https://review.opendev.org/c/openstack/keystoneauth/+/928805 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Enable ruff, ruff-format https://review.opendev.org/c/openstack/keystoneauth/+/928806 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: typing: Make abstract method actually abstract https://review.opendev.org/c/openstack/keystoneauth/+/928807 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: typing: Synchronise overridden method signatures https://review.opendev.org/c/openstack/keystoneauth/+/928808 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: typing: Add variable for auth plugin opts https://review.opendev.org/c/openstack/keystoneauth/+/928809 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: typing: Resolve remaining initial issues https://review.opendev.org/c/openstack/keystoneauth/+/928810 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Enable mypy https://review.opendev.org/c/openstack/keystoneauth/+/928811 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: discover: Expand use of kwargs https://review.opendev.org/c/openstack/keystoneauth/+/928812 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: discover: Rework normalization https://review.opendev.org/c/openstack/keystoneauth/+/928813 | 14:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: typing: Annotate keystoneauth1.discover https://review.opendev.org/c/openstack/keystoneauth/+/928814 | 14:06 |
| stephenfin | gtema: jfyi, I have keystoneauth fully typed (except for tests and test fixtures). Might be looking for your help getting them merged in the next while. I want them to continue my typing SDK journey | 14:19 |
| stephenfin | Also, I was going to ask if you'd seen https://review.opendev.org/c/openstack/codegenerator/+/927704 but it seems you have (y) | 14:19 |
| gtema | sure stephenfin, thanks. I got reformat and typing merged in keystone itself | 14:19 |
| gtema | so adding keystoneauth is great | 14:19 |
| gtema | I am on your change | 14:20 |
| gtema | https://review.opendev.org/c/openstack/codegenerator/+/928761 fixes it (you used ' in the commit message and it broke the job preparing the commit) | 14:20 |
| gtema | sadly there is apparently also one race that I was not able to reproduce locally ever that causes codegenerator to produce "broken" code and only recheck helps - I am scratching my head with no ideas why is that happening and how I can reproduce it reliably | 14:21 |
| gtema | btw stephenfin - we started looking at ironic. They use different decorators and not everything is easy, but at least I hope they will decide how to deal with it (whether to switch to decorators like everywhere else or to tweak codegenerator to support that deviation as well) | 14:23 |
| stephenfin | sweet \o/ | 14:23 |
| stephenfin | I did talk to them at the last PTG about doing this and TheJulia noted that they would just like to ensure adding decorators didn't introduce a significant performance regression but were otherwise open to it | 14:24 |
| stephenfin | I've been focused on Manila and Nova (plus, to a lesser degree, Cinder) since though, so I haven't had a chance to work on Ironic | 14:25 |
| gtema | JayF asked his colleague to start looking at openapi@ironic so hopefully we get a better traction. It's however still not trivial due to use of pecan with a very dynamic routing | 14:25 |
| JayF | adamcarthur5 is in here FYI | 14:26 |
| gtema | right, thanks for bringing us together JayF | 14:27 |
| JayF | Absolutely. We're going to be chatting today and he's going to bring me up to date. | 14:28 |
| gtema | perfect. I told him also that we could have another joint meeting to further align on strategy | 14:29 |
| JayF | I'm more of a person who likes to find a specific action we can take and take it rather than talking more about strategy. Such as maybe putting a point on that question for how ironic is going to implement this, so that we can pilot it and get an answer from the other cores | 14:30 |
| gtema | you speak to my soul JayF ;-) | 14:31 |
| stephenfin | JayF: Agreed. The main thing we need is the ability to retrieve a schema for a given microversion. | 14:32 |
| JayF | I don't know any of the mechanics for what y'all been working on so far, but I'll have a better idea after talking to Adam this afternoon. | 14:33 |
| JayF | I think you'll find that ironic tends to let the implementers have a lot of free reign on implementation details. We're not dogmatic about specific ways of doing things | 14:33 |
| JayF | So if you find a way that works, as long as it makes sense and is updatable I don't think anyone is going to have a problem with it [knock on wood] | 14:34 |
| stephenfin | idk how up-to-speed Adam is, but most of nova's logic lives here nova/api/validation/__init__.py | 14:34 |
| gtema | stephenfin - ironic has something similar and though deviating heavily | 14:35 |
| stephenfin | It's not complicated and relies on a webob request object being the first parameter to any controller method | 14:35 |
| stephenfin | gtema: good to know | 14:35 |
| stephenfin | so does manila and cinder, but those are a lot close to Nova that Ironic is (all use webob + routes) | 14:35 |
| stephenfin | ...hence my focus on those three together so far :) | 14:36 |
| gtema | and I "copied" the same into Keystone (which is similarly as ironic relies on a completely different framework) | 14:36 |
| gtema | from that pov it would be strategically more helpful if ironic would adopt the same decorators (at least syntaxically) | 14:37 |
| stephenfin | yeah, flask (with plugins) iirc. Nova et al probably need to switch themselves at some point since webob seems to be going the way of the dinosaur, but I already have more work that I can keep on top of so... | 14:40 |
| JayF | Ironic uses some webob headers too -- you're saying those need to change? | 14:57 |
| JayF | Let me put it this way; if Ironic has tech debt in our API controllers around how we pass around headers/information, I'm not aware of it and so please help if you know any :D | 14:58 |
| opendevreview | Merged openstack/codegenerator master: Escape commit message https://review.opendev.org/c/openstack/codegenerator/+/928761 | 15:06 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Drop support for Python 3.8 https://review.opendev.org/c/openstack/keystoneauth/+/928833 | 16:45 |
| opendevreview | Merged openstack/codegenerator master: Add getting started guide to README https://review.opendev.org/c/openstack/codegenerator/+/927704 | 16:48 |
| opendevreview | Stephen Finucane proposed openstack/keystoneauth master: Drop support for Python 3.8 https://review.opendev.org/c/openstack/keystoneauth/+/928833 | 16:49 |
| stephenfin | gtema: Are we ready to merge the "drop python3.8" patches now that we've branched? https://review.opendev.org/q/topic:%22drop-py37-py38%22 | 16:50 |
| gtema | yupp, shoot it | 16:51 |
| gtema | oh but not in the cli | 16:52 |
| stephenfin | huh? | 16:52 |
| gtema | cli is not branched, not that I see a branch | 16:52 |
| stephenfin | I think it's independent | 16:52 |
| stephenfin | nope | 16:53 |
| stephenfin | only thing I can see is that it's 'type: other' | 16:54 |
| gtema | ehm, I am confused | 16:55 |
| stephenfin | looks like this has happened before https://review.opendev.org/c/openstack/releases/+/913494 | 16:56 |
| stephenfin | I think there's a tooling issue | 16:56 |
| gtema | fwik it should have happened in https://review.opendev.org/c/openstack/releases/+/927944 where sdk and osc-lib were branched | 17:03 |
| stephenfin | discussing on #openstack-release now. elodilles confirmed it's because of the type | 17:04 |
| stephenfin | osc-lib and sdk are 'type: library' so we can't do the same there. client-library doesn't really make sense either since it's not a library | 17:04 |
| opendevreview | Slawek Kaplonski proposed openstack/python-openstackclient master: Add "trusted" attribut to the "port" https://review.opendev.org/c/openstack/python-openstackclient/+/927727 | 17:49 |
| opendevreview | Slawek Kaplonski proposed openstack/openstacksdk master: Add "trusted" attribute to "port" resource https://review.opendev.org/c/openstack/openstacksdk/+/927723 | 18:25 |
| samcat116 | Has --os-profile been removed from the OpenStack CLI? Its not a recognized argument for anything on client version 6.3.0 | 19:15 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Change block storage volume summary parameter https://review.opendev.org/c/openstack/openstacksdk/+/928867 | 20:29 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Change block storage volume summary parameter https://review.opendev.org/c/openstack/openstacksdk/+/928867 | 20:30 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Change block storage volume summary parameter https://review.opendev.org/c/openstack/openstacksdk/+/928867 | 21:16 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Change block storage volume summary parameter https://review.opendev.org/c/openstack/openstacksdk/+/928867 | 21:26 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Change block storage volume summary parameter https://review.opendev.org/c/openstack/openstacksdk/+/928867 | 21:49 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Add backup summary API https://review.opendev.org/c/openstack/openstacksdk/+/928870 | 21:52 |
| opendevreview | Tobias Urdin proposed openstack/openstacksdk master: Add backup summary API https://review.opendev.org/c/openstack/openstacksdk/+/928870 | 21:52 |
| opendevreview | Tobias Urdin proposed openstack/python-openstackclient master: Add volume backup summary command https://review.opendev.org/c/openstack/python-openstackclient/+/928871 | 22:17 |
| JayF | gtema: took a browse with adamcarthur5 through the generator, and looking at ironic v nova microversion implementations, it seems extremely unlikely that the same approach will work unless significant changes are made to Ironic. Those changes likely will need another catalyst to happen (e.g. if we need to migrate something in our api library stack anyway) and would be a hot topic for PTG | 22:51 |
| JayF | gtema: if you wanna see an example, start here: https://opendev.org/openstack/ironic/src/branch/master/ironic/api/controllers/v1/versions.py look at history and you'll see many commits adding microversions to see how we handle them, e.g.: https://opendev.org/openstack/ironic/commit/8e34d622aff72d7dd286add31e3d7cd366629bc2 | 22:52 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!