| opendevreview | Merged openstack/ironic master: Address conductors by Target.server, not host-in-topic https://review.opendev.org/c/openstack/ironic/+/997789 | 00:22 |
|---|---|---|
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: Prevent container steps from shadowing existing steps https://review.opendev.org/c/openstack/ironic-python-agent/+/998376 | 02:06 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: Prevent container steps from shadowing existing steps https://review.opendev.org/c/openstack/ironic-python-agent/+/998376 | 04:14 |
| opendevreview | Jed Preist proposed openstack/ironic master: Add AgentBaseMixin to AutodetectDeploy https://review.opendev.org/c/openstack/ironic/+/1000074 | 08:59 |
| opendevreview | Yorick proposed openstack/sushy master: system: guard HttpBootUri writes against iDRAC 10 removal https://review.opendev.org/c/openstack/sushy/+/999743 | 11:23 |
| opendevreview | Esther Domfeh proposed openstack/ironic master: Add state, target_provision_state, and duration_seconds to node history https://review.opendev.org/c/openstack/ironic/+/989994 | 11:49 |
| opendevreview | Esther Domfeh proposed openstack/ironic master: Populate node_history state fields https://review.opendev.org/c/openstack/ironic/+/993208 | 11:49 |
| opendevreview | Esther Domfeh proposed openstack/ironic master: Add request context user tracking to node history records https://review.opendev.org/c/openstack/ironic/+/993901 | 11:49 |
| opendevreview | Esther Domfeh proposed openstack/ironic master: Add power tracking telemetry in node history https://review.opendev.org/c/openstack/ironic/+/996347 | 11:49 |
| dtantsur | JayF: https://review.opendev.org/c/openstack/ironic/+/997998/comment/a2d0bcff_fc4431ce/. I can move this sentence to the future metal3 document, but I think we crossed the bridge long ago. | 12:35 |
| dtantsur | This statement won't even make it to top 10 questionable things in the Ironic docs that we claim to be supported | 12:35 |
| JayF | Well I'm glad to make sure it's a top 19 and not a top 20 then 😂 | 13:54 |
| dtantsur | All I'm saying is that "remove a provisioned host from the provisioner" is more or less on the default path for the use case of "ironic as a general provisioner". | 13:58 |
| dtantsur | The fact that it has a weird API that involved maintainance is a self-inflicted pain | 13:58 |
| dtantsur | "That's not how Nova works" (tm) | 13:59 |
| TheJulia | dtantsur: I feel like there is a t-shirt option somewhere in this spectrum | 14:13 |
| JayF | dtantsur: I will put it this way: if Ironic has somehow garnered the additional scope of being a persistent oneshot deployment mechanism, it happened without me noticing. Oneshot deployment? A-OK. Long running Ironic doing life-cycle management? A-OK. Spinning nodes in and out of Ironic just to provision them ... literally that patch is the first I've heard of that pattern | 14:24 |
| JayF | "I spun up a temporary Ironic to provision 200 nodes" != "I have a persistent Ironic running that we enroll/unenroll nodes from to provision them" | 14:24 |
| JayF | I'm not even saying it's wrong, just that if that somehow became a general use case it did without me having any awareness of it | 14:25 |
| TheJulia | It is definitely a pattern which has been embraced although there are always folks who want one or both of the worlds | 14:48 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: Prevent container steps from shadowing existing steps https://review.opendev.org/c/openstack/ironic-python-agent/+/998376 | 14:53 |
| opendevreview | Doug Goldstein proposed openstack/python-ironicclient master: Add --disable-ramdisk to runbook create https://review.opendev.org/c/openstack/python-ironicclient/+/1000110 | 15:16 |
| opendevreview | Doug Goldstein proposed openstack/python-ironicclient master: Remove broken --traits option from runbook create https://review.opendev.org/c/openstack/python-ironicclient/+/1000111 | 15:16 |
| opendevreview | Doug Goldstein proposed openstack/python-ironicclient master: Remove broken --step option from runbook unset https://review.opendev.org/c/openstack/python-ironicclient/+/1000112 | 15:16 |
| opendevreview | Doug Goldstein proposed openstack/python-ironicclient master: Add --disable-ramdisk to runbook set/unset https://review.opendev.org/c/openstack/python-ironicclient/+/1000113 | 15:16 |
| opendevreview | Merged openstack/sushy master: Handle incomplete TaskMonitor responses https://review.opendev.org/c/openstack/sushy/+/999741 | 15:57 |
| TheJulia | If I can get some ironic folks to look at https://review.opendev.org/c/openstack/security-doc/+/1000120 it would be appreciated | 16:05 |
| opendevreview | Merged openstack/ironic master: Make Redfish read timeout configurable https://review.opendev.org/c/openstack/ironic/+/998711 | 16:15 |
| TheJulia | One item regarding the security note, we may be lacking a more explicit override for inspection url generation | 16:17 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.openstack.org will be offline momentarily at 18:00 UTC for a patch upgrade (just under an hour from now), but should return within a few minutes. | 17:11 | |
| opendevreview | Merged openstack/ironic-python-agent master: Prevent container steps from shadowing existing steps https://review.opendev.org/c/openstack/ironic-python-agent/+/998376 | 17:20 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Ensure AutodetectDeploy always cleans https://review.opendev.org/c/openstack/ironic/+/1000074 | 17:21 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Ensure AutodetectDeploy always cleans https://review.opendev.org/c/openstack/ironic/+/1000074 | 17:54 |
| JayF | TheJulia: ^ althoguh I'm reviewing it for claudeisms now and will see how CI likes it | 17:55 |
| TheJulia | Yeah, I'm definitely curious | 17:59 |
| JayF | I think this'll work | 17:59 |
| JayF | I put a review on it, lots of cleanup stuff, typical claude BS, but the primary logic looks good | 18:00 |
| TheJulia | yeah | 18:00 |
| JayF | will see what CI thinks first, then make sure we aren't missing anywhere that needs a switch_interface() and cleanup the various claudeisms | 18:00 |
| JayF | do you know how to point that github/metal3 change to my updated code here so I can see if it's happier? | 18:00 |
| TheJulia | The change on our repo is supposed to do that, that being said I think there was a question w/r/t the newer job which was recently added | 18:02 |
| JayF | I think they made a change to the job on the github side to make it fail the initial clean | 18:02 |
| JayF | https://github.com/metal3-io/baremetal-operator/pull/3495 | 18:02 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org is going offline momentarily for a patch upgrade, but should return within a few minutes. | 18:02 | |
| JayF | dtantsur: rpittau: How do I point CI for ^ at my change in gerrit? Or can I at all? | 18:03 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Only honor container step priorities while cleaning https://review.opendev.org/c/openstack/ironic-python-agent/+/998377 | 18:21 |
| TheJulia | I rebased ^, I'm not sure why it didn't like being 3 deep, so time will tell | 18:25 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Accept whitespace separated container options https://review.opendev.org/c/openstack/ironic-python-agent/+/998378 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Check that the configured container runtime is present https://review.opendev.org/c/openstack/ironic-python-agent/+/998379 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Report errors reading the container steps file https://review.opendev.org/c/openstack/ironic-python-agent/+/998380 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Ignore the transport prefix when matching allowed containers https://review.opendev.org/c/openstack/ironic-python-agent/+/998381 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Drop the unused container_conf_file option https://review.opendev.org/c/openstack/ironic-python-agent/+/998382 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Rename container_clean_step to generic_container_step https://review.opendev.org/c/openstack/ironic-python-agent/+/998383 | 18:26 |
| opendevreview | cid proposed openstack/ironic-python-agent master: Verify registry TLS by default https://review.opendev.org/c/openstack/ironic-python-agent/+/999725 | 18:26 |
| TheJulia | rebase button party | 18:27 |
| cid | \o/. tks | 18:53 |
| TheJulia | happy to | 19:12 |
| TheJulia | looks like CI is pretty happy, at least on ipa | 19:19 |
| TheJulia | the failures related to openstacksdk are annoying | 19:19 |
| TheJulia | Well, if CI stays happy, a lot will land over the next few hours | 19:48 |
| opendevreview | Jay Faulkner proposed openstack/ironic-specs master: Add Portgroup/LAG support to ironic-networking https://review.opendev.org/c/openstack/ironic-specs/+/999448 | 20:14 |
| TheJulia | cardoe: so, I had someone raise an interesting prospect, the idea of clustering baremetal machines with rdma (roce v2) with VMs.. has anyone raised this in your sphere? | 20:21 |
| TheJulia | I would suspect the ha chasiss group would never get assigned, so it might get "weird" in ovn... | 20:22 |
| TheJulia | That is as long as there i no router and all, and the question would also be if ovn is handling the packets with same prioritzation, etc. | 20:23 |
| cardoe | Not yet. | 20:24 |
| cardoe | But somehow whenever you ask me about something.... the same thing comes up sometime. | 20:25 |
| cardoe | https://usercontent.irccloud-cdn.com/file/QOY0KSZv/image.png | 20:26 |
| cardoe | mmmm so tasty | 20:26 |
| TheJulia | Is that claude pondering the same? | 20:26 |
| TheJulia | :) | 20:26 |
| cardoe | So that's my it's Friday and I'm about to walk away prompt. | 20:27 |
| TheJulia | heh | 20:27 |
| cardoe | I'll give ya the TL;DR of the prompt. | 20:27 |
| cardoe | The Ironic team has python-ironicclient which is a CLI client but is also the python API. The combined openstack project has openstacksdk which is also the python API. This means the Ironic team has to maintain both places. Let's come up with a plan to migrate python-ironicclient to utilize openstacksdk while not breaking the python-ironicclient and its APIs but planning for eventual deprecation of the python API. | 20:29 |
| JayF | I would be -1.5 to the output of that. | 20:29 |
| JayF | We'd have to be ridiculously careful to ensure we didn't break `baremetal` (standalone) CLI use | 20:29 |
| cardoe | Yeah of course. | 20:29 |
| JayF | and frankly I am not sure I want to trust our standalone client to an SDK that's integrated-oriented | 20:29 |
| TheJulia | well, plan, it might be doing... knowing claude's desire and weighting to do | 20:29 |
| cardoe | The baremetal use case already depends on the openstacksdk | 20:29 |
| cardoe | It's just writing a plan and analyzing stuff. | 20:30 |
| TheJulia | Yeah, it does make sense to clean some of that | 20:30 |
| TheJulia | I look forward to reading it | 20:30 |
| cardoe | Cause I'm sitting here throwing patches at python-ironicclient and openstacksdk | 20:30 |
| TheJulia | in the mean time, I'm going to head to a movie theater with the wife | 20:30 |
| cardoe | Cause BOTH of them implement microversions v1.100 and up incorrectly. | 20:30 |
| cardoe | It's just tiring to have to use one for 1 API call and the other for the other API call. | 20:31 |
| cardoe | Cause folks only test one or the other. | 20:31 |
| TheJulia | ugh | 20:32 |
| cardoe | JayF: we already use the same "Proxy" interface objects. | 20:32 |
| cardoe | We just copied a version from a much older commit and didn't stay in sync. | 20:33 |
| cardoe | Per commit history it was to avoid having python-ironicclient depend on openstacksdk | 20:34 |
| cardoe | https://opendev.org/openstack/python-ironicclient/src/commit/7721ba3744e3cc85a727a893ecb7e9ef3b42d592/requirements.txt#L11 | 20:34 |
| cardoe | If ya don't wanna click the link... that's python-ironicclient hard depending on openstacksdk | 20:35 |
| cardoe | It's already figured out what commit the Manager object in python-ironicclient was copied from the openstacksdk source tree | 20:38 |
| JayF | yeah that sounds to me like we were trying to espouse the attitude I had earlier (don't require standalone users to install openstacksdk) and somehow it just got ... droppped | 20:40 |
| JayF | that tends to happen when stuff like that is expressed in random IRC conversations and not in docs :P | 20:40 |
| cardoe | Claude is writing a spec that I'll read before I push | 20:44 |
| cardoe | So I mean how far are we gonna take that? Should we do our own keystone auth? | 20:44 |
| JayF | keystoneauth middleware is only executed by integrated openstack installations | 20:46 |
| JayF | my main thing is I don't want us repurposing openstacksdk for standalone use | 20:46 |
| JayF | without their enthusiastic consent and approval for use in standalone cases | 20:46 |
| cardoe | I mean the first commit by dtantsur which adds "baremetal" is hard depending on openstacksdk | 20:46 |
| JayF | if sdk team is on board, I'm meh about all the other stuff | 20:46 |
| cardoe | And in fact it uses osc-lib | 20:48 |
| opendevreview | Merged openstack/ironic-python-agent master: Only honor container step priorities while cleaning https://review.opendev.org/c/openstack/ironic-python-agent/+/998377 | 20:49 |
| cardoe | So "baremetal" is just OSC. We're just registering the "baremetal" subcommands. And then stripping "baremetal" from the prefix chain of the subcommands. | 20:50 |
| JayF | so then what are you proposing? | 20:51 |
| JayF | we use openstacksdk, but have vestigages of "no, don't use it" with copied around code? | 20:51 |
| cardoe | Well I'm just tired of somethings work with python-ironicclient and somethings with openstacksdk... speaking from a... I'm writing a Python script and importing the API and using the API. | 20:52 |
| cardoe | So exploring how we can improve that. | 20:54 |
| opendevreview | Merged openstack/ironic-python-agent master: Accept whitespace separated container options https://review.opendev.org/c/openstack/ironic-python-agent/+/998378 | 20:56 |
| cardoe | The entire underside of the requests usage is copy/paste. So like the object we have that say Node inherits from. It gets a .get() .list() .delete() etc automatically. | 20:59 |
| cardoe | I don't necessarily have an answer to reducing the API split but reducing the duplication seems good. | 21:01 |
| cardoe | You also said that keystoneauth is only executed in openstack installations but we always build up a CloudRegion object | 21:03 |
| cardoe | keystoneauth1.loading._plugins.noauth:NoAuth is what we load in that case | 21:05 |
| JayF | interesting | 21:25 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Ensure AutodetectDeploy always cleans https://review.opendev.org/c/openstack/ironic/+/1000074 | 21:52 |
| opendevreview | Merged openstack/ironic-python-agent master: Check that the configured container runtime is present https://review.opendev.org/c/openstack/ironic-python-agent/+/998379 | 21:57 |
| opendevreview | Merged openstack/ironic-python-agent master: Report errors reading the container steps file https://review.opendev.org/c/openstack/ironic-python-agent/+/998380 | 21:57 |
| opendevreview | Merged openstack/ironic-python-agent master: Ignore the transport prefix when matching allowed containers https://review.opendev.org/c/openstack/ironic-python-agent/+/998381 | 21:57 |
| opendevreview | Merged openstack/ironic-python-agent master: Drop the unused container_conf_file option https://review.opendev.org/c/openstack/ironic-python-agent/+/998382 | 21:57 |
| opendevreview | Merged openstack/ironic-python-agent master: Rename container_clean_step to generic_container_step https://review.opendev.org/c/openstack/ironic-python-agent/+/998383 | 21:57 |
| opendevreview | Merged openstack/ironic-python-agent master: Verify registry TLS by default https://review.opendev.org/c/openstack/ironic-python-agent/+/999725 | 21:57 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
