| opendevreview | Kaifeng Wang proposed openstack/ironic-python-agent-builder master: Remove unnecessary firmware from deploy image https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/940844 | 02:33 |
|---|---|---|
| cardoe | JayF: https://github.com/mfencik/nova/commit/bf5bd652768ff6393917d15c8167bb2b154ce3c5 | 03:54 |
| cardoe | That's his first swag. | 03:55 |
| cardoe | I've some some feedback on how to make it generic. I'll publish it publicly and you can see if you agree | 03:55 |
| opendevreview | Doug Goldstein proposed openstack/ironic-prometheus-exporter master: fix sphinx errors with incorrect backticks https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940575 | 04:58 |
| opendevreview | Doug Goldstein proposed openstack/ironic-prometheus-exporter master: fix bashate error in script https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940576 | 04:58 |
| opendevreview | Doug Goldstein proposed openstack/ironic-prometheus-exporter master: enable pre-commit lints from main ironic repo https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940577 | 04:58 |
| opendevreview | Doug Goldstein proposed openstack/ironic-prometheus-exporter master: Revert "temp hack for failing flake8 check" https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940588 | 04:58 |
| rpittau | good morning ironic! o/ | 07:57 |
| opendevreview | Kaifeng Wang proposed openstack/ironic-specs master: Tag base port/porgroup scheduling https://review.opendev.org/c/openstack/ironic-specs/+/940861 | 09:16 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Run metal3 integration job using UEFI boot (default) https://review.opendev.org/c/openstack/ironic/+/939694 | 11:02 |
| masghar | re cardoe: wrt to 'not a lot going on at the moment - Taylor Swift' she dropped folklore and evermore after this instagram post during covid, IIRC :D XD | 11:12 |
| masghar | (2 albums) | 11:13 |
| milan | JayF: this is my better formatted version to what cardoe posted - https://github.com/mfencik/nova/commit/6db52b81efeee700a500bbe75ce46661f7e7e0f7 but in general I agree that I could make trunk data more general | 11:49 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 12:14 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 12:42 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-python-agent bugfix/10.1: Update .gitreview for bugfix/10.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/940875 | 12:49 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic bugfix/28.0: Update .gitreview for bugfix/28.0 https://review.opendev.org/c/openstack/ironic/+/940876 | 12:53 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 13:03 |
| opendevreview | cid proposed openstack/ironic-prometheus-exporter master: Trivial: Fix 'node_uuid' KeyError https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940571 | 13:07 |
| opendevreview | cid proposed openstack/ironic master: Log secure boot access failures at INFO level https://review.opendev.org/c/openstack/ironic/+/940433 | 13:18 |
| TheJulia | good morning | 13:20 |
| cid | o/ | 13:21 |
| cid | cardoe: I made a second pass at https://etherpad.opendev.org/p/ironic-cardoe-inspect-hooks-and-rules, no comments and I'm sure that's not a bad idea at all. | 13:21 |
| TheJulia | cid: by chance, did you find the float typeerror issue which was mentioned in IRC ?last week? | 13:23 |
| cid | I'm not sure I remember the convo, ;) | 13:24 |
| JayF | cardoe: that makes zero sense to me. How will it fix vms to edit ironic driver? | 13:24 |
| TheJulia | https://meetings.opendev.org/irclogs/%23openstack-ironic/%23openstack-ironic.2025-01-31.log.html#t2025-01-31T15:49:37 | 13:26 |
| cardoe | JayF: it's not. that's just ironic approach. | 13:31 |
| cardoe | JayF: https://github.com/openstack/nova/blob/d267ede98f1b0aa25630b1efe07db5a8b478d473/nova/network/model.py#L434 is where we'd either need to add a list of nested VIFs (I think that's over kill) or create another smaller object that is just tagged interfaces | 13:32 |
| JayF | I am confused about what you're fixing | 13:32 |
| JayF | I thought your complaint was vlans and bonds didn't work in vms | 13:33 |
| cid | TheJulia, oh I remember. No haven't even tried. That's something I could take a look at some time. And the logs are still available. | 13:37 |
| opendevreview | Kaifeng Wang proposed openstack/ironic master: Remove ibmc leftovers https://review.opendev.org/c/openstack/ironic/+/940877 | 13:47 |
| TheJulia | oh interesting | 13:54 |
| TheJulia | w/r/t that float error | 13:55 |
| opendevreview | Kaifeng Wang proposed openstack/ironic-specs master: Tag based port/porgroup scheduling https://review.opendev.org/c/openstack/ironic-specs/+/940861 | 13:58 |
| opendevreview | Julia Kreger proposed openstack/ironic-prometheus-exporter master: Trivial: Fix KeyError from None value being returned in sensor data https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940878 | 14:09 |
| TheJulia | cid: ^^^ | 14:09 |
| TheJulia | super easy fix | 14:09 |
| TheJulia | err | 14:09 |
| TheJulia | Need to fix the commit message | 14:09 |
| cid | That was fast | 14:09 |
| cid | tks | 14:10 |
| opendevreview | Julia Kreger proposed openstack/ironic-prometheus-exporter master: Trivial: Fix TypeError from None value being returned in sensor data https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940878 | 14:10 |
| TheJulia | I remembered what was going on | 14:10 |
| TheJulia | and actually had to revise it because there is a 0 value test already | 14:10 |
| cardoe | Since all eyes are on ironic-prometheus-exporter... https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940577 that series gets the pre-commit based lints to run in CI | 14:14 |
| cardoe | JayF: in a round about fashion. don't really care about bonds. What I care about are trunk ports (non-access port) for Ironic. | 14:17 |
| cardoe | The easiest way to get this fixed upstream is to care about VMs. | 14:18 |
| cardoe | The only reason ironic has it's own override for the metadata is that the neutron ports don't have the MAC of the hardware. So the metadata is generated and then the MAC is updated. | 14:20 |
| parasitid | hi! i'm getting here to ask if ever someone here remembers a pretty old initiative to propose an alternative to nova to boot nodes as a end customer., more suited to BM usecases. i'm not refferring to metalsmith. in my memory, it was a new api, which wanted to support sutff like adhoc raid config and sutff. i just cant remember the name of this project and can't find it either. maybe it was even just a blueprint. | 14:28 |
| JayF | You're basically describing use of ironic standalone without Nova | 14:32 |
| JayF | Bifrost is a tool we built an ansible to make that easier, but we also have documented how to do this without it. | 14:33 |
| maxh[m] | there is also metal3.io which uses ironic | 14:35 |
| parasitid | JayF: hi. i know of metal3 but to me it was an api next to nova api. and i know of ironic standalone of course. but it's hard to use. i think it was something like metalsmith, but as an api. maybe it was just a blueprint i dreamt of then... | 14:38 |
| parasitid | found it | 14:41 |
| parasitid | https://wiki.openstack.org/wiki/Mogan | 14:41 |
| JayF | That's the first I've ever heard of it. It seems rather redundant | 14:52 |
| JayF | Yeah that page hasn't been edited in 8 years | 14:52 |
| cardoe | So how can I figure out what zuul jobs can be used? | 15:09 |
| cardoe | I think ironic-specs needs to add "openstack-tox-codespell" or something like htat. | 15:10 |
| cardoe | I thought "openstack-tox-linters" was gonna run pre-commit for us which would run codespell | 15:11 |
| JayF | In other repos, we just run code spell as a pre-commit check right? | 15:11 |
| cardoe | Yeah. | 15:11 |
| JayF | Yeah that should work. If it's not something weird is going on and putting a specific tox target for code spell is just going to work around it at best | 15:11 |
| TheJulia | Yeah, Mogan never went beyond the "idea" phase, really. | 15:26 |
| TheJulia | AT least, that is my perception. They were trying to move really fast by copying other project's code and mashing things together | 15:26 |
| rpittau | it does look way more complex than ironic standalone to me :D | 15:28 |
| TheJulia | Yeah, fully in the openstack context | 15:29 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: Configure ironic bugfix 28 https://review.opendev.org/c/openstack/ironic/+/940889 | 15:31 |
| JayF | you can "ironic standalone" in full openstack with no nova too; we integrate as much or as little as folks want | 15:33 |
| JayF | obviously not really user friendly to use ironic that way without some kinda of client help but the bits are there | 15:33 |
| TheJulia | Yeah, realistically the thing we're "missing" is "go make me a vif" | 15:33 |
| TheJulia | (which might be something we should just support.. somehow) | 15:33 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-python-agent master: Configure ipa bugfix 10.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/940890 | 15:34 |
| opendevreview | Merged openstack/ironic-python-agent bugfix/10.1: Update .gitreview for bugfix/10.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/940875 | 15:52 |
| JayF | TheJulia: I think we should have the conversation we've been dodging for a long time at the next ptg re: improving ironic/nova/neutron interactions | 15:53 |
| JayF | TheJulia: and that, if successful, might lead to us being able to do that | 15:53 |
| TheJulia | doing it now would be easy, the unknown is default networks or preferred networks | 15:54 |
| TheJulia | Nova has ?5? ways you can get a VIF | 15:55 |
| -opendevstatus- NOTICE: nominations for the OpenStack PTL and TC positions are now open, for details see https://governance.openstack.org/election/ | 16:08 | |
| opendevreview | Julia Kreger proposed openstack/ironic master: bootc deploy interface - for bootable containers https://review.opendev.org/c/openstack/ironic/+/937897 | 16:20 |
| opendevreview | Riccardo Pittau proposed openstack/ironic-python-agent master: Configure ipa bugfix 10.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/940890 | 16:22 |
| opendevreview | Merged openstack/ironic bugfix/28.0: Update .gitreview for bugfix/28.0 https://review.opendev.org/c/openstack/ironic/+/940876 | 16:25 |
| opendevreview | Julia Kreger proposed openstack/ironic master: OCI container adjacent artifact support https://review.opendev.org/c/openstack/ironic/+/937896 | 16:30 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Automatic zstd detection and decompression... https://review.opendev.org/c/openstack/ironic/+/938904 | 16:31 |
| TheJulia | JayF: your message on https://review.opendev.org/c/openstack/ironic/+/937897/9 has been truncated | 16:32 |
| JayF | merge before the spec lands | 16:32 |
| JayF | is what I was going to say, then I deleted apparently 90% of it | 16:32 |
| JayF | because I saw it needed the OCI code, which I started to review | 16:32 |
| JayF | then you ninja-updated it LOL :P | 16:33 |
| opendevreview | Julia Kreger proposed openstack/ironic master: bootc deploy interface - for bootable containers https://review.opendev.org/c/openstack/ironic/+/937897 | 16:33 |
| TheJulia | yes, very fair | 16:33 |
| TheJulia | JayF: just the text formatting on one of the methods | 16:33 |
| TheJulia | I noticed some silent errors in the docs build | 16:34 |
| JayF | I was in the docs anyway, no real loss :) | 16:34 |
| TheJulia | sorry! | 16:34 |
| TheJulia | :) | 16:34 |
| JayF | yeah, apologize to the person who has probably owed you a review for a week /o\ | 16:35 |
| TheJulia | eh, the demo was kind of key to demonstrate the idea and get some questions into the open | 16:36 |
| TheJulia | I replied to comments on https://review.opendev.org/c/openstack/ironic-python-agent/+/940178 based upon thoughts, I will revise | 16:39 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Run metal3 integration job using UEFI boot (default) https://review.opendev.org/c/openstack/ironic/+/939694 | 16:42 |
| JayF | satoshi: o/ | 17:11 |
| rpittau | good night! o/ | 17:12 |
| opendevreview | Merged openstack/ironic master: Remove ibmc leftovers https://review.opendev.org/c/openstack/ironic/+/940877 | 17:42 |
| JayF | Everyone welcome satoshi -- he'll be working with cid and I over the next few weeks as part of the MLH fellowship. | 17:43 |
| TheJulia | Welcome satoshi! | 17:44 |
| cid | satoshi, o/ | 17:45 |
| satoshi | Glad to be here. Thanks for the warm welcome. | 17:48 |
| JayF | would folks be +1 to making https://github.com/openstack/ironic/blob/69b3ff4052b4f7d02769c1d77a0b2ab441b7fb6d/ironic/drivers/modules/agent_base.py#L869 overridable per node? | 18:20 |
| JayF | it seems like a glaring omission | 18:20 |
| TheJulia | .... I feel like its been discussed in the past because of risks, but maybe that is no longer valid and maybe we need to just revisit it | 18:32 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 18:32 |
| JayF | TheJulia: honestly my thought was "the reason this doesn't exist is probably me, 8 years ago" but the "don't allow cleaning to be meaningfully disabled at a node level" ship sailed already years ago | 18:34 |
| TheJulia | quite possible | 18:34 |
| TheJulia | my memory in this regard is fuzzy | 18:34 |
| JayF | same, but I remember that was the primary plank in my Ironic platform for a while; disabling cleaning was setting up operators for failure | 18:35 |
| TheJulia | yeah | 18:35 |
| JayF | but the real world sometimes needs the ability to do wrong-seeming things | 18:35 |
| TheJulia | And, We also added more granulated access controls | 18:35 |
| JayF | that is a /very/ good point | 18:35 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 18:36 |
| TheJulia | Admin is now truly an admin, users have varying level of even admin-ish access. Its no longer "oh, yeah, I gave admin role to $rando-user because he needs to do a different thing" | 18:38 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 18:41 |
| cardoe | Speaking of steps per node and such. I know there was a plan to add an API to fetch the available / possible steps. | 18:48 |
| cardoe | I also know that the cleaning docs tell you to grab the driver_internal_info and parse out the data to figure out what cleaning step you're on (by extension service steps but that's not covered). | 18:48 |
| cardoe | How would folks feel about a CLI addition to show the current step and maybe have it extensible so it's natural to add the list available steps in the future being added? | 18:49 |
| JayF | So the reason we didn't add the full list of steps | 18:50 |
| JayF | is to get that list, and include in-band steps, requires booting the agent | 18:50 |
| JayF | since available steps are dynamic based on available hardware | 18:50 |
| cardoe | Yep. Not proposing to tackle that piece. But just not making the CLI awkward if something like that does appear. | 18:51 |
| -opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily while we upgrade for a new jeepyb feature and switch our database container image source repository | 18:51 | |
| JayF | CLI for getting current clean step and only that exists too, any addition would be syntax sugar effectively | 18:52 |
| JayF | baremetal node show node-0 -c clean_step -f value | 18:52 |
| JayF | I think? | 18:52 |
| * TheJulia awaits gerrit to return | 18:54 | |
| cardoe | ah so it's not that bad then. | 18:55 |
| TheJulia | cardoe: and, we had people objecting to caching | 18:55 |
| cardoe | https://docs.openstack.org/ironic/latest/admin/cleaning.html#what-cleaning-step-is-running that's the current docs. | 18:55 |
| TheJulia | and folks demanding the api worked even if the node was deployed | 18:55 |
| JayF | that's the old way iirc | 18:55 |
| TheJulia | so requirements which didn't align | 18:55 |
| JayF | yeah there was no way to shape the API that made it 1) useful, 2) not lie and 3) be available without waiting ten minutes for a server to boot | 18:56 |
| cardoe | I was gonna pitch out "baremetal node step show node-0" | 18:56 |
| JayF | cardoe: I'm not crazy am I? node.clean_step exists now, right? | 18:56 |
| * JayF checks | 18:56 | |
| cardoe | Yeah it does. | 18:56 |
| TheJulia | yes, it does | 18:56 |
| JayF | okay yeah | 18:56 |
| JayF | that's just an old doc then | 18:56 |
| cardoe | I've been parsing driver_internal_info like the docs tell me to. :D | 18:56 |
| JayF | if nobody cleans it up soon I'll jfdi fix it | 18:56 |
| TheJulia | driver_internal_info shows you the totality of what it has in the list | 18:57 |
| TheJulia | the field is the current step | 18:57 |
| cardoe | Yep. Just a docs update and I'm off to the races. | 18:58 |
| opendevreview | Julia Kreger proposed openstack/ironic master: OCI container adjacent artifact support https://review.opendev.org/c/openstack/ironic/+/937896 | 19:38 |
| TheJulia | I need to sit down and re-test that with the auth changes, but looks like I've completely excised some of the old locking/caching behavior which wouldn't apply in our context that I didn't get to originally | 19:58 |
| opendevreview | cid proposed openstack/ironic master: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 20:50 |
| cardoe | wrt https://bugs.launchpad.net/ironic/+bug/2040217 is there any reason why we wouldn't want to ask neutron to allow physical_network for VXLAN? | 21:05 |
| cardoe | For me personally I've got my own vxlan type driver for neutron that adds physical_network. | 21:06 |
| TheJulia | I suspect it makes sense to do so | 22:44 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Automatic zstd detection and decompression... https://review.opendev.org/c/openstack/ironic/+/938904 | 22:47 |
| opendevreview | Julia Kreger proposed openstack/ironic master: bootc deploy interface - for bootable containers https://review.opendev.org/c/openstack/ironic/+/937897 | 23:08 |
| TheJulia | okay, I feel a lot better about the oci client change in my last revision | 23:27 |
| cardoe | TheJulia: I'm happy to write up up a use case to share to the ironic-networking group. I word blehed in #-neutron just now. | 23:27 |
| TheJulia | might help. I need to spend some mental energy on the entire topic since I suspect I'm going to need to champion lots of work on networking-generic-switch | 23:34 |
| cardoe | Well I left my comments on Kaifeng's spec PR and on his bug... https://bugs.launchpad.net/ironic/+bug/2040217 | 23:56 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!