| JayF | anyone running a multitenant cloud should look at this: https://github.com/google/security-research/security/advisories/GHSA-4xq7-4mgh-gp6w tl;dr AMD microcode issues for any Zen processors. | 00:01 |
|---|---|---|
| *** diablo_rojo_phone is now known as Guest7964 | 00:30 | |
| opendevreview | Steve Baker proposed openstack/ironic-specs master: Graphical Console Support https://review.opendev.org/c/openstack/ironic-specs/+/938526 | 04:11 |
| opendevreview | Merged openstack/ironic master: Trivial: Clean up residual `.vbmc` dir after clean https://review.opendev.org/c/openstack/ironic/+/940573 | 04:49 |
| frickler | codespell is failing on this backport on unrelated issues, how did that even get activated in the first place? https://zuul.opendev.org/t/openstack/build/7a61050406834c298c06fcb5c0fa9c7e | 07:16 |
| opendevreview | cid proposed openstack/ironic master: Trivial: Remove the long deprecated `token` arg https://review.opendev.org/c/openstack/ironic/+/940567 | 07:19 |
| rpittau | good morning ironic! o/ | 07:53 |
| opendevreview | cid proposed openstack/ironic-prometheus-exporter master: Trivial: Fix 'node_uuid' KeyError https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940571 | 08:07 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 08:11 |
| opendevreview | Vasyl Saienko proposed openstack/ironic master: Test patch to check portgroups https://review.opendev.org/c/openstack/ironic/+/940612 | 08:12 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 09:18 |
| opendevreview | cid proposed openstack/ironic-prometheus-exporter master: Trivial: Fix 'node_uuid' KeyError https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940571 | 10:13 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 10:39 |
| opendevreview | Riccardo Pittau proposed openstack/ironic master: [WIP] Run metal3 integration job using UEFI boot (default) https://review.opendev.org/c/openstack/ironic/+/939694 | 10:59 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 11:00 |
| opendevreview | Merged openstack/ironic master: Support querying node history with sort_key and sort_dir https://review.opendev.org/c/openstack/ironic/+/940522 | 11:12 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 11:29 |
| opendevreview | Vasyl Saienko proposed openstack/ironic-tempest-plugin master: Validate network data for portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 | 12:47 |
| opendevreview | Vasyl Saienko proposed openstack/networking-generic-switch master: Add vlan aware VMs support https://review.opendev.org/c/openstack/networking-generic-switch/+/928490 | 13:45 |
| opendevreview | cid proposed openstack/ironic master: [WIP] Agentless Kea DHCP backend. https://review.opendev.org/c/openstack/ironic/+/935645 | 13:52 |
| TheJulia | good morning | 14:08 |
| opendevreview | Vasyl Saienko proposed openstack/ironic master: [devstack] Allow deploy environment with portgroups https://review.opendev.org/c/openstack/ironic/+/940611 | 14:39 |
| opendevreview | Vasyl Saienko proposed openstack/ironic master: Test patch to check portgroups https://review.opendev.org/c/openstack/ironic/+/940612 | 14:39 |
| cardoe | cid: ironic-prometheus-exporter master fails today. | 14:52 |
| cardoe | cid: https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940587 is the hack to make it pass tests | 14:52 |
| cardoe | But the whole series fixes up the linting stack to match what we have in ironic | 14:52 |
| cardoe | And the last commit reverts that hack. | 14:52 |
| cardoe | Since I see you're fighting with a fix there. | 14:53 |
| cid | I am :D | 14:54 |
| cid | I think I can also add a #noqa to the last line there as well right? | 14:54 |
| cardoe | So if ya wanna land my hack first ;) | 14:54 |
| cid | That works too | 14:54 |
| TheJulia | I can look at that stuff in a little over an hour | 14:55 |
| rpittau | cid cardoe I've approved that to unblock CI | 14:56 |
| cid | tks ;) | 14:56 |
| cardoe | Thanks | 14:56 |
| cardoe | frickler: yeah the codespell thing is weird. my only guess is that it floated forward version wise? | 14:57 |
| JayF | cardoe: frickler: you know, considering, we should probably *completely disable* codespell at the stable release boundary | 15:01 |
| cardoe | I'd be cool with that. | 15:11 |
| dtantsur | cardoe: hey, good morning! Do you have any sort of a write-up on where you want to end up with your effort to generalize inspection features? (formal or informal) | 15:13 |
| dtantsur | JayF++ | 15:17 |
| frickler | JayF: sounds reasonable, try to keep newly added code clean, don't care about old mistakes ;) | 15:27 |
| JayF | I'm happy to land a change if someone writes it, I'm in meetings all morning though | 15:28 |
| frickler | I would do it, but I fail to find where ironic-tox-codespell is even defined | 15:33 |
| JayF | ~/.pre-commit-config.yaml | 15:34 |
| JayF | or somethign like that | 15:34 |
| JayF | well, not home, but root of repo | 15:34 |
| JayF | we run all lint thru pre-commit now | 15:34 |
| cardoe | Can we not just turn off that step running | 15:37 |
| JayF | lint is part of the overall template for openstack tox jobs | 15:37 |
| JayF | I think you'd have to list it and mark it nonvoting | 15:37 |
| opendevreview | Merged openstack/ironic-prometheus-exporter master: temp hack for failing flake8 check https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940587 | 15:47 |
| opendevreview | cid proposed openstack/ironic-prometheus-exporter master: Trivial: Fix 'node_uuid' KeyError https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/940571 | 15:48 |
| frickler | ah, https://review.opendev.org/c/openstack/ironic/+/933686 was done in master only, 2024.2 has still the previous explicit jobs. so I'll do a stable-only patch to drop the job there | 15:50 |
| frickler | and then maybe someone remembers to do further cleanup when 2025.1 is branched ... :) | 15:51 |
| opendevreview | Dr. Jens Harbott proposed openstack/ironic stable/2024.2: stable-only: Drop ironic-tox-codespell job https://review.opendev.org/c/openstack/ironic/+/940717 | 15:53 |
| frickler | JayF: ^^ | 15:53 |
| JayF | +2A (we only require one core review for CI unblocking) | 15:54 |
| opendevreview | Dr. Jens Harbott proposed openstack/ironic stable/2024.1: stable-only: Drop ironic-tox-codespell job https://review.opendev.org/c/openstack/ironic/+/940718 | 15:59 |
| frickler | thx. ^^ seems to be the only other branch that would be affected | 15:59 |
| opendevreview | Vasyl Saienko proposed openstack/ironic master: [devstack] Allow deploy environment with portgroups https://review.opendev.org/c/openstack/ironic/+/940611 | 16:15 |
| opendevreview | Vasyl Saienko proposed openstack/ironic master: Test patch to check portgroups https://review.opendev.org/c/openstack/ironic/+/940612 | 16:27 |
| vsaienko | JayF: please add to your review queue https://review.opendev.org/c/openstack/networking-generic-switch/+/940606 fixes port validation of ngs. | 16:40 |
| opendevreview | cid proposed openstack/ironic master: WIP: make port binding fatel https://review.opendev.org/c/openstack/ironic/+/699953 | 16:45 |
| *** awb_ is now known as awb | 16:46 | |
| opendevreview | cid proposed openstack/ironic master: WIP: Make port binding (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 16:46 |
| opendevreview | cid proposed openstack/ironic master: WIP: Make port binding failure (configurably) fatal https://review.opendev.org/c/openstack/ironic/+/699953 | 16:47 |
| JayF | vsaienko: I think there's something I don't understand about that change, because it looks like you remove more code than you replace | 16:48 |
| vsaienko | the same code is in update_port_postcommit, its not needed in bind_port method | 16:49 |
| vsaienko | I've left only code that do if driver can handle the request in bind_port. Actual logic of switch configuration is triggered later in port_update_postcommit. Please check docstring for bind_port there is a description why it should not be done there | 16:50 |
| vsaienko | please review 1 line fix for ngs that removes unneeded hasattr check https://review.opendev.org/c/openstack/networking-generic-switch/+/940296 | 16:57 |
| JayF | vsaienko: oh wow, that's almost certainly a badly forward port of a downstream patch on my part | 17:04 |
| JayF | vsaienko: thanks for pointing me at it, +2 | 17:04 |
| JayF | marking it for backport, too | 17:05 |
| JayF | +2 the hasattr removal | 17:06 |
| JayF | someone else should review and land 940606 ^^^ | 17:06 |
| vsaienko | thanks | 17:08 |
| rpittau | good night! o/ | 17:15 |
| vsaienko | TheJulia: I've added tests to cover network_data with portgroups https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678, please add to your review queue | 17:16 |
| vsaienko | https://review.opendev.org/c/openstack/ironic/+/940584/3 | 17:16 |
| vsaienko | https://review.opendev.org/c/openstack/ironic/+/940611/5 | 17:17 |
| vsaienko | integration patch that combines tempest plugin change and enables portgroups is here https://review.opendev.org/c/openstack/ironic/+/940612. We need to decide on which job we want to enable portgroups. Also it would be good if we can add job with portgroups into n-g-s as well. | 17:18 |
| cardoe | JayF: If you want to +W https://review.opendev.org/c/openstack/networking-generic-switch/+/940606 I +2'd it. | 18:05 |
| JayF | +a | 18:05 |
| cardoe | https://review.opendev.org/c/openstack/networking-generic-switch/+/940296 too | 18:06 |
| cardoe | dtantsur: https://etherpad.opendev.org/p/ironic-cardoe-inspect-hooks-and-rules and https://gist.github.com/cardoe/82ca7d2a33ae963dbe30f2b43ac13f64 should hopefully be rough. I wanna get my example more concrete. | 18:24 |
| opendevreview | Takashi Kajinami proposed openstack/ironic-inspector master: Fix missing ironic-inspector-status when GLOBAL_VENV is True https://review.opendev.org/c/openstack/ironic-inspector/+/935363 | 18:39 |
| *** iurygregory_ is now known as iurygregory | 18:55 | |
| opendevreview | Merged openstack/ironic stable/2024.2: stable-only: Drop ironic-tox-codespell job https://review.opendev.org/c/openstack/ironic/+/940717 | 19:04 |
| opendevreview | Merged openstack/ironic stable/2024.1: stable-only: Drop ironic-tox-codespell job https://review.opendev.org/c/openstack/ironic/+/940718 | 19:08 |
| opendevreview | Jay Faulkner proposed openstack/ironic-inspector master: Migrate from ironic-lib https://review.opendev.org/c/openstack/ironic-inspector/+/939407 | 19:23 |
| opendevreview | Doug Goldstein proposed openstack/ironic stable/2024.2: Fix hold/wait step logic in step validation https://review.opendev.org/c/openstack/ironic/+/940637 | 19:34 |
| opendevreview | Merged openstack/networking-generic-switch master: Fix ngs_allowed_vlans validation https://review.opendev.org/c/openstack/networking-generic-switch/+/940606 | 20:00 |
| opendevreview | Verification of a change to openstack/networking-generic-switch master failed: Remove unneeded hasattr check for bond methods https://review.opendev.org/c/openstack/networking-generic-switch/+/940296 | 20:00 |
| opendevreview | Jay Faulkner proposed openstack/ironic master: Fix invalid command in grenade destroy https://review.opendev.org/c/openstack/ironic/+/940745 | 20:39 |
| opendevreview | Jay Faulkner proposed openstack/ironic-inspector master: Fix missing ironic-inspector-status when GLOBAL_VENV is True https://review.opendev.org/c/openstack/ironic-inspector/+/935363 | 20:39 |
| JayF | well I found /something/ broken, who the hell knows if it's the thing failing inspector runs | 20:40 |
| * JayF crosses fingers | 20:40 | |
| cardoe | So stupid idea... Been asked to deploy an ISO based OS. The ISO supplied will install everything. I'm however using nova-ironic which doesn't allow for this. I could in theory have a "deploy_iso" step that uses boot_iso and then waits for the box to reboot and be booted off the local disk. That state identification should be possible via redfish. | 22:20 |
| cardoe | Cause today the ISO based boot method assumes you'll always be booting from an ISO. There's no logic that says this boot_iso is an installer. | 22:21 |
| JayF | cardoe: I've thought for a long, long time that if someone had the gumption, they could use the anaconda deploy interface as a template for an "installer" deploy interface | 22:21 |
| cardoe | hrm. So I could potentially enable nova to use that. | 22:23 |
| cardoe | Cause that's essentially what they're doing. They're rolling kickstart stuff into an ISO. | 22:23 |
| JayF | I've used the anaconda deploy through nova before* | 22:24 |
| cardoe | It's VMware ESXi. | 22:24 |
| JayF | the * is because it was the *downstream* impleemntation that I later helped upstream | 22:24 |
| cardoe | hrm I don't see nova/virt/ironic/patcher.py setting the necessary bits. | 22:25 |
| JayF | What bits do you refer to? | 22:26 |
| JayF | https://docs.openstack.org/ironic/latest/admin/anaconda-deploy-interface.html#creating-an-os-image | 22:27 |
| JayF | the integrated limitation is a 1:1:1 mapping between kickstart image/kickstart template file/glance image object | 22:30 |
| JayF | so you can't allow arbitrary users to run arbitrary ks scripts | 22:30 |
| cardoe | ah now I follow | 22:31 |
| cardoe | openstack baremetal node set <node> --instance_info ks_template=glance://uuid from the docs however which seems to keep it separate from the image. | 22:31 |
| cardoe | And that's the part I don't see being set in nova/virt/ironic/patcher.py | 22:32 |
| JayF | the text right below "creating a bare metal server" cleans that up | 22:33 |
| JayF | basically you use the kickstart template associated with the glance image for non-standalone | 22:33 |
| JayF | that is an instance_info level override allowing per-node changes /if not doing boots in nova/ | 22:34 |
| cardoe | Oh... | 22:34 |
| cardoe | "Apart from uploading a custom kickstart template to glance and associating it with the OS image via the ks_template property in glance" | 22:34 |
| cardoe | Reading comprehension. | 22:34 |
| JayF | of course, if one has a patched downstream nova and patched downstream nova clients, you could take an unofficial arg at that layer and pass it through | 22:35 |
| JayF | that might lead you to shout Yahoo! about the freedom given to your users <.< >.> | 22:35 |
| cardoe | Well I'd really only wanna dynamically patch in instance_metadata | 22:35 |
| cardoe | Like NICs and such. | 22:35 |
| JayF | https://review.opendev.org/c/openstack/ironic/+/940745 is a trivial patch which can land; sadly did not fix inspector grenade | 22:47 |
| JayF | I'm slowly eliminating errors seen in the log while the tests themselves cackle and fail in the same manner lolsob | 22:47 |
| opendevreview | Steve Baker proposed openstack/ironic master: Checkout dnsmasq in a tmp directory https://review.opendev.org/c/openstack/ironic/+/940754 | 22:57 |
| opendevreview | Steve Baker proposed openstack/ironic master: Utility functions for graphical console drivers https://review.opendev.org/c/openstack/ironic/+/939505 | 22:59 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add ironic-novncproxy service https://review.opendev.org/c/openstack/ironic/+/939191 | 22:59 |
| JayF | stevebaker[m]: it's possible we don't even need to compile a dnsmasq anymore | 23:04 |
| * JayF checks | 23:04 | |
| JayF | we're using noble, right? | 23:04 |
| stevebaker[m] | I'm using noble, and that installs version 2.90 | 23:06 |
| JayF | I'm looking to see if the build2 of that has the needed patch | 23:06 |
| JayF | 2.91 is still in testing upstream | 23:06 |
| JayF | they like to test them for months, apparently :-| | 23:06 |
| stevebaker[m] | can't be too careful | 23:06 |
| JayF | (yes you can) | 23:07 |
| JayF | lol | 23:07 |
| stevebaker[m] | oh I know :P | 23:07 |
| JayF | how wonderful, the only ubuntu patch in that dnsmasq version is doing the important work of removing the upstream author's donation request | 23:08 |
| JayF | nice to see they're patching the things that matter while not having the patch for a simple crasher bug ಠ_ಠ | 23:08 |
| JayF | and from the ubuntu bug it looks like they passed the buck to stable maint team which just ... isn't doing anything | 23:09 |
| JayF | +2 your change | 23:10 |
| cardoe | stevebaker[m]: we can probably land https://review.opendev.org/c/openstack/ironic/+/939505 to minimize your rebasing in the future. | 23:11 |
| stevebaker[m] | cardoe: that would be nice thanks. The change to add novncproxy is ready now I think. I have just -2ed adding the idrac driver because TheJulia has convinced me it should be treated the same as iLO and Supermicro, with a vnc container | 23:13 |
| cardoe | stevebaker[m]: btw I'm trying to merge sushy-oem-idrac into sushy proper. | 23:15 |
| cardoe | https://review.opendev.org/c/openstack/sushy/+/940557 is the change (it's not ready yet). I'm like 12 yaks deep unfortunately. The tests in sushy are broken and the act of making another oem extension available exposed the brokenness. | 23:16 |
| stevebaker[m] | cardoe: Oh! | 23:16 |
| stevebaker[m] | cardoe: does this theoretically enable other oem extensions too? | 23:17 |
| cardoe | My ultimate goal is to propose something of a "sushy.quirks.<oem>" in the tree where we can gather up some quirks / workarounds. | 23:18 |
| cardoe | Yeah I'll pull in the iLO needful as well. | 23:18 |
| stevebaker[m] | neat | 23:18 |
| cardoe | The quirks would be an auto-load if that oem is exposed. | 23:19 |
| stevebaker[m] | yeah cool | 23:23 |
| cardoe | Dell's just the "biggest" quirk so we have the "idrac" driver. But it's really just redfish. | 23:23 |
| cardoe | For other vendors we've got a pile of if's in the redfish driver. | 23:23 |
| JayF | ilo exposes a lot of oem-specific stuff too | 23:24 |
| cardoe | Yep. | 23:24 |
| JayF | that can change based on what you get with the device in question | 23:24 |
| cardoe | Well the Dell support we've got sucks. | 23:24 |
| cardoe | The redfish driver doesn't do the "if vendor" like the others. You gotta use idrac. Which is fine. But even that doesn't work correctly unless you explicitly install sushy-oem-idrac. | 23:25 |
| cardoe | And some of the requirements for installing that are as lame as just adding another constant to a device status. | 23:25 |
| cardoe | iLO has the same thing in the proliant-utils (I forget exactly the name of the package) for RAID. | 23:26 |
| cardoe | https://opendev.org/x/proliantutils/src/branch/master/proliantutils/redfish/resources that's the oem bits for iLO | 23:29 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/2024.2: Fix redfish session cache on missing password https://review.opendev.org/c/openstack/ironic/+/940634 | 23:32 |
| opendevreview | Jay Faulkner proposed openstack/ironic stable/2024.2: Fix agent from being locked out with complex steps https://review.opendev.org/c/openstack/ironic/+/940471 | 23:33 |
| cardoe | I wiped the +W on those two with the rebase. The rebase was necessary because the codespell failure on stable which has since been fixed. I think JayF and stevebaker[m] had them respectively. | 23:34 |
| stevebaker[m] | Done | 23:37 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!