| opendevreview | OpenStack Release Bot proposed openstack/bifrost stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/bifrost/+/1006453 | 08:39 |
|---|---|---|
| opendevreview | OpenStack Release Bot proposed openstack/bifrost stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/bifrost/+/1006454 | 08:39 |
| opendevreview | OpenStack Release Bot proposed openstack/bifrost master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/bifrost/+/1006455 | 08:39 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-prometheus-exporter stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006456 | 08:39 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-prometheus-exporter stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006457 | 08:39 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-prometheus-exporter master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006458 | 08:39 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-python-agent stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006459 | 08:40 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-python-agent stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006460 | 08:40 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-python-agent master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006461 | 08:40 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006462 | 08:40 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006463 | 08:40 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006464 | 08:40 |
| opendevreview | Merged openstack/ironic-prometheus-exporter stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006456 | 09:44 |
| opendevreview | Merged openstack/bifrost master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/bifrost/+/1006455 | 09:44 |
| opendevreview | Merged openstack/ironic-python-agent stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006459 | 09:45 |
| opendevreview | Merged openstack/ironic-python-agent stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006460 | 09:47 |
| opendevreview | Merged openstack/ironic-python-agent master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic-python-agent/+/1006461 | 09:47 |
| opendevreview | Merged openstack/ironic-prometheus-exporter stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006457 | 09:51 |
| opendevreview | Merged openstack/ironic-prometheus-exporter master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/1006458 | 09:57 |
| opendevreview | Merged openstack/ironic master: Update master for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006464 | 10:00 |
| zigo | Hi there ! | 10:45 |
| zigo | I'm trying to backport https://review.opendev.org/c/openstack/ironic/+/1003598/1/ironic/common/image_service.py#1119 to Dalmatian. The last hunk of this file doesn't apply, as the function get_image_service_auth_override() doesn't exist in Dalmatian. Does anyone have a clue on what I should do ? Is it fine to ignore this hunk ? | 10:45 |
| rpittau | zigo: dalmatian is EOL since April now, I doubt we will merge anything there | 11:10 |
| rpittau | in any case since that function was introduced in a feature in ironic 29 (epoxy), it's safe to skip that chunk, although you will have to adapt the rest of the patch | 11:10 |
| opendevreview | Merged openstack/ironic stable/2026.2: Update .gitreview for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006462 | 11:29 |
| opendevreview | Merged openstack/ironic stable/2026.2: Update TOX_CONSTRAINTS_FILE for stable/2026.2 https://review.opendev.org/c/openstack/ironic/+/1006463 | 11:29 |
| zigo | rpittau: https://wiki.debian.org/OpenStack/CVE-2026 I am maintaining all of OpenStack up to Zed, doing backports by myself. | 11:48 |
| zigo | Thanks for the help, I'll do that. | 11:48 |
| rpittau | zigo: sure, np, so that's on debian side, let me know if I can help anyway :) | 11:49 |
| zigo | Yeah, Debian only. | 11:50 |
| zigo | priteau: CVE-2026-90461 fix is about doing more checks before the requests.auth.HTTPBasicAuth() call, right? | 12:24 |
| zigo | In such case, then bobcat and before aren't affected, right? | 12:24 |
| opendevreview | Merged openstack/networking-generic-switch master: Add RestconfSwitch base class for HTTP transport https://review.opendev.org/c/openstack/networking-generic-switch/+/992887 | 12:53 |
| TheJulia | good morning | 13:03 |
| iurygregory | good morning ironic o/ | 13:08 |
| TheJulia | good morning | 13:08 |
| kubajj | hello everyone o/ | 13:11 |
| clif | gm o/ | 13:18 |
| opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/37.0: Fix fast-track deployment failure with autodetect deploy interface https://review.opendev.org/c/openstack/ironic/+/1006517 | 13:19 |
| opendevreview | Merged openstack/networking-generic-switch master: Add RestconfOpenConfigSwitch driver https://review.opendev.org/c/openstack/networking-generic-switch/+/992888 | 13:20 |
| opendevreview | Merged openstack/networking-generic-switch master: Add RESTCONF documentation https://review.opendev.org/c/openstack/networking-generic-switch/+/992889 | 13:20 |
| opendevreview | Merged openstack/networking-generic-switch master: Rename netconf_models to yang_models https://review.opendev.org/c/openstack/networking-generic-switch/+/993090 | 13:20 |
| cardoe | TheJulia: ^ that'll conflict with some of your stuff | 13:21 |
| TheJulia | yeah, I know | 13:21 |
| TheJulia | :( | 13:21 |
| TheJulia | best to caffinate and do more | 13:22 |
| cardoe | Can I get ya do head nod two backports while ya caffeinate? https://review.opendev.org/c/openstack/networking-baremetal/+/1005793 https://review.opendev.org/c/openstack/networking-baremetal/+/1005795 | 13:25 |
| cardoe | Gonna try and rebase some builds and have the team run a bigger validation today. | 13:27 |
| opendevreview | Riccardo Pittau proposed openstack/ironic bugfix/37.0: Fix fast-track deployment failure with autodetect deploy interface https://review.opendev.org/c/openstack/ironic/+/1006517 | 13:33 |
| TheJulia | cardoe: done | 13:34 |
| cardoe | Thanks! | 13:40 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost stable/2025.2: CI: Fall back to HTTPS when cirros 0.5.3 is not cached https://review.opendev.org/c/openstack/bifrost/+/1006542 | 14:03 |
| rpittau | ^ fixing upgrade job in new 2026.2 branch | 14:03 |
| rpittau | yes, the change is in 2025.2 :) | 14:04 |
| zigo | TheJulia: Hi there! | 14:04 |
| zigo | I'm backporting the fix for CVE-2026-90461. Am I right that fix is about doing more checks before the requests.auth.HTTPBasicAuth() call, right? | 14:04 |
| zigo | In such case, since there's no such call before Caracal, then Bobcat and earlier aren't affected? | 14:04 |
| TheJulia | Your read on the fix is correct | 14:12 |
| zigo | :) | 14:13 |
| zigo | Thanks for confirming. | 14:13 |
| TheJulia | The "no such call before caracal" is incorrect depending on the path invoked" | 14:13 |
| zigo | Oh... | 14:13 |
| TheJulia | if you look we also prohibit secrets from getting shipped to the agent | 14:13 |
| TheJulia | and we inhibit on the load, there is just a few different ways the clients/connections can get invoked | 14:13 |
| TheJulia | and basic auth itself has been a thing for a longish time | 14:14 |
| zigo | Ok, so I can just get rid of the changes to ironic/common/image_service.py and keep the rest? | 14:14 |
| TheJulia | The bottom part of it as you modified it, I believe | 14:14 |
| TheJulia | or, as you noted earlier | 14:14 |
| TheJulia | The upper part, I don't think would really conflict, although if its not getting called in caracal, it can likely be dropped | 14:15 |
| zigo | Right, that was my understanding. | 14:15 |
| opendevreview | Merged openstack/networking-baremetal stable/2026.1: Remove binding:host_id setting for L2VNI subports https://review.opendev.org/c/openstack/networking-baremetal/+/1005793 | 14:18 |
| zigo | TheJulia: is_host_auth_permitted() is called from many other places, so I'll keep the function available. | 14:20 |
| TheJulia | Yeah, its the main helper | 14:20 |
| zigo | s/many places/one other place/ | 14:21 |
| zigo | Rest of is tests... | 14:21 |
| zigo | TheJulia: Bobcat doesn't even have a image_server_auth_strategy configuration option, so it really feels like the feature isn't there. | 14:23 |
| zigo | There's no image_server_user / image_server_password either. | 14:23 |
| TheJulia | oh! | 14:24 |
| TheJulia | then it might not be affected | 14:24 |
| opendevreview | Merged openstack/networking-baremetal stable/2026.1: Remove invalid requested-chassis opt for localnet ports https://review.opendev.org/c/openstack/networking-baremetal/+/1005795 | 14:31 |
| alegacy | Hi folks! Now that https://review.opendev.org/c/openstack/networking-generic-switch/+/984832 has merged, can we get a ngs version bump so that we can set it as the new minimum requirement for ironic for standalone networking? FYI, I'm preparing a change to ironic-networking to make use of this change now. | 14:39 |
| rpittau | alegacy: we've just released NGS as part of the new Openstack Hibiscus release | 14:44 |
| rpittau | we're probably going to wait for the official announcement before proposing new releases, most likely next week | 14:44 |
| TheJulia | Yeah, technically it is a library so all that has to pass before master branch version gets updated with it. | 14:45 |
| alegacy | oh wait, did I get that wrong. Is it in 10.1.0? | 14:59 |
| rpittau | alegacy: it's probably there | 14:59 |
| JayF | there's a yaml file in openstack/releases which maps sha:TAG:branch_created | 14:59 |
| TheJulia | I think merged after | 14:59 |
| JayF | #startmeeting ironic | 15:00 |
| opendevmeet | Meeting started Mon Sep 21 15:00:11 2026 UTC and is due to finish in 60 minutes. The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:00 |
| opendevmeet | The meeting name has been set to 'ironic' | 15:00 |
| clif | o/ | 15:00 |
| JayF | Welcome all to the weekly Ironic meeting. As usual we all gotta follow the OpenInfra Code of Conduct :) | 15:00 |
| TheJulia | o/ | 15:00 |
| JayF | #topic Announcements/Reminders | 15:00 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add Cisco NDFC REST API driver for VXLAN L2VNI https://review.opendev.org/c/openstack/networking-generic-switch/+/968484 | 15:00 |
| iurygregory | o/ | 15:00 |
| estherd[m] | o/ | 15:00 |
| Mahnoor | o/ | 15:00 |
| JayF | #note Ironic release targeted for integrated OpenStack 2026.2 release was just cut. Congratulations on creating more Irony! | 15:01 |
| kubajj | o/ | 15:01 |
| rpittau | o/ | 15:01 |
| JayF | Usually I'd put here to review ironic-week-prio hashtag things -- please do -- but maybe on release day is the one time I can avoid bothering folks to review :D | 15:01 |
| JayF | except those required branch patches :D | 15:01 |
| JayF | #note https://releases.openstack.org/hibiscus/schedule.html it's R-1. Next week is the official marketing release. | 15:02 |
| JayF | I suggest contributors have a hibiscus tea in celebration lol | 15:02 |
| JayF | #link https://etherpad.opendev.org/p/ironic-ptg-2027.1 | 15:02 |
| TheJulia | I was just going to note we should mention the ptg etherpad | 15:03 |
| JayF | #note Most important thing in the coming weeks: review and participate in PTG etherpad. Prepare new topics. Participate asynchronously in other folks' topics. If you are new to OpenStack, a hint: the etherpad *is the PTG* moreso than the video calls themselves. | 15:03 |
| JayF | I will suggest if folks have PTG topics up they think folks will be interested in, maybe a link w/line number during open discussion can kick things off | 15:04 |
| JayF | anything else for Announcements before we move on? | 15:04 |
| cid | o/ | 15:05 |
| JayF | #topic Working Group Updates | 15:06 |
| JayF | Is there anything from the AsyncIO folks? | 15:06 |
| iurygregory | o/ | 15:07 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add multicast BUM replication support to NDFC driver https://review.opendev.org/c/openstack/networking-generic-switch/+/984815 | 15:07 |
| iurygregory | I'm planning on run some downstream tests before pushing the patch with the async calls upstream | 15:07 |
| JayF | Sounds like a *very* good thing to merge early in the cycle | 15:07 |
| iurygregory | currently we already have a patch to add the async redfish client to be used | 15:08 |
| iurygregory | hopefully things will work well <fingers crossed> | 15:08 |
| JayF | We might get a 2027.1 release headline very very early :) Sounds great! Thanks. | 15:08 |
| JayF | Security coresec team update: | 15:08 |
| JayF | We've been on an advisory blackout. VMT usually enforces this project-wide for release week (next week) but we're honoring it this week for Ironic as well since we just cut our 2026.2 release. | 15:09 |
| JayF | There are no new public security bugs. | 15:09 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add global fallback device for controller-style drivers https://review.opendev.org/c/openstack/networking-generic-switch/+/1001200 | 15:10 |
| JayF | Moving on; next topic would be Discussion Topics but none were added to the agenda. | 15:10 |
| JayF | #topic Bug Deputy Updates | 15:10 |
| JayF | looks like Mahnoor identified some for us | 15:10 |
| Mahnoor | Yes, there were 3 in ironic | 15:11 |
| Mahnoor | It appears Julia has already fixed the 3rd one | 15:11 |
| TheJulia | what?!? | 15:12 |
| TheJulia | ;) | 15:12 |
| Mahnoor | :) | 15:12 |
| Mahnoor | I should probably mark that as completed | 15:12 |
| TheJulia | yeah | 15:12 |
| Mahnoor | 'Fix completed'? | 15:12 |
| JayF | there are a couple other interesting ones there | 15:12 |
| JayF | Mahnoor: if I mark it manually, I use released | 15:13 |
| Mahnoor | alrightie | 15:13 |
| JayF | Mahnoor: if the automation is not marking it as completed, it won't mark it as released later ime | 15:13 |
| Mahnoor | oh okay | 15:13 |
| JayF | Any volunteers to be the bug deputy next week? I'll note this might be a 2 week stint as I have a "cancel meeting in release week" in my back pocket for Open Discussion ;) | 15:14 |
| Mahnoor | I could do it | 15:14 |
| TheJulia | w/r/t https://bugs.launchpad.net/ironic/+bug/2167590 we likely need to treat read timeout as a harder failure instead of one to treat as a transient error. It can be transient, but we shouldn't go down a rabbit hole 10 times for 30 seconds :\ | 15:15 |
| JayF | ++ | 15:15 |
| JayF | Mahnoor: thanks, I'll mark you down as continuing to bug deputy until next meeting | 15:15 |
| TheJulia | its more of a hard failure anyhow, sever connects, but the software ain't home as it were | 15:15 |
| JayF | I do wanna note: thank you to everyone who has done bug deputy work. I don't know if newer folks in the project know how awful we used to do with bugs--people would file them and they would sit and sit | 15:15 |
| JayF | hundreds and hundreds of stale bugs nobody even looked at | 15:16 |
| TheJulia | cardoe: do you have any ideas on https://bugs.launchpad.net/ironic/+bug/2167283 how we would make right decision ? | 15:16 |
| JayF | very thank you :) | 15:16 |
| TheJulia | I can likely do it this week | 15:16 |
| TheJulia | would be good time to clean up some of the older stuffs | 15:16 |
| JayF | TheJulia: we have to have a token renewal for that to be useful | 15:16 |
| JayF | TheJulia: a cold boot can take longer than our current console timeouts :( | 15:16 |
| TheJulia | yup, if they are connected we should try to know and do something maybe | 15:17 |
| TheJulia | its a weirdly shaped problem | 15:17 |
| JayF | Yeah, I'd just suggest centering around an operator use case of troubleshooting a cleaning or deployment. That'd mean you'd likely need a console working, uninterrupted, for at least 45 minutes | 15:17 |
| cardoe | looking | 15:18 |
| JayF | cardoe: it's your "the vnc token is short" | 15:18 |
| TheJulia | I think we're going to need an async call to "extend the lifetime!" | 15:18 |
| TheJulia | but that means more surface/thing/logic | 15:18 |
| JayF | at a minimum, it might be wise to change the default token lifetime | 15:18 |
| JayF | assuming we have a call/method to revoke the token/tear down console | 15:18 |
| TheJulia | quite likely, yeah | 15:19 |
| JayF | Any further discussion on new bugs before continuing? | 15:20 |
| TheJulia | nah | 15:20 |
| JayF | No RFEs to review, so no need for that agenda item. | 15:21 |
| JayF | #topic Open Discussion | 15:21 |
| JayF | First thing I'd like to propose here: should we cancel the meeting next week, since it's R+0 | 15:22 |
| TheJulia | ... I think that might be nice | 15:22 |
| iurygregory | I'm ok with canceling the meeting | 15:22 |
| Mahnoor | No complaints from me as well | 15:22 |
| JayF | I don't see anyone racing to the defense of a meeting ;) | 15:23 |
| JayF | #agreed no meeting September 28, 2026 to celebrate 2026.2 release | 15:23 |
| cardoe | So I think we need a DB table and not use internal_driver_info | 15:23 |
| JayF | Any other items folks wanna bring up for open discussion? | 15:23 |
| cardoe | and an RPC between the console container and the conductor | 15:23 |
| cardoe | and a periodic when the console is opened to refresh the token periodically | 15:24 |
| TheJulia | cardoe: please write it up then, that might be reasonable! | 15:24 |
| cardoe | will do. I had to clear the brain fog of being gone. :-D | 15:24 |
| JayF | Last call for items for open discussion. | 15:26 |
| JayF | In the meantime, any volunteer to run the October 5th meeting? | 15:26 |
| TheJulia | I can | 15:26 |
| JayF | WFM | 15:26 |
| JayF | Agenda in place for 10/5. Thanks all. | 15:28 |
| JayF | #stopmeeting | 15:28 |
| JayF | #endmeeting | 15:28 |
| opendevmeet | Meeting ended Mon Sep 21 15:28:57 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:28 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/ironic/2026/ironic.2026-09-21-15.00.html | 15:28 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/ironic/2026/ironic.2026-09-21-15.00.txt | 15:28 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/ironic/2026/ironic.2026-09-21-15.00.log.html | 15:28 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Describe internal per-switch options as cfg.Opt instances https://review.opendev.org/c/openstack/networking-generic-switch/+/1006314 | 15:49 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Generate per-switch config option reference from cfg.Opt https://review.opendev.org/c/openstack/networking-generic-switch/+/1006315 | 15:49 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Deduplicate switch option docs against generated reference https://review.opendev.org/c/openstack/networking-generic-switch/+/1006316 | 15:49 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Recognise and document VXLAN L2VNI EVPN/multicast options https://review.opendev.org/c/openstack/networking-generic-switch/+/1006317 | 15:49 |
| JayF | https://bugs.launchpad.net/ironic/+bug/2166499 is now public | 16:30 |
| TheJulia | cardoe: so, thought regarding vnc | 16:47 |
| TheJulia | You could just update_node over rpc now, since remotable methods are enabled... | 16:47 |
| TheJulia | that could then respect the lock and all | 16:48 |
| cardoe | That would work. | 16:48 |
| cardoe | I didn't think of that. | 16:48 |
| * cardoe stares at Zuul and pokes with with a stick and sighs. | 17:41 | |
| TheJulia | ci unhappy? | 18:13 |
| cardoe | GitHub burped earlier and needed to recheck things. But it was on the workflow step so now gotta let it re-run to get it verified and then it workflows again. | 18:24 |
| cardoe | So wanted to confirm https://github.com/rackerlabs/ironic/blob/da6a25808d3d1d1c6c16178b21b885982e9b483f/ironic/conductor/manager.py#L1138-L1187 clears the instance info no matter what if the tear down fails. There's not a case that we'd want to keep that for a retry? | 18:25 |
| TheJulia | we don't know where it failed at off-hand, if the host was modified, so there is no real retry at the end. It is a "go back to a clean state" and try again on the outer loop | 18:36 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Ansible: allow disabling per-node python interpreter override https://review.opendev.org/c/openstack/ironic/+/1006616 | 18:38 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Ansible: default to ignoring per-node python interpreter https://review.opendev.org/c/openstack/ironic/+/1006617 | 18:38 |
| opendevreview | Merged openstack/ironic bugfix/37.0: Fix fast-track deployment failure with autodetect deploy interface https://review.opendev.org/c/openstack/ironic/+/1006517 | 18:48 |
| cardoe | TheJulia: well it clears the instance UUID off unconditionally. | 19:17 |
| cardoe | That's the "issue" | 19:17 |
| cardoe | And might not be an issue. Having folks sketch out in more detail the state machine | 19:18 |
| cardoe | But I do think I understand what they're saying. | 19:19 |
| cardoe | basically if the failure is at a vif detach for example then that port in theory is unknown... maybe it's plugged up... maybe it's not. | 19:20 |
| cardoe | But what is responsible for re-attempting the undeploy and to clear that? | 19:20 |
| cardoe | Nova could retry but now it's no longer associated with that node so it won't | 19:20 |
| JayF | undeploy leaves machines in clean failed | 19:21 |
| JayF | which means they aren't associated with tenants anymore | 19:21 |
| JayF | as soon as undeploy gets far enough to nuke the instance uuid, the server doesn't belong to that tenant anymore even if cleaning hasn't succeeded | 19:21 |
| JayF | nodes in clean failed with an instance uuid | 19:22 |
| JayF | **nodes in clean failed with an instance uuid used to be a failure case caused by race conditions in nova-compute | 19:22 |
| JayF | it's a feature not a bug that we don't do that anymore :D | 19:22 |
| cardoe | okay so back over to telling folks to sketch out that state machine operation that they believe happened and explain to me why it's bad we cleared that info | 19:24 |
| JayF | yeah basically only nodes in deploying/wait call-back (aka deploywait) should have a instance uuid | 19:25 |
| JayF | active is the only stable state where a node should have an instance uuid | 19:25 |
| JayF | even on deploy fail we pull the instance uuid | 19:25 |
| cardoe | makes sense to me | 19:26 |
| JayF | a rule of thumb is: would it make sense for nova to still think this is backing an instance | 19:27 |
| JayF | for the delete case: no; nova sends the call and considers it deleted | 19:27 |
| JayF | for the deploy fail case: no; nova reschedules to another node so it has to give up this one | 19:27 |
| JayF | I know not everyone uses nova, but instance uuid is essentially "does nova care about this?" | 19:28 |
| cardoe | yeah this was some nova delete case and I'm just giving feedback to the folks that came to the conclusion they came to that no assumptions are wrong... try again. | 19:29 |
| JayF | or you're still using onmetal with clustered compute manager from like, a decade+ ago :P | 19:38 |
| JayF | in which case you deserve all the breakage you get lol | 19:38 |
| JayF | (we had this breaky case all the time with that leaky/racey HA setup) | 19:38 |
| cardoe | no this was in 2026.1 versions | 19:39 |
| cardoe | well ironic is 38.0 instead of 36.0 | 19:39 |
| TheJulia | So, I've been a bit distracted | 20:14 |
| TheJulia | The other reason not to retry on the same node is because it could be misconfigured. In the tripleo days we saw some of that where deployes would "reschedule" around a broken or misconfigured node | 20:15 |
| TheJulia | we should be nuking things like vifs from high orbit if they ever got even partially attached but the record in ironic is not really really really attached until we move to hand over the node. Nova and Ironic both try to ensure the records are gone. | 20:16 |
| opendevreview | Julia Kreger proposed openstack/networking-baremetal master: Docs: Fix L2VNI driver name and reconcile driver ordering https://review.opendev.org/c/openstack/networking-baremetal/+/1006635 | 20:37 |
| JayF | Cores: If someone wants to chime in on the thread and do an Ironic update for OpenInfra Live it'd be ideal. I'm going to be wearing a VMT hat on that call, and having more faces on the call is better | 21:12 |
| TheJulia | I've already got an openinfra live recording on the books, so ideally someone else will volunteer | 21:23 |
| JayF | Maybe cardoe? Has he done one for Ironic yet? /me trying to think of DPL members who haven't taken a turn | 21:23 |
| TheJulia | ... I was already thinking of cardoe... | 21:24 |
| TheJulia | Prepare the mind control devices! | 21:24 |
| * TheJulia didn't sleep again last night | 21:26 | |
| TheJulia | wow, https://docs.openstack.org/releasenotes/ironic/2026.2.html is surprisingly long :) | 21:29 |
| clif | are those notes final now? I noticed a typo in one of my release notes :( | 21:35 |
| TheJulia | you can rev it on master if you must | 21:35 |
| TheJulia | its also not a huge deal | 21:35 |
| JayF | rev on master and backport to 2026.2 if you are :) | 21:35 |
| JayF | you might want to dig and do a bigger run through and fix all typos | 21:36 |
| clif | rgr | 21:36 |
| opendevreview | Doug Goldstein proposed openstack/networking-baremetal master: Make enable_l2vni_trunk_reconciliation the master switch https://review.opendev.org/c/openstack/networking-baremetal/+/1006660 | 21:48 |
| cardoe | There was a big push on release notes so we've got a lot more. | 21:48 |
| cardoe | I can do the Ironic update for OpenInfra Live | 21:49 |
| cardoe | We just gotta backport ^ if Claude's conclusion is correct. I'll have to read through it tomorrow cause I've gotta run. But noticed something weird from the test logs and threw it at Claude and that's what it said I needed to do. | 21:50 |
| opendevreview | Allain Legacy proposed openstack/ironic master: Add MTU support to ironic-networking https://review.opendev.org/c/openstack/ironic/+/1006662 | 22:19 |
| opendevreview | Doug Goldstein proposed openstack/networking-baremetal master: Name the listener pool queue after [DEFAULT] host https://review.opendev.org/c/openstack/networking-baremetal/+/1006663 | 22:32 |
| opendevreview | Allain Legacy proposed openstack/ironic master: Add MTU support to ironic-networking https://review.opendev.org/c/openstack/ironic/+/1006662 | 22:45 |
| opendevreview | Jay Faulkner proposed openstack/ironic-specs master: Add Portgroup/LAG support to ironic-networking https://review.opendev.org/c/openstack/ironic-specs/+/999448 | 22:54 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!