| opendevreview | Doug Goldstein proposed openstack/ironic master: build runbook traits via ORM relationship in create_runbook https://review.opendev.org/c/openstack/ironic/+/984663 | 00:12 |
|---|---|---|
| TheJulia | cardoe: super fair, and I'm sorry I'm sort of out full spoon depleation today | 00:20 |
| opendevreview | Merged openstack/networking-generic-switch stable/2025.2: Add Arista bond trunk support https://review.opendev.org/c/openstack/networking-generic-switch/+/983176 | 01:23 |
| opendevreview | Verification of a change to openstack/networking-generic-switch master failed: Add Dell OS10 bond trunk support https://review.opendev.org/c/openstack/networking-generic-switch/+/983131 | 01:23 |
| *** hroy_ is now known as hroy | 07:20 | |
| *** dmellado47174 is now known as dmellado4717 | 07:31 | |
| *** jroll02 is now known as jroll0 | 08:48 | |
| *** dmellado47172 is now known as dmellado4717 | 09:39 | |
| *** ykarel_ is now known as ykarel | 09:57 | |
| opendevreview | Harald Jensås proposed openstack/networking-generic-switch master: SONiC: Create VLAN before adding members/VNI map https://review.opendev.org/c/openstack/networking-generic-switch/+/984648 | 10:41 |
| opendevreview | mabulgu proposed openstack/ironic master: Improve vmedia insertion error messages https://review.opendev.org/c/openstack/ironic/+/982626 | 12:51 |
| TheJulia | good morning | 13:11 |
| cardoe | Since we've had a surge in contributors recently with our outreach projects I thought it'd be good to mention that I'm hiring an India based OpenStack developer. If folks that see this want to let the different folks they're working with know and I can provide more info. | 13:35 |
| cardoe | TheJulia: cocaine claude's code is slop. But I think maybe the rough idea with your deferred tasks might be good. | 13:36 |
| TheJulia | cardoe: great callout :) | 13:38 |
| cardoe | So one thing that bites is restarting conductor when there's on going work. Like a deployment or a firmware update. I know we've added stuff like don't actually go away until your done. But maybe we can somehow pickle/serialize that "on going work" state and have a new conductor pick it back up? | 13:38 |
| TheJulia | as for your... concainified (is that a word?!?) claude code output, I guess I'm sort of curious but also I'm feeling very ENOSPOON at the moment. | 13:39 |
| TheJulia | yeah, so the ideal pattern is in some long running processes put a "should I exit and punt" logic flow so the taking over conductor can pick up stuff like that which is sort of mid-flight but might be hung on a BMC reboot or something like that | 13:40 |
| cardoe | It's a Sonnet 4.5 that's not hosted directly by Anthropic and it's just.... different.... It's what I'd imagine OpenClaw people do. | 13:40 |
| cardoe | Someone turned the token output quantity up. And it loves words like "excited" and "amped". | 13:42 |
| TheJulia | Which was sort of always my vision of deferred tasks, get rid of anything that looks or smells like a waiting sleep, and try to get into a pattern of letting stuff get picked back which is really what is needed at a high level for allowing conductors to shutdown quicker and be able to resume as well. | 13:42 |
| TheJulia | so someone vibe modeled their own model and told it to be a vibe coder? | 13:42 |
| cardoe | So I threw the movement of sushy, the async sushy that Dmitri talked about, your deferred tasks, and some of Iury's BMC waits/states stuff at it as a big long context at it to see what it made. | 13:43 |
| TheJulia | ahh, yeah, it would have ended up with a semi-similar model | 13:43 |
| cardoe | And the result is that like you said... anything that is a sleep. It wanted to turn that into an async operation where it could put it down and come back to it at check intervals. | 13:43 |
| TheJulia | Cool, where do I hire this AI at? ;) | 13:44 |
| cardoe | Vibe modeled their own model and told it to be a vibe coder is exactly what this model is. | 13:44 |
| cardoe | The problem is its a really bad coder. Tests are failing? What if we just patch those tests to be happy. | 13:45 |
| * TheJulia facepalms | 13:47 | |
| TheJulia | sounds like it was way more aggressive than I ever expected we would be with the feature | 13:47 |
| TheJulia | it "amped" it up, man! | 13:48 |
| JayF | cardoe: you have a public job listing for that? | 14:06 |
| TheJulia | JayF: I think it is https://www.linkedin.com/jobs/view/4398509303. I saw it get shared a few days ago | 14:32 |
| cardoe | ha Julia was quicker than me. Yeah that's it. | 15:13 |
| cardoe | It's a direct report to me. And you'd be working on some of the PTG items and specs in the pipeline. | 15:14 |
| cardoe | Like... https://review.opendev.org/c/openstack/ironic-specs/+/984596 | 15:15 |
| cardoe | Did I do https://review.opendev.org/c/openstack/ironic-specs/+/984600 correctly for marking TBN as implemented in 35.0? | 15:16 |
| opendevreview | David Nwosu proposed openstack/ironic master: Add project field to node history and record deployment transitions https://review.opendev.org/c/openstack/ironic/+/983349 | 15:18 |
| kubajj | cardoe: I saw the change because of which you gave -1 to https://review.opendev.org/c/openstack/ironic/+/983349 was merged, do you wanna check it again? | 15:20 |
| cardoe | kubajj: I honestly don't understand the point of the change. | 15:21 |
| kubajj | cardoe: let me have a look at it again :D | 15:22 |
| cardoe | You're recording the project_id that the conductor is running as into the node history record? | 15:22 |
| kubajj | cardoe: I think for this it would be very useful: https://review.opendev.org/c/openstack/ironic/+/983349/6/ironic/conductor/deployments.py | 15:24 |
| kubajj | If I understand it right - this logs which project the node was provisioned from | 15:24 |
| cardoe | No it doesn't | 15:24 |
| cardoe | The leesee field is what you need. | 15:26 |
| kubajj | cardoe: what if the node doesn't have lessee or an owner | 15:26 |
| cardoe | Yep. That's why this needs some more design and thought. | 15:27 |
| kubajj | we have nodes like this - call it the hardware pool - users only get a flavour and they never see the node which the instance lands on | 15:27 |
| cardoe | But all this is doing is adding a DB field to log the project_id of the service project. | 15:27 |
| cardoe | That's exactly how my hardware pool is. | 15:27 |
| kubajj | ok, I see it now | 15:28 |
| kubajj | the task.context.project_id is different than I thought | 15:28 |
| cardoe | I'm just looking at where task.context is used and right around that cleaning log line that context is used to talk to neutron and the context's project_id is the service project. | 15:30 |
| TheJulia | yeah, it is only the requestor for the initial hit on the api and by requestor, the core account authenticated against ironic, which doesn't necessarilly mean the *user* | 15:44 |
| cardoe | and when you use nova to drive this it's "service" | 15:45 |
| TheJulia | +++++ | 15:46 |
| cardoe | cause that's who started the deploy or the cleaning | 15:46 |
| TheJulia | or whatever other project you use for talking to ironic | 15:46 |
| cardoe | And the original bug report was about the nova use case. | 15:46 |
| cardoe | Yeah. I'm using that as the default answer. I don't actually use "service" | 15:46 |
| opendevreview | Harald Jensås proposed openstack/networking-generic-switch master: SONiC: Enable privilege escalation for show cmds https://review.opendev.org/c/openstack/networking-generic-switch/+/984814 | 15:52 |
| 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 | 15:54 |
| 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 | 15:54 |
| JayF | cardoe: posted that job in our internal GR-OSS slack; we have a lot of mlh fellows and others pass through so I'll let you know if any referrals come thru :) | 15:59 |
| JayF | cardoe: hiring for Ironic is hiring for all of us :D | 15:59 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: De-duplicate VXLAN admin docs https://review.opendev.org/c/openstack/networking-generic-switch/+/984819 | 16:07 |
| opendevreview | Merged openstack/python-ironicclient master: Add typing to many `ironicclient/osc/v1/` modules https://review.opendev.org/c/openstack/python-ironicclient/+/983352 | 16:29 |
| opendevreview | Merged openstack/python-ironicclient master: Pre-test typing cleanup https://review.opendev.org/c/openstack/python-ironicclient/+/983360 | 16:31 |
| opendevreview | Merged openstack/python-ironicclient master: Add type annotations to `tests/unit/common/` and `tests/unit/utils.py` https://review.opendev.org/c/openstack/python-ironicclient/+/984285 | 16:31 |
| opendevreview | Merged openstack/python-ironicclient master: Add type annotations to `tests/unit` directory-level test files https://review.opendev.org/c/openstack/python-ironicclient/+/984287 | 16:31 |
| opendevreview | Merged openstack/python-ironicclient master: Add type annotations to `ironicclient/tests/unit/v1` https://review.opendev.org/c/openstack/python-ironicclient/+/984290 | 16:35 |
| opendevreview | Merged openstack/python-ironicclient master: Add type annotations to `ironicclient/tests/unit/osc` https://review.opendev.org/c/openstack/python-ironicclient/+/984295 | 16:44 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add opt-in MTU management for switch ports https://review.opendev.org/c/openstack/networking-generic-switch/+/984832 | 17:28 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add opt-in port carrier bounce on bind https://review.opendev.org/c/openstack/networking-generic-switch/+/984833 | 17:28 |
| cardoe | before I do any more of these clean ups... is this correct? https://review.opendev.org/c/openstack/ironic-specs/+/984600 | 18:19 |
| TheJulia | you link to the implemented spec, and remove the not-implemented link | 18:24 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: Add opt-in STP edge port and BPDU guard support https://review.opendev.org/c/openstack/networking-generic-switch/+/984847 | 18:32 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: ci: add "vxlan" ci job https://review.opendev.org/c/openstack/networking-generic-switch/+/981024 | 18:35 |
| 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 | 18:56 |
| 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 | 18:58 |
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: De-duplicate VXLAN admin docs https://review.opendev.org/c/openstack/networking-generic-switch/+/984819 | 18:58 |
| TheJulia | ugh, looks like the recent chnages to the devstack plugin have made some jobs super unhappy https://zuul.opendev.org/t/openstack/builds?job_name=ironic-tempest-ipa-wholedisk-direct-multinode&project=openstack/networking-generic-switch | 21:00 |
| cardoe | I didn’t do it… maybe | 23:54 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!