Tuesday, 2026-04-21

opendevreviewMichal Nasiadka proposed openstack/bifrost master: Add Debian 13 (Trixie) support  https://review.opendev.org/c/openstack/bifrost/+/97965105:56
opendevreviewMichal Nasiadka proposed openstack/bifrost master: Add Debian 13 (Trixie) support  https://review.opendev.org/c/openstack/bifrost/+/97965106:02
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:02
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:03
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:06
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:07
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:08
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:08
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:09
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:09
opendevreviewMichal Nasiadka proposed openstack/bifrost master: [WIP] Experimental support for ARM64 hosts on Debian  https://review.opendev.org/c/openstack/bifrost/+/92005706:09
opendevreviewMichal Nasiadka proposed openstack/bifrost master: Add Debian 13 (Trixie) support  https://review.opendev.org/c/openstack/bifrost/+/97965111:27
opendevreviewMichal Nasiadka proposed openstack/bifrost master: Add Debian 13 (Trixie) support  https://review.opendev.org/c/openstack/bifrost/+/97965111:27
dtantsurTheJulia: hey, quick q about OCI: is pointing at the digest of a *layer*, not the root manifest, supposed to work?11:54
TheJuliadtantsur: I think it would depend on the data in the oci registry for the layer. The intended form and entry is a tag or digest of a base line layer  or singluar file object. Technically a layer is a singlar file as well if I'm recalling the structural model correctly12:19
dtantsurIt looks like right now Ironic is consistent with ORAS CLI in not accepting layers, at least not the ones from Quay12:21
TheJuliawhy would you want a single layer for a container?12:23
* TheJulia begins sipping coffeeeee12:23
dtantsurIt's about artifacts, not containers. And the answer to question "why" is somebody just tried it downstream and complained that the feature does not work :)12:25
TheJuliaexactly, container layers themselves are not technically file artifacts even if under the hood they are really just tgz files with extended attributes12:26
dtantsurOn an unrelated note, "c" in "IPA inventory" is for "consistency" :-/12:28
TheJuliaI feel like more coffee is needed to understand that12:28
dtantsurWe're arguing naming downstream, and I'm looking for prior art in the inventory12:29
dtantsurWhere we now have 2 instances of "vendor" and 1 of "manufacturer", as well as 1 instance of "serial" and 1 of "serial_number"12:29
dtantsurWe also have "product", "product_name" and "model"12:30
TheJuliaAnd vendors themmselves who are often inconsistent12:30
dtantsurSurely. Does not excuse us though..12:30
TheJuliaI'm amazingly divorced from that portion of code12:32
dtantsurIt may be helpful for your mental health :)12:33
TheJuliahealth over all12:33
TheJuliaseriously though, vendor can be programmed to be different from manufacturer12:34
TheJuliaFor example, nvidia is a great example of this, they are the vendor in some cases for items, but one of three OEM vendors.12:35
TheJuliaAt one time when we were musing with the idea of branded solutions, our employer would have been the vendor, but oem would have been independently stamped12:35
TheJuliaserial/serial_number, yeeeah.12:35
TheJulialots of networking discussion and cross-project discussion today12:39
dtantsurI will be sooooo lost :D12:46
TheJuliac'est la vie ?12:49
TheJuliaPTG Time!13:03
JayFPTG time! cc cid clif :D 13:03
* TheJulia starts handing out virtual coffee13:06
cidOn my way...13:06
cidCross session w/Nova in 13mins @ https://meetpad.opendev.org/nova-ptg-2026.213:47
dtantsurI'll rejoin for Ironic topics afterwards13:54
TheJuliahttps://meetpad.opendev.org/nova-2026.2-ptg is the correct link13:56
stephenfinCould I get another +2 on https://review.opendev.org/c/openstack/ironic/+/985421 when you've time during/after today's PTG sessions, please and thanks13:56
cidTheJulia, tks.14:03
TheJuliaseems like we're likely going to need to jump to the neutron cross-project14:55
dtantsurso, back to Ironic topics in 1 hour?14:58
TheJuliaLooks like it15:00
cidIronic w/ Neutron  in the Ironic room15:03
opendevreviewMerged openstack/ironic master: objects: Prepare for oslo.versionedobjects 3.10.0  https://review.opendev.org/c/openstack/ironic/+/98542115:51
dtantsurbtw, our schedule is overflowing for today: we have 1 hour left and 100 minutes of planned topics15:53
opendevreviewJay Faulkner proposed openstack/ironic-specs master: WIP: Ironic talking to placement  https://review.opendev.org/c/openstack/ironic-specs/+/98570815:57
TheJuliacardoe: tried to capture more notes regarding the existing evpn/port scheduling topic your raised during the neutron session16:08
TheJuliaI think at least one of those topics set for 25 minutes should only take 5-ish minutes16:08
cardoeThank you!16:08
TheJuliaits super high leve16:08
TheJulialevel, but I think the scheduling is more about upfront host selection, the storage physnet (i hate I'm using physnet, but thats sort of the physical modeling mental model) is kind of along those lines16:09
cardoeMy son has a band trip and drop off for his stuff was this prior hour. So I was juggling many things.16:09
TheJuliaso $words and $challenges16:09
* JayF has been working on getting local inference working to get off the claude roller coaster17:02
JayFit's such a weird milestone, but I finally have a fully working local harness with qwen3 smart enough to search the web and summarize Ironic's release notes for 2026.217:02
-opendevstatus- NOTICE: The wiki.openstack.org service will be offline for a few minutes while we perform a database upgrade18:07
* TheJulia requests one freshly steamed brain18:08
TheJuliaso, question of the day which is unrelated to ironic: why do my new cameras and old $different_and_now_evil_vendor's cameras identify cats as people, but dogs as pets?18:15
cardoehttps://review.opendev.org/c/openstack/ironic/+/985117 is an important one-ish for me18:40
cardoeSo that I don't have to smack people.18:40
cardoeSave people from a hand print!18:41
* TheJulia builds a room around cardoe with a HandKey II for cardoe to exit it. 18:43
* TheJulia wonders if $employer_from_the_early_2000s ever removed her from the handkeys as an admin18:44
rm_work[m]oh hey, is there official guidance now on AI-assisted commits, like do we need to flag them or something19:28
rm_work[m]oh wait is it PTG week>?19:28
JayFyes and yes19:30
JayFopeninfra ai policy is on the foundation website, if you search 'openinfra ai policy' you can find it19:30
JayFplease read it, but tl;dr label LLM inference usage with "Assisted-by: mySuperCoder with BobModel37" or "Generated-by: aClanker with BrainPointOh"19:31
JayFand please for the love of everything good in the world do not push code you couldn't explain without having your robot friend alongside :D 19:31
rm_work[m]lol19:34
rm_work[m]don't worry, keeping all of that garbage internal19:35
JayFthat last line is 100% jay and 0% policy lol19:35
JayFLLMs are great at being a force multiplier.19:35
JayFThey don't check if they are multiplying the good or the bad :D 19:36
opendevreviewVerification of a change to openstack/ironic master failed: fix: bios.apply_configuration needs to validate before starting  https://review.opendev.org/c/openstack/ironic/+/98511719:46
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: Add opt-in MTU management for switch ports  https://review.opendev.org/c/openstack/networking-generic-switch/+/98483220:24
opendevreviewVerification of a change to openstack/ironic master failed: fix: bios.apply_configuration needs to validate before starting  https://review.opendev.org/c/openstack/ironic/+/98511720:34
TheJuliahjensas, stevebaker[m], cardoe: https://review.opendev.org/c/openstack/neutron/+/98573221:07
TheJuliaLikely needs to be in neutronlib, but we'll see what they say21:08
TheJuliaJayF: I, for one, am deeply concerned if Clanker's from the star wars universe are appearing ;)21:28
opendevreviewVerification of a change to openstack/ironic master failed: fix: bios.apply_configuration needs to validate before starting  https://review.opendev.org/c/openstack/ironic/+/98511722:59
* cardoe side eyes the test infra23:13
TheJulia.... should we bring a corgi to yield ?23:13
opendevreviewVerification of a change to openstack/ironic master failed: fix: bios.apply_configuration needs to validate before starting  https://review.opendev.org/c/openstack/ironic/+/98511723:18

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!