Thursday, 2026-07-02

opendevreviewAdam McArthur proposed openstack/ironic master: api: Add schema for conductors API  https://review.opendev.org/c/openstack/ironic/+/99576300:55
opendevreviewAdam McArthur proposed openstack/ironic master: api: Add schema for volume API  https://review.opendev.org/c/openstack/ironic/+/99576400:58
opendevreviewAdam McArthur proposed openstack/ironic-specs master: Add node workflows  https://review.opendev.org/c/openstack/ironic-specs/+/95253302:39
opendevreviewAdam McArthur proposed openstack/ironic-python-agent master: Add pyright type checking  https://review.opendev.org/c/openstack/ironic-python-agent/+/95833303:25
opendevreviewAdam McArthur proposed openstack/ironic-python-agent master: Add pyright type checking  https://review.opendev.org/c/openstack/ironic-python-agent/+/95833303:52
opendevreviewRiccardo Pittau proposed openstack/ironic-prometheus-exporter stable/2025.1: [stable only] Replace pkg_resources with importlib.resources  https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/99476806:56
opendevreviewMerged openstack/ironic master: Fix release codename in webapi version history  https://review.opendev.org/c/openstack/ironic/+/99541907:32
elodillesJayF: in general i'm helping out with CI maintenance on all open unmaintained branches. if the gates are not unrepairably broken, then i'll fix them, if it is reasonable to fix (with relevant CI coverage). so in short, i say let's try to leave it open, and if we reach the point when it is not feasible to keep them, then of course those branches needs to be EOL'd. 08:02
opendevreviewMerged openstack/bifrost master: Drop support for ironic-staging-drivers  https://review.opendev.org/c/openstack/bifrost/+/98613109:52
opendevreviewJed Preist proposed openstack/tenks master: Add Redfish support  https://review.opendev.org/c/openstack/tenks/+/99569710:04
skrobulcid / JayF :hi there! Quick question about https://specs.openstack.org/openstack/ironic-specs/specs/approved/kea-dhcp-backend.html - did you manage to start any implementation on the Ironic's DHCP provider? If so, is the code available anywhere (even if it's not ready)? 11:00
cidskrobul. We did. It's mostly unfinished.11:29
cidhttps://review.opendev.org/c/openstack/ironic/+/93564511:29
skrobulcid: thank you! that's still very helpful11:35
cid++11:39
opendevreviewEsther Domfeh proposed openstack/ironic master: feat: add state, target_provision_state, and duration_seconds to node history  https://review.opendev.org/c/openstack/ironic/+/98999411:42
opendevreviewEsther Domfeh proposed openstack/ironic master: feat: populate node_history state fields  https://review.opendev.org/c/openstack/ironic/+/99320811:42
opendevreviewEsther Domfeh proposed openstack/ironic master: Add request context user tracking to node history records  https://review.opendev.org/c/openstack/ironic/+/99390111:42
opendevreviewElod Illes proposed openstack/ironic-python-agent unmaintained/zed: Fix CI issues  https://review.opendev.org/c/openstack/ironic-python-agent/+/99520113:00
opendevreviewEsther Domfeh proposed openstack/ironic master: feat: add state, target_provision_state, and duration_seconds to node history  https://review.opendev.org/c/openstack/ironic/+/98999414:40
opendevreviewEsther Domfeh proposed openstack/ironic master: feat: populate node_history state fields  https://review.opendev.org/c/openstack/ironic/+/99320814:40
opendevreviewEsther Domfeh proposed openstack/ironic master: Add request context user tracking to node history records  https://review.opendev.org/c/openstack/ironic/+/99390114:40
clifJayF: I am loathe to admit it but I think claude was actually right, while *some* options are validated at config loading time (namely cli opts), the rest are not type validated until access time (this is a HUGE wtf to me)15:31
clifI have been nerd sniped into caring about oslo.config on my day off15:31
clifI cannot fathom *why* you would want to wait until you try to use or access a configuration option to validate it15:32
clifI want my program to blow up before it starts if the configuration is wrong at such a basic level that a configuration option fails type checking15:33
clifso claude was subtly wrong while being mostly right15:34
clifthe fact that type coercion/validation doesn't happen until access time is making me irrationally angry15:37
clifoslo-config-validator exists, but why wouldn't you want something like that to run when your program ingests the config at startup?15:48
opendevreviewDoug Goldstein proposed openstack/ironic master: Add ironic-manage command for driver discoverability  https://review.opendev.org/c/openstack/ironic/+/99584716:15
JayFclif: always remember the answer can be "someone wrote it this way in a hurry 15 years ago and changing it across dozens of applications may not be worth it"16:16
JayFclif: sounds like we have a path forward though: explicitly access any of our config items that require validation after loading of configs16:17
cardoeclif: speaking of config options... ^ that's for discoverability... and I'm looking to add your TBN evaluator into that command if folks are good with it16:21
JayFsgtm16:24
JayFclif: I thought you were off today? I didn't know you were susceptible to this level of nerd sniping :D 16:24
opendevreviewcid proposed openstack/ironic master: WIP: Console container provider configuration  https://review.opendev.org/c/openstack/ironic/+/99585216:30
opendevreviewcid proposed openstack/ironic master: WIP: Console container provider  https://review.opendev.org/c/openstack/ironic/+/99585316:30
opendevreviewcid proposed openstack/ironic master: WIP: Console container provider plugin support  https://review.opendev.org/c/openstack/ironic/+/99585416:30
opendevreviewcid proposed openstack/ironic master: WIP: Console container provider documentation  https://review.opendev.org/c/openstack/ironic/+/99585516:30
opendevreviewcid proposed openstack/ironic master: WIP: Console container provider tests  https://review.opendev.org/c/openstack/ironic/+/99585616:30
opendevreviewDoug Goldstein proposed openstack/ironic master: fix manual-management hardware type to use noop per the docs  https://review.opendev.org/c/openstack/ironic/+/99585716:51
cardoecid: ^ I believe that's what you had intended with your original change so I've marked that as backport17:05
cidIndeed. Thanks!17:53
opendevreviewMerged openstack/ironic master: Default TLS minimum version to 1.3  https://review.opendev.org/c/openstack/ironic/+/98795318:19
clifI'm going to hurl a chair through a window, does oslo.config even validate configuration type coercion when using `oslo-config-generator`? because so far in my limited testing it does not19:20
clifI have a string specified for an option that is explicitly an IntOpt and the validator doesn't bat an eye about it19:20
JayFThe folks who work on that are in #openstack-oslo, but it's like, 10% time of two people.19:21
JayFIf you can make it better, please do.19:21
clifit looks like validate makes sure the config parses and the options in the config file map to options defined in the config namespace specified19:23
clifand that's about it19:24
clifthat's a problem for future me19:24
JayFSounds good :) there's a lot of stuff like this hidden in the depths; a long time for tech debt to accumulate and sometimes you don't see the dusty corner until you go looking 19:33
opendevreviewMerged openstack/ironic master: fix manual-management hardware type to use noop per the docs  https://review.opendev.org/c/openstack/ironic/+/99585719:41
opendevreviewDoug Goldstein proposed openstack/ironic stable/2026.1: fix manual-management hardware type to use noop per the docs  https://review.opendev.org/c/openstack/ironic/+/99590020:56
cardoeclif: I believe it does not validate type coercion. It was a limitation of argparse and I had a mental note to revisit it.20:58
opendevreviewDoug Goldstein proposed openstack/ironic bugfix/37.0: fix manual-management hardware type to use noop per the docs  https://review.opendev.org/c/openstack/ironic/+/99590220:59
opendevreviewDoug Goldstein proposed openstack/ironic master: add types to hardware types for supported interfaces  https://review.opendev.org/c/openstack/ironic/+/99590421:06
cardoeIf I was a hardware type that was effectively two paired devices with two different BMCs would you make one a parent? No relationship at all? Or make some fake parent device and both real things as children?22:20
opendevreviewMerged openstack/ironic-python-agent unmaintained/zed: Fix CI issues  https://review.opendev.org/c/openstack/ironic-python-agent/+/99520122:21
JayFcardoe: depending on how my workflows worked, I might make that *one* custom hardware type with an expectation that the less-common bmc get a --management-interface= when required23:47

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