Wednesday, 2026-02-18

opendevreviewJulia Kreger proposed openstack/ironic master: CI: Add logic for vxlan/geneve network types  https://review.opendev.org/c/openstack/ironic/+/97708000:33
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708100:35
opendevreviewMerged openstack/networking-generic-switch master: Avoid unnecessary functional test runs  https://review.opendev.org/c/openstack/networking-generic-switch/+/97705300:55
opendevreviewJulia Kreger proposed openstack/ironic master: CI: Add logic for vxlan/geneve network types  https://review.opendev.org/c/openstack/ironic/+/97708001:53
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708101:57
TheJuliahopefully getting close01:59
opendevreviewSyed Haseeb Ahmed proposed openstack/ironic master: Support pluggable API middleware via entry points  https://review.opendev.org/c/openstack/ironic/+/97649404:47
opendevreviewRiccardo Pittau proposed openstack/ironic master: Workaround for latest docker in metal3 job  https://review.opendev.org/c/openstack/ironic/+/97704708:52
opendevreviewSyed Haseeb Ahmed proposed openstack/ironic master: Support pluggable API middleware via entry points  https://review.opendev.org/c/openstack/ironic/+/97649412:26
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: vxlan: Fix ironic-neutron-agent reporting  https://review.opendev.org/c/openstack/networking-baremetal/+/97721813:07
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708113:07
opendevreviewRiccardo Pittau proposed openstack/ironic bugfix/34.0: [bugfix only] remove metal3 integration job  https://review.opendev.org/c/openstack/ironic/+/97722414:15
opendevreviewMerged openstack/networking-baremetal master: tox: Drop redundant injection of VIRTUAL_ENV variable  https://review.opendev.org/c/openstack/networking-baremetal/+/97687514:21
opendevreviewRiccardo Pittau proposed openstack/ironic master: Workaround for latest docker in metal3 job  https://review.opendev.org/c/openstack/ironic/+/97704715:12
opendevreviewMerged openstack/ironic-python-agent master: tox: Drop redundant injection of VIRTUAL_ENV variable  https://review.opendev.org/c/openstack/ironic-python-agent/+/97687315:19
cardoeclif: I replied to your review15:21
cardoeTheJulia: on Haseeb's API middleware, I was going to as a follow on make the existing ones we had fall under his docs and be configured the same way if that makes sense.15:22
clifcardoe: Replied back. I hope what I'm saying makes sense.15:37
cardoeclif: No I don't follow.15:39
TheJuliacardoe: cool cool15:41
cardoeSo you want an instance/class variable that's the string names of the fields?15:41
cardoeCause those are available programmatically from the class definition. Otherwise keeping them as manual strings like that relies on humans to keep strings in sync with the names of parameters.15:42
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: vxlan: Fix ironic-neutron-agent reporting  https://review.opendev.org/c/openstack/networking-baremetal/+/97721815:48
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708115:49
TheJuliaso lets see if this works15:49
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: WIP: Crazy idea: NDFC support in NGS?  https://review.opendev.org/c/openstack/networking-generic-switch/+/96848416:35
TheJuliaAs an aside, I do wonder if nautobot or apstra might make a lot of sense to do something similar16:35
clifcardoe: I'm trying to say that ALL_KEYS and NECESSARY_KEYS should be defined as part of TraitAction so TraitAction.{ALL_KEYS,NECESSARY_KEYS} not defined in config_file.py17:52
*** dking is now known as Guest309418:02
cardoeclif: okay. They're computed off the class definition now. When you previously had them as just a list of arbitrary strings.18:02
*** Guest3094 is now known as dking18:02
cardoeI can compute them off the class definition and set them as class variables then?18:03
cardoeFunctionally the code in config_file.py is just a very basic implementation of jsonschema18:03
TheJuliaokay, where was my brain18:04
cardoeAll I want is to not have the case where something changes a human has to update two places. Because that's how the inspection_rules where done. The config file parser had the ALL_KEYS and NECESSARY_KEYS thing as well. And then someone updated the fields in the class and didn't update the fields in the config file parser and it was broken.18:05
cardoeTheJulia: what do ya mean with Nautobot?18:05
TheJulialike, port config management, see the change18:05
TheJuliasimilar to ndfc18:05
cardoeUltimately it's really too slow to do operations like that.18:05
TheJuliao'rly :(18:06
TheJuliaQueue endless sadness18:06
cardoeWe do the port operations ourselves with nornir/napalm and then let Nautobot's golden config and other stuff just act as a compliance checker18:06
cardoeBut we use the same repo of templates so the end result is the same18:06
TheJuliaHmm18:07
cardoeNautobot/NetBox take a change and generate a Job. They use Celery to farm out the job. But when you scale the Job stuff really becomes slow because you have to have a generic Job which then creates a pool specific Job which then runs your change.18:07
* TheJulia starts heating up some lunch to reset brain from meeting18:07
cardoeAnd the whole System Change management stuff they have in there must run and must capture the diffs18:08
TheJuliaoh, ugh18:08
TheJuliaokay18:08
cardoeOur folks reallly wanted us to have Nautobot drive it all.18:08
cardoePlus Nautobot's expression of VNIs is... lacking... NetBox is much better.18:09
TheJuliaso, it looks like I'm down to "why devstack do you not log or do anything around actual external_ids for bridge mappings with ovn18:09
cardoeNidhi, that did the recent Redfish inspection changes, is actually writing the improvements to Nautobot's VNI handling as we speak.18:09
cardoehttps://github.com/nautobot/nautobot/pull/846718:11
TheJuliaoh well, anyway, oven is heating to make me a small personal pizza and I have the hope it brings happy belly and devstack clarity as in why does it hate me18:12
clifcardoe: but why wouldn't the config file use TraitActions.ALL_KEYS and TraitActions.NECESSARY_KEYS? I agree with not duplicating code or logic, which is why I think those constants should stay on TraitActions18:20
clifanything that needs to use them can refer directly to them instead of re-deriving them18:21
cardoeI can add them on there then. They weren’t ever constants before. You had them as instance variables.18:22
opendevreviewcid proposed openstack/ironic stable/2025.1: Make RPCService picklable for oslo.service  https://review.opendev.org/c/openstack/ironic/+/97726118:54
dkingdtantsur: Would you happen to still be around?19:25
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: vxlan: Fix ironic-neutron-agent reporting  https://review.opendev.org/c/openstack/networking-baremetal/+/97721819:35
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708119:36
dkingMaybe somebody else could help me. I'm receiving a cleaning error that driver_info is missing deploy_kernel and deploy_ramdisk. Those should be set, so I'm trying to troubleshoot. I'm using the metal3-io ironic image. the same settings work in a different environment, so I'm confused.19:37
cardoecid: should that be against master?19:37
cidIt should :D. 19:51
cidI only noticed I haven't checkout out to master after the change was up.19:51
dkingI see that there's an 'OS_CONTROLLER__DEPLOY_KERNEL' env variable, but I'm not sure how in the working environment that gets into the ironic configuration.19:51
opendevreviewcid proposed openstack/ironic master: Make RPCService picklable for oslo.service  https://review.opendev.org/c/openstack/ironic/+/97726619:56
opendevreviewcid proposed openstack/ironic master: Make RPCService picklable for oslo.service  https://review.opendev.org/c/openstack/ironic/+/97726620:04
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: vxlan: Fix ironic-neutron-agent reporting  https://review.opendev.org/c/openstack/networking-baremetal/+/97721820:05
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708120:05
TheJuliacardoe: I feel like I'm chasing either broken ovn config or some weird slight incompatability20:06
cardoePossible?20:08
* TheJulia gets another data point20:09
cardoeLogs on the ci job?20:10
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: DNM: Disable segments to see if things play nice without them  https://review.opendev.org/c/openstack/networking-baremetal/+/97726720:10
TheJuliadisable segments20:11
TheJuliaSo, there is like a double-edged sword configuration wise20:11
TheJuliaif you want to dig, see: https://48f44a00bc5b9605050d-37d30e8ca4786c547186bac4ea4dd159.ssl.cf2.rackcdn.com/openstack/f358bff37ae54684a4e6cbbbfcba6126/controller/logs20:12
TheJuliaRegarding the upcoming PTG, We should start gathering thoughts/plans/ideas20:37
TheJuliahttps://etherpad.opendev.org/p/ironic-ptg-2026.220:37
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708120:55
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: DNM: Disable segments to see if things play nice without them  https://review.opendev.org/c/openstack/networking-baremetal/+/97726720:57
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: vxlan: Fix mech looping and supported type declaration  https://review.opendev.org/c/openstack/networking-baremetal/+/97721822:03
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708122:03
opendevreviewJulia Kreger proposed openstack/networking-baremetal master: WIP: CI: implement vxlan job  https://review.opendev.org/c/openstack/networking-baremetal/+/97708123:16

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!