Tuesday, 2026-02-03

opendevreviewSteve Baker proposed openstack/ironic master: Add an autodetect deploy interface  https://review.opendev.org/c/openstack/ironic/+/97318700:27
opendevreviewSteve Baker proposed openstack/ironic master: Replace ironic-tempest-bios-ipmi-direct with -autodetect  https://review.opendev.org/c/openstack/ironic/+/97426000:27
opendevreviewSteve Baker proposed openstack/ironic master: Add bootc deploy to autodetect default  https://review.opendev.org/c/openstack/ironic/+/97536500:27
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: devstack: Ignore error with file existing on restack  https://review.opendev.org/c/openstack/networking-generic-switch/+/97331000:48
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: l2vni plug case with Cisco NXOS  https://review.opendev.org/c/openstack/networking-generic-switch/+/96837700:48
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Arista EOS and vendor neutral SONiC  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276300:48
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Add Junos, Cumulus NVUE, and denote Dell OS10 as unsupported  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276400:48
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: OVS testing patch for 'vxlan' binding model  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276500:48
opendevreviewMerged openstack/networking-generic-switch master: devstack: Ignore error with file existing on restack  https://review.opendev.org/c/openstack/networking-generic-switch/+/97331004:21
rpittaugood morning ironic! o/08:11
rpittaugouthamr: just -1ed :)08:11
rpittauTheJulia: I was (still kind of am) sick :/08:11
jandersgood morning Ironic o/08:20
jandersrpittau hope you feel better soon!08:21
rpittauthanks janders, I'm feeling better, I will be online today, just reduced capacity :)08:22
jandersfolks, after W+1 for https://review.opendev.org/c/openstack/ironic/+/974985 (hw health monitoring doco, has 2x +2)08:37
jandersthanks in advance!08:37
rpittaujanders: workflowed :D08:43
jandersrpittau TY!08:44
opendevreviewMerged openstack/ironic master: Add documentation for node health monitoring feature  https://review.opendev.org/c/openstack/ironic/+/97498509:17
janders^ \o/ - last hardware health monitoring change for now09:22
jandersapi/client/sdk/doco DONE09:22
janderssince it's a feature we put on the roadmap during the PTG, do I need to note anywhere this is now complete?09:23
rpittaujanders: if we had a spec we can remove that from the not-implemented section09:41
rpittaubut besides that, release notes already serve the scope, and we will mention it during the open infra live for Gazpacho :) 09:41
jandersthanks rpittau. There wasn't a spec, so I think we're all done.09:41
jandersappreciate your assistance! :) 09:41
rpittauanytime :)09:41
opendevreviewAbhishek Bongale proposed openstack/python-ironicclient master: Fix field names for selection and json  https://review.opendev.org/c/openstack/python-ironicclient/+/97394810:36
opendevreviewAbhishek Bongale proposed openstack/python-ironicclient master: Fix field names for selection and json  https://review.opendev.org/c/openstack/python-ironicclient/+/97394810:51
opendevreviewAbhishek Bongale proposed openstack/python-ironicclient master: Fix field names for selection and json  https://review.opendev.org/c/openstack/python-ironicclient/+/97394812:27
TheJuliarpittau: eek :( Feel better! 14:13
rpittauTheJulia: thanks :)14:13
opendevreviewMerged openstack/ironic master: Add ironic-networking network interface  https://review.opendev.org/c/openstack/ironic/+/96647014:47
opendevreviewMerged openstack/ironic-tempest-plugin master: Enable ironic-tempest-bios-ipmi-direct in check/gate  https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/97478014:57
opendevreviewJad Haj Yahya proposed openstack/ironic master: Fix Redfish boot mode handling for read-only BMCs  https://review.opendev.org/c/openstack/ironic/+/97552115:00
opendevreviewAllain Legacy proposed openstack/ironic master: Improve exception handling in switch driver factory  https://review.opendev.org/c/openstack/ironic/+/96985215:41
opendevreviewAllain Legacy proposed openstack/ironic master: Address remaining review comments for rpc methods  https://review.opendev.org/c/openstack/ironic/+/97118415:41
opendevreviewAllain Legacy proposed openstack/ironic master: Add standalone networking service installation guide  https://review.opendev.org/c/openstack/ironic/+/96647115:41
opendevreviewJad Haj Yahya proposed openstack/ironic master: Fix Redfish boot mode handling for read-only BMCs  https://review.opendev.org/c/openstack/ironic/+/97552115:57
alegacycid:  thanks for reviews my remaining changes.  saw your comment and updated with a new revision.16:08
cidSure. I will take another look now. tks16:13
alegacythanks, and unfortunately the others lost their +2 in the rebase as well.16:20
opendevreviewJulia Kreger proposed openstack/ironic master: ci: Exclude test_baremetal_vm_multitenancy from multinode  https://review.opendev.org/c/openstack/ironic/+/97554316:56
* cardoe looks around.17:01
cardoeHey there was an ironic meeting. 17:01
* TheJulia gives cardoe a cookie17:01
TheJuliaYou okay?17:02
cardoeJust a little pre-PTO diet care of a stomach bug.17:02
TheJuliaugh :(17:03
cardoeSpent some time snuggling the porcelain yesterday and Sunday.17:03
cardoeI did speak with someone that's smarter than me on networking just a little bit ago and I'm diving into the VXLAN patches.17:04
TheJuliaokay17:21
TheJuliaDid that conversation include BUM traffic?17:21
* dtantsur is a responsible adult and is not giggling at "bum traffic"17:22
TheJuliaoh, the timing on that is excellent17:22
* TheJulia totally doesn't believe dtantsur... and concludes he is only pretending to be a responsible adult ;)17:22
dtantsur:D17:24
* TheJulia suddenly worries we might have broken cardoe somehow17:26
cardoeSorry. Context switching.17:27
cardoeYeah I was asking about BUM traffic17:27
TheJuliaSo, broadcasts, unknowns, and multicast.....17:30
* TheJulia dresses up as an awards announcer, and stands at a fancy podium and says "The award for killing the irc disucssion goes to... *opens envelop* ME!"17:35
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Arista EOS and vendor neutral SONiC  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276317:44
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Add Junos, Cumulus NVUE, and denote Dell OS10 as unsupported  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276417:44
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: OVS testing patch for 'vxlan' binding model  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276517:44
TheJuliaso, on the cisco patch, I backed away from the physnet mapping to singular mcast group for bum traffic handling as its against best practices at this point. The others... are in various states. 17:50
cardoeYeah I'm not sure how to best expose that.17:53
cardoeSo long term for a large scale I'd like a conductor groups, neutron XXXX, and nova cells? to maybe be the segmentation things.17:54
TheJulia... to articulate, at some point the building block is the physnet17:55
cardoeSo where do you see the recommendations for best practices and mcast groups btw?17:56
JayF.o(people use mcast groups still?!)17:56
TheJuliaIn that, and sort of disjointed in a way, is EVPN sufficient or is the expectation that disjointed patterns will be required with some management. I guess on the cisco side the limit is like 8k groupings for distinct routes, but I can't see mapping that to each VNI gets its own mcast address or group either.17:57
TheJuliacardoe: uhhh, hjensas left a comment with a link on a earlier cisco n-g-s revision17:57
cardoeYeah I saw that.17:57
cardoeHe used a version that was EOLed in 2019 and EOS in 2022.17:57
TheJuliahttps://review.opendev.org/c/openstack/networking-generic-switch/+/968377/5/networking_generic_switch/devices/netmiko_devices/cisco.py17:58
TheJuliagood callouts17:58
cardoeSo there's a design trade off as well. When you have a low number of VTEPs and a high number of VNIs then yes ingress replication. But in other scenarios multicast works out better.18:01
cardoeYou definitely cannot give each VNI its own mcast address.18:01
cardoeMost of the hardware you're working with is going to be limited to 128 mcast groups for the ASIC18:02
TheJuliaLooks like the docs have largely leaned *hardcore* over to just doing evpn18:02
cardoeBut honestly this is why you're wanting to do BGP EVPN for updates and not a flood and learn model.18:03
cardoeYeah.18:03
TheJuliaYeah, the math doesn't really work out there, and 128 mcast groups is also problematic. Seem slike just stripping out everything that is not evpn is sort of ideal18:03
TheJuliaand if thats a limitation on the asics, that is likely across vendors because some of those same ASICs are from a 3rd party18:05
cardoeThere's a lot of ways to go. I think the single use case initially makes more sense.18:05
TheJuliaso, really the model is likely just "go evpn only"18:05
cardoeYeah that's what I would say.18:05
cardoeCertainly can entertain other approaches as a 2.0 maybe?18:05
TheJuliaThe AI tools have largely gotten to pointing out issues with other vendors as reason to keep/lean on such, but strictly speaking, the evpn style is really best overall18:07
* TheJulia wonders if a spell was cast that caused james to appear :)18:14
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Arista EOS and vendor neutral SONiC  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276318:43
JayFalegacy: +2s throughout, including one with meaningful, actionable comments on the docs19:01
jrosserwhat actual switch ("Most of the hardwre youre working with") only supports 128 multicast groups?19:06
jrossercardoe: ^19:06
cardoeFor BUM traffic for the underlay configuration.19:07
cardoeFor VXLAN ingress replication19:07
cardoeNot generically multicast groups19:07
jrosserah important caveat19:08
alegacyjayf: thanks.  i've responded to your comments.  you ok with a follow-up patch for these or re-spin?19:20
JayFup to you, I wouldn't have +2'd if I needed it before it landed19:20
alegacyk19:20
jamesdenton@TheJulia one unchecks 'Connect Securely' in Textual. ¯\_(ツ)_/¯  19:51
JayFjoin the cult of irccloud James :D 19:53
jamesdentoni'd probably still be using IRCle if i could19:53
TheJuliaLuckily irccloud is nicely insulated, thus preventing creation of a pluribus.20:08
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Arista EOS and vendor neutral SONiC  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276320:11
opendevreviewVerification of a change to openstack/ironic master failed: Improve exception handling in switch driver factory  https://review.opendev.org/c/openstack/ironic/+/96985220:15
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Add Junos, Cumulus NVUE, and denote Dell OS10 as unsupported  https://review.opendev.org/c/openstack/networking-generic-switch/+/97276420:25
rm_work[m]JayF: Yeah, we have that validation step before the boxes enter the "Ready" state in our inventory management, so I just use that filter and the difference between Current State and Ideal State would just be that instead of me running the "get Ready machines from IM" that Ironic does it. The script to pull them already exists and I do nothing except take the YAML and put it somewhere and run a manual enroll script21:20
TheJuliacardoe: https://review.opendev.org/q/topic:%22ironic-vxlan%2221:26
cardoeGoing through them.21:26
* TheJulia goes on etsy to see if anyone has a cardboard cut-out of hypno-toad21:26
JayFif there's not a "we make a die-cut cardboard cutout of your svg" service on etsy or the internet, someone is missing a market opportunity lol21:27
opendevreviewVerification of a change to openstack/ironic master failed: Improve exception handling in switch driver factory  https://review.opendev.org/c/openstack/ironic/+/96985221:28
TheJuliawell, die-cut might be a bit much for a one off to convince reviewers.21:28
TheJuliaLASER!21:36
opendevreviewJay Faulkner proposed openstack/ironic-prometheus-exporter master: Add mypy infrastructure for type annotation migration  https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/97557821:53
opendevreviewJay Faulkner proposed openstack/ironic-prometheus-exporter master: Add type annotations to parsers/descriptions.py  https://review.opendev.org/c/openstack/ironic-prometheus-exporter/+/97557921:53
JayFThese are the PRs we'll be pointing our MLH fellow at as examples for the type annotations work. A perfunctory review by interested parties (cc cardoe cid ) would be appreciated 22:07
JayFplease do not approve those until after the MLH fellow has gotten a chance to onboard with them though :D 22:07
cid++23:31

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