Tuesday, 2026-03-10

opendevreviewyatin proposed openstack/ironic unmaintained/2024.1: [Stable 2024.1 Only] pin ironic-tempest-plugin to last working tag  https://review.opendev.org/c/openstack/ironic/+/97977304:11
opendevreviewyatin proposed openstack/ironic unmaintained/2024.1: [Stable 2024.1 Only] pin ironic-tempest-plugin to last working tag  https://review.opendev.org/c/openstack/ironic/+/97977306:58
rpittaugood morning ironic! o/07:51
opendevreviewVerification of a change to openstack/networking-generic-switch master failed: docs: address nits for 978799  https://review.opendev.org/c/openstack/networking-generic-switch/+/97976108:15
kubajjgood morning rpittau, and Ironic! o/09:02
rpittauhey kubajj :)09:02
jandershey Ironic'ers o/11:29
janderso/11:29
jandersdtantsur cardoe when you're online it would be great to have a quick chat about outstanding comments in https://review.opendev.org/c/openstack/ironic/+/97756911:30
jandersit feels like we're close to be able to merge it but I'm not 100% sure if the outstanding observations/concerns need further revisions or are more lab-testing material11:31
janderswould appreciate your insights!11:31
opendevreviewKonstantinos Karampogias proposed openstack/ironic master: Skip redundant BIOS reboot when settings already match  https://review.opendev.org/c/openstack/ironic/+/97982812:31
TheJuliagood morning13:08
opendevreviewDoug Goldstein proposed openstack/ironic master: fix(redfish): correct submission of DateTime to BMC  https://review.opendev.org/c/openstack/ironic/+/97972213:55
opendevreviewKonstantinos Karampogias proposed openstack/ironic master: Skip redundant BIOS reboot when settings already match  https://review.opendev.org/c/openstack/ironic/+/97982815:01
cardoejanders: yeah I need to get back to testing it15:31
opendevreviewRiccardo Pittau proposed openstack/ironic master: Add [redfish]connect_timeout config for faster unreachable BMC detection  https://review.opendev.org/c/openstack/ironic/+/97987215:42
opendevreviewRiccardo Pittau proposed openstack/ironic master: Add [redfish]connect_timeout config for faster unreachable BMC detection  https://review.opendev.org/c/openstack/ironic/+/97987215:43
opendevreviewRiccardo Pittau proposed openstack/ironic master: Add [redfish]connect_timeout config for faster unreachable BMC detection  https://review.opendev.org/c/openstack/ironic/+/97987215:44
cardoedtantsur, rpittau: https://review.opendev.org/c/openstack/ironic/+/979722 I think that's one you guys will see in Metal3 since I tried using that verify you've enabled there.16:23
cardoeI don't understand why its not commenting on the bug.16:24
dtantsurI don't think we use this particular step just yet, but thanks!16:25
rpittaucardoe: thanks! looks good besides the missing test16:36
TheJuliabrraaaains17:32
TheJuliaAnyone seen my brains today?17:32
JayFTheJulia: I just learned that vm brasseur was your old grandboss! I had no idea.17:35
JayF(she just sat for a gr-oss out interview)17:36
TheJuliaShe was not my grandboss. Separate management chain17:37
JayFeither way, still neat to find openstack connections I didn't realize existed17:38
TheJuliacool cool17:38
cardoedtantsur: so fwiw now that I am thinking about it. We should always strip microseconds.18:49
cardoePer the spec they’re always optional. But if you are going to send them, you must match the precision supported by the hardware. Which can be 3 or 6 digits.18:50
cardoeThe test always worked by not sending it.18:50
*** mdfr5 is now known as mdfr19:39
TheJulia+1 strip microseconds19:47
TheJuliabecause your going to drift throught the interaction, the most reliable thing is to add time but even then... yeah.19:48
clifJayF: do you know the 'right' place to teardown dynamic portgroups?19:51
clifvif_detach?19:54
JayF_do_node_tear_down19:54
JayFoooooh19:54
JayFthat's probably a better answer19:54
JayFwe need to ensure it gets run no matter what though, I wonder if we have strong assurances that gets called19:55
clifif the vif is attached to a dynamic_portgroup, detach it, then tear down the dynamic_portgroup?19:55
JayFyes19:55
JayFI'm just trying to think of any edge cases there20:01
JayFfor servicing, we should be detaching and reattaching to the existing network, we should ensure that's doing a detach/attach20:02
JayFsame idea for rescuing20:02
opendevreviewJulia Kreger proposed openstack/ironic master: redfish: increase cache password hashing iterations  https://review.opendev.org/c/openstack/ironic/+/97990320:04
TheJuliayeah, tear down as part of tear down of the deployed instance20:05
JayFare you saying vif_detach wasn't the right answer?20:06
TheJuliaoh20:06
TheJuliawell, hold up20:06
JayFclif: I think vif_detach is wrong20:06
TheJuliano, vif_detach is good20:06
JayFit doesn't get called, e.g., in rescuing cases20:06
TheJuliaits dynamic, so should be okay to it on attach/detach20:06
TheJuliano, it doesn't20:06
JayFbut isn't that *wrong* for this case?20:06
TheJuliais this where we get a state diagram out20:06
TheJulia?20:06
JayFhere's what I'm trying to reconcile20:07
JayFvif_detach only gets called from nova/external api user afaict20:07
JayFin cases of things like rescuing or servicing where we tear down/swap network on an active server20:07
JayFis it OK that the dynamic portgroup (at least the data model for it) is still in place when it's on a service or rescue network20:08
TheJuliaso vif_detach *is* dynamic in that you can ask for new things to get applied, I guess the question is "at deploy time" or "at attach/detach" time, detach is fine, but yeah.20:08
TheJuliayes, should be perfectly okay20:08
TheJuliajust because it exists, doesn't mean it has to be used20:08
JayFokay good, from claude and my digging20:09
JayFfor vif in vifs; do vif_detach; done gets called after unconfigure_tenant_networks removes the binding from neutron in _do_node_tear_down()20:09
JayFso vif_detach is the right place, and it's a feature-not-a-bug that we keep the dynamic portgroup in the data model (representing the port config if returned to active)20:10
JayFeven if we're servicing/rescuing and therefore on a servicing/recusing network20:10
JayFclif: ^ tl;dr yes vif_detach is the right place, re-confirmed20:10
TheJuliayup20:10
clifty I will proceed with vif_detach20:14
TheJuliaSo, feels like we need to figure out what date we want to speak about our advances?20:42
JayFI'm unsure what you mean?20:43
JayFI do owe a reponse to that ironic video chain, it's on my list for today20:43
JayFmy task list has not gotten below "overflowing" since being back20:43
TheJuliaYeah, step 0 is sort of figuring out a date20:53
JayFack; I was waiting to reply until I got slides started for us to start populating20:54
JayFbut I went ahead and kicked that reply out20:54
TheJuliaI think we're sort of in too many cooks trying to organize, but I think bullets on slides is a good starting point21:10
JayFI was literally just talking about taking the oif live standard slides -> clone -> s/openstack/ironic/ in a few places21:11
JayF"hey everyone here's a collab space"21:11
JayFit's just competing with a thousand other tiny things for my time21:11
JayFI'll get there today21:11
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: Fix port binding race by using the original segment data  https://review.opendev.org/c/openstack/networking-generic-switch/+/97991421:12
TheJuliaok21:15
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: vxlan: Multicast traffic options for cisco driver  https://review.opendev.org/c/openstack/networking-generic-switch/+/97933221:46
TheJuliacardoe: I feel you may have opinions on ^, lmk and we can discuss at some point21:47
JayFTheJulia: actually, I'm not doing that, because apparently the actual openinfra live slide decks aren't out yet :) 21:49
TheJuliastart a new deck maybe ?21:52
JayFthe existing ones had starter slides that looked hard for me to replicate21:52
TheJuliafloat the idea to allison to give us a clean starting point21:52
TheJuliadunno21:52
JayFe.g. sponsor logos everywhere21:52
JayFfloating the idea to allison is the right answer21:52
TheJulia++21:52
TheJuliaokay, running into town then21:52
cardoeI’ll peek.23:00
cardoestevebaker[m]: novnc doesn’t have to have ironic right? Like I can have novnc in a smaller container in the same pod as conductor?23:02
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/common/http.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97891723:14
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/events.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992423:20
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/shard.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992523:23
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/utils.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992623:24
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/conductor.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992723:28
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/deploy_template.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992823:32
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/inspection_rule.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97992923:37
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/runbook.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97993023:40
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/volume_connector.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97993123:45
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/volume_target.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97993223:46
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/client.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97993323:51
opendevreviewKaran Anand proposed openstack/python-ironicclient master: Add type annotations to `ironicclient/v1/port.py`  https://review.opendev.org/c/openstack/python-ironicclient/+/97993423:54

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