Monday, 2025-12-01

opendevreviewcid proposed openstack/ironic master: WIP: Redfish VirtualMedia NFS and SMB/CIFS Support  https://review.opendev.org/c/openstack/ironic/+/95850500:44
*** hroy_ is now known as hroy02:36
rpittaugood morning ironic! o/07:04
kubajjgood morning rpittau, and ironic o/07:20
rpittauhey kubajj :)07:21
opendevreviewRiccardo Pittau proposed openstack/bifrost master: [DNM] TEST CI  https://review.opendev.org/c/openstack/bifrost/+/96879107:33
rpittaunew ovmf packages are out for noble and jammy, it looks like the issue with uefi has been fixed07:38
opendevreviewVerification of a change to openstack/ironic master failed: Add standalone networking service for ironic  https://review.opendev.org/c/openstack/ironic/+/96646609:05
rpittauadn we're good again to switch to DIB https://review.opendev.org/c/openstack/bifrost/+/964404 \o/10:13
opendevreviewVerification of a change to openstack/ironic master failed: Add standalone networking service for ironic  https://review.opendev.org/c/openstack/ironic/+/96646610:26
dtantsurrpittau: job names still need fixing, everything else looks good10:29
rpittau/o\10:32
dtantsurA new idea for today's RFE review: https://bugs.launchpad.net/ironic/+bug/213349910:47
rpittaudtantsur: just to clarify, I'm going to remove dibipa from all the jobs as we don't have other methods for ipa ramdisk now, and add "on" to indicate the node, and "src" to indicate we build the ipa image and we dont use the pre-built one10:52
rpittauso we'll ahve10:52
rpittaubifrost-integration-dibipa-debian-centos-10 -> bifrost-integration-debian-src-on-centos-1010:52
rpittaubifrost-integration-dibipa-centos-10 -> bifrost-integration-on-centos-1010:52
rpittausuggestions welcome for other renaming! :D10:52
dtantsurs/src/ipasrc/ otherwise I quite like this!10:54
rpittauperfect! :D11:01
opendevreviewMerged openstack/ironic-python-agent master: Fix slow test_find_routable_addr_no_reachable_urls  https://review.opendev.org/c/openstack/ironic-python-agent/+/96848812:02
opendevreviewJacob Anders proposed openstack/ironic master: Add hardware health monitoring via management interface  https://review.opendev.org/c/openstack/ironic/+/96694613:13
opendevreviewJacob Anders proposed openstack/ironic master: Add hardware health monitoring via management interface  https://review.opendev.org/c/openstack/ironic/+/96694613:27
opendevreviewRiccardo Pittau proposed openstack/bifrost master: Remove tinyipa support and switch to DIB-based IPA  https://review.opendev.org/c/openstack/bifrost/+/96440414:01
cidSo, I tagged Kolla on this bug, is really a Kolla bug, https://bugs.launchpad.net/ironic/+bug/213296014:01
rpittaucid: indeed14:02
cidAlright, thanks :D14:02
rpittau:)14:02
opendevreviewRiccardo Pittau proposed openstack/ironic master: Use new bifrost ci job name  https://review.opendev.org/c/openstack/ironic/+/96907714:07
opendevreviewRiccardo Pittau proposed openstack/bifrost master: Removed unused tinyipa CI jobs  https://review.opendev.org/c/openstack/bifrost/+/96908314:15
TheJuliaGood morning14:18
opendevreviewMerged openstack/ironic master: Add standalone networking service for ironic  https://review.opendev.org/c/openstack/ironic/+/96646614:20
opendevreviewJacob Anders proposed openstack/ironic master: Add hardware health monitoring via management interface  https://review.opendev.org/c/openstack/ironic/+/96694614:26
opendevreviewJacob Anders proposed openstack/ironic master: Add hardware health monitoring via management interface  https://review.opendev.org/c/openstack/ironic/+/96694614:29
janders^ kubajj apologies I added depends-on in the wrong patch14:30
janders(this push removed your review vote)14:30
* TheJulia blinks14:30
dtantsurmorning TheJulia14:31
cardoeHello all.14:59
TheJuliagood morning!14:59
cardoehttps://review.opendev.org/c/openstack/ironic/+/968418 if someone will nudge that. I'll start the backports.14:59
TheJuliaWho is going to be running today's meeting14:59
JayFI can but I will be eating breakfast while doing so :)15:00
TheJuliaThis is fine.15:00
TheJulia:)15:00
JayF#startmeeting ironic15:00
opendevmeetMeeting started Mon Dec  1 15:00:34 2025 UTC and is due to finish in 60 minutes.  The chair is JayF. Information about MeetBot at http://wiki.debian.org/MeetBot.15:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.15:00
opendevmeetThe meeting name has been set to 'ironic'15:00
TheJuliaI've not had breakfast much less coffee yet :)15:00
dtantsuro/15:00
janderso/15:00
kubajjo/15:00
TheJuliao/15:00
JayFWelcome to the Ironic weekly meeting. We operate under the usual Code of Conduct. 15:00
JayF#topic Announcements15:00
JayF#note  Standing reminder to review patches tagged ironic-week-prio and to hashtag any patches ready for review with ironic-week-prio: https://tinyurl.com/ironic-weekly-prio-dash15:01
rpittauo/15:01
JayFIt's R-17. Trailing release cycle stuff for Flamingo is due by today; that doesn't impact us.15:01
JayFs/today/this week/15:01
iurygregoryo/15:01
clifo/15:01
JayF#topic Working Group Updates15:02
JayFDo we have an update on Standalone Networking?15:02
alegacyo/15:02
alegacyYep15:02
alegacyupdated my patch series last week to address everyone comments15:02
cido/15:02
alegacyi see that one of them has merged as of this morning.15:03
TheJulia\o/15:03
alegacyif there are further comments on the remaining patches I'll take a look!15:03
JayFYou wanna link the chain here for the notes?15:03
alegacyhttps://review.opendev.org/q/topic:%22feature/standalone-networking%22+status:open15:04
JayF#link https://review.opendev.org/q/topic:%22feature/standalone-networking%22+status:open15:04
JayF#note patches are landing for standalone networking; please prioritize review15:04
JayFThank you for the update alegacy, I assume that's it?15:04
alegacyyes sir15:04
JayFAnything new on AsyncIO work?15:04
dtantsurNo, and won't be for a week more15:06
JayFAck.15:06
dtantsur2 weeks really15:06
JayF#note No new info on AsyncIO, probably not for 2+ weeks.15:06
JayFVXLAN Networking, then?15:06
TheJuliaI believe I revised the spec I've got posted. Discussions seem stalled at the moment.15:06
TheJuliadtantsur: I did reproduce the toc or lack of toc entry issue. That is weird, but I'll try to figure it out this week.15:07
TheJuliaI won't be around next week, so likely no progress on my end until sometime until later in the month15:07
JayF#note VXLAN spec updated; please review. Progress note expected until later in the month.15:08
JayF#topic Discussion Topics15:08
JayFI think the first one here, based on my reading of the ML earlier, is a leftover?15:08
JayFWe have already decided to cancel Dec. 22, Dec 29 meetings, yes?15:08
rpittauyep15:08
TheJuliaYes, I sent out that email already15:08
TheJuliaits a left over15:08
JayFWhat about iRMC/DMTF items, new or leftover cc: janders15:08
jandersnew15:09
JayFyou have the floor15:09
jandersw/r/t iRMC we are very much done. Depreciation change is merged. FJ were consulted and are on board, so we're nearly done. Should I go ahead and send an email to the ML stating what we decided to close the loop?15:09
TheJuliaYes please15:10
JayFI'd say yes, plese15:10
jandersOK, will do. Once that's done we can tick this item off. I will team up with someone more experienced since it's the first depreciation I am handling.15:11
jandersAny closing comments on iRMC? If not I will move to the DMTF topic.15:11
jandersOK, moving on.15:11
jandersI approached DMTF folks about the issue of machines with key resources spread across multiple Systems. They are happy to talk about it and advised best way to get attention is to open an issue in their GH space.15:12
JayFI also have had an interaction with someone invested in DMTF, who pointed me at two people, one of whom is mraineri, for feedback. I suspect this will be duplicating noise but I won't turn down the opportunity to help advocate for a more Ironic-friendly Redfish spec :D 15:13
jandersUnfortunately this space is private, but I do have access to it. I think it's worth pursuing this. However I think it would be best if we can colaborate on the problem description to best explain to them what we want to know15:13
jandersyeah I also exchanged emails with Mike15:13
jandersJayF since you raised this issue, I thought perhaps the two of us collaborate via gDoc with whoever else may be interested?15:13
jandersonce we have a draft for the issue I can post it and raise it in upcoming DMTF meetings15:14
janders(upcoming/tomorrow one is heavily booked, but the following one should be better for this)15:14
janderswould this be a good course of action? WDYT folks?15:14
TheJuliaI think a doc makes sense to get an issue in place as a starting point15:14
JayFjanders: I think that's a good idea, but I don't really have cycles this week. I'm flying out to Dallas on Wednesday.15:14
jandersJayF noted. I will share the gDoc with you anyway so you can contribute when you have time.15:15
TheJuliaSounds like this may be a slow burn until the end of the year, tbh15:15
jandersTheJulia would you have some time to look at the initial version of what I am putting together?15:16
JayF#note Those interested in collaborating on a feedback packet for DMTF should contact janders.15:16
janders++15:16
JayF#action janders to email the list about iRMC retirement15:16
jandersOK, that's me, thanks folks15:16
TheJuliajanders: I would be happy to, I'm not sure I will have anything constructive in the short term, but if you ping me I'll take a look.15:16
jandersthank you TheJulia15:16
JayF#topic Bug Deputy Updates15:16
JayFcid: it says you were the bug deputy 15:17
cidYeah. I was15:17
cidUpdate, 1 RFE 15:18
JayF#link https://bugs.launchpad.net/ironic/+bug/213349915:18
TheJuliaI can see where folks might be pushing us and dmitry wants to go to solve it, I think a slightly deeper discussion may be needed because it feels like we're unlocking a closet of foot-guns.15:18
TheJuliaMaybe that is a spec, I dunno.15:19
JayFAt a minimum I don't fully grok the use case15:19
TheJuliaI kind of do, but it feels like a footgun shaped giant lever15:19
JayFonce I understand what the use case is and why a hard node delete doesn't solve it I would have a better opinion15:20
TheJuliaThat is great point15:20
TheJuliaAnd it would help15:20
janderswas thinking the same - why not force-delete?15:20
TheJuliawe've been skittish of anything which is a forcing behavior in the past because it begins to reset expectations15:21
TheJuliaI guess I'm worried the next is "force active with a verb", which heads us on a path to "force ipa to make me whipped cream for my pies"15:21
JayFWe'll get more detail.15:21
JayFI put a comment in the bug just now.15:21
TheJuliaI did as well earlier15:22
JayFI don't care if that detail is in spec or more-fleshed-RFE form15:22
TheJuliaTBF, we're talking verbs, behavior changes, and state machine as well15:22
TheJuliaThat really puts it in spec territory15:22
TheJuliaBut, a little more detail first might help the discussion15:22
JayFOK, who wants to be bug deputy this week?15:24
cidI will be bug deputy this week. I think most folks will be out.15:24
TheJuliaI can, but I won't be around next monday15:24
TheJuliacid: ack15:24
JayF#note CID to be bug deputy again15:24
JayF#topic Open Discussion15:24
JayFAs a first note for open discussion, I'll be out the last half of this week traveling so if you need something from me speak soon :)15:25
TheJuliaFair, I'll be out next week15:26
jandersSince it's quiet I will go: I have a question related to hardware health monitoring (https://review.opendev.org/c/openstack/ironic/+/966946). In one of the reviews Steve asked me if there is an issue or a spec for this. From the PTG discussion it seemed to me this is lightweight enough not to need one. Am I right or do you feel there is further15:27
jandersdiscussion needed about this? Asking so I can give Steve a solid answer.15:27
JayFmy usual personal rule is to make a rfe bug if someone asks for one15:28
JayFand upgrade that to a spec if demanded in the meeting15:28
JayFbut that's not like, policy, just now I approach ti15:28
TheJuliaI think consensus was largely "we agree to the idea", but we didn't require a spec. RFE bug does make sense to just treat as something approved to track it15:28
* dtantsur finally back to the screen15:28
dtantsurJayF: force deletion won't clean up anything, that's a big issue now15:28
dtantsur(cache, PXE configs, neutron configs, you name it)15:29
dtantsuralso right now force deletion is only possible in a couple of states15:29
JayFThat helps focus the use case for sure15:29
dtantsurTheJulia: retired is an entire different purpose (planned retirement vs "this thing burned in its track")15:30
TheJuliafair enough, why an entirely new db field though?15:30
dtantsurThink about getting a bricked machine out of the provision state samsara 15:30
dtantsurTheJulia: so that various code throughout Ironic can stop doing what it's tryign to do15:31
dtantsure.g. trying to power it on/off during cleaning, even though the thing will never respond to IPMI packages ever again15:31
TheJuliaWell, if there is an open task, they aren't going see if if they currently hold another task15:31
TheJulia(sort of feels like something we should be doing with the state machine too, but more pondering needed)15:32
dtantsurThat's a fair point. Another part of it is to be able to reject new tasks.15:32
dtantsurAlso, your long-running task spec may make TaskManager re-entrant15:32
JayFIf y'all are Ok with this discussion happening outside the meeting, I'm going to close up the official logger?15:33
TheJuliaeh, sort of, but i have no time to revisit that at hte moment15:33
dtantsursure, it's not urgent15:33
dtantsurin fact, I won't have time to work on it in any near future15:33
TheJuliaReally, the deferred task bit expects the state might change on the task itself, at least, that is how I envisioned it15:33
dtantsur(but maybe some interns...)15:33
TheJuliafair far15:33
TheJuliadtantsur: could you put a little more clarity in to the RFE and lets try to revisit the discussion in a few days?15:34
TheJuliaJust give brains time to do the brain thing15:34
dtantsurdefinitely. I want to write the thought down before I forget it15:34
dtantsur(squirell brain, you know)15:34
TheJuliafair fair15:34
JayFAny other topics for Open Discussion?15:35
jandersI will create that RFE for monitoring.15:35
jandersThanks!15:36
JayFThat does it for the Ironic meeting, thanks all, see you next week same metal-time, same metal-channel o/15:36
JayF#endmeeting15:36
opendevmeetMeeting ended Mon Dec  1 15:36:40 2025 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:36
opendevmeetMinutes:        https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-12-01-15.00.html15:36
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-12-01-15.00.txt15:36
opendevmeetLog:            https://meetings.opendev.org/meetings/ironic/2025/ironic.2025-12-01-15.00.log.html15:36
jandersdtantsur while I have you, would you have cycles to collaborate on DMTF issue gDoc for multi-System-machines?15:37
janders(lines timed from 16:12 in the meeting)15:38
dtantsurjanders: not before next week, sorry15:38
dtantsurYou don't imagine how busy I am downstream this week15:38
jandersdtantsur totally understand, leave this with me15:38
jandersI may share the gDoc with all interested but that's just so you have an option to comment when you have time15:40
opendevreviewIury Gregory Melo Ferreira proposed openstack/ironic master: Make post-firmware-update reboot conditional on component  https://review.opendev.org/c/openstack/ironic/+/96634415:41
opendevreviewMerged openstack/ironic master: fix: ensure that the bootloader path exists  https://review.opendev.org/c/openstack/ironic/+/96841816:10
dtantsurOh damn, should have mentioned the broken (?) SIGHUP support on the meeting16:59
* dtantsur is curious if anyone cares in reality16:59
jandersdtantsur as in kill -HUP triggering low-disruption Ironic config reload?17:02
TheJuliaso17:02
dtantsurWell, that's what it used to do17:02
TheJuliawell17:02
dtantsurnow it's reloading the entire conductor worker17:02
TheJuliamy opinion is that oslo broke it and its actually a correct change for mutlithreaded universe17:03
TheJulia(likely, not the popular opinion, but its much more a side effect from my point of view)17:03
dtantsurFrom the container perspective, I don't care too much since any configuration change is a restart for us anyway17:03
TheJuliaI think that is the other aspect, the world is shifting to that as a model17:03
dtantsurbut I wonder if some consumers appreciated being able to e.g. change to debugging mode without draining the entire conductor17:04
TheJuliawhich gives me a bit less urgency than to fight the "mutate should mutate!"17:04
TheJuliaWell, you can do a quick restart, you just break stuff in flight17:04
opendevreviewDoug Goldstein proposed openstack/ironic stable/2025.2: fix: ensure that the bootloader path exists  https://review.opendev.org/c/openstack/ironic/+/96912017:04
TheJuliaeven then, steve has shorted a lot of that window (if I'm recalling correctly)17:05
TheJuliaThere is also proper drain and such17:05
opendevreviewDoug Goldstein proposed openstack/ironic stable/2025.1: fix: ensure that the bootloader path exists  https://review.opendev.org/c/openstack/ironic/+/96912217:05
TheJuliaThats a bit different, and actually is broken depending on the eventlet version on older versions (*twitch* *twitch*)17:05
opendevreviewDoug Goldstein proposed openstack/ironic stable/2024.2: fix: ensure that the bootloader path exists  https://review.opendev.org/c/openstack/ironic/+/96912317:06
opendevreviewDoug Goldstein proposed openstack/ironic bugfix/31.0: fix: ensure that the bootloader path exists  https://review.opendev.org/c/openstack/ironic/+/96912417:06
jandersI retrospectively created RFE for Redfish health monitoring: https://bugs.launchpad.net/ironic/+bug/2133522 17:30
cardoeclif: you good with my feedback on TBN?18:42
TheJuliaBRRAAAAINS18:44
clifcardoe: yep, I'm working on it18:45
clifdidn't know about ddt 18:45
cardoegood deal. just making sure I'm not a blocker18:45
TheJuliaddt is awesome, but... I think there was soemthing that makes me hesitant18:45
clifalthough I'm a little peeved that it doesn't provide an easy way to set the __name__ of the test data object18:46
clifeven in the example it does a little dance to set it... why not just provide that as part of the library or make it easy18:46
opendevreviewMerged openstack/python-ironicclient master: A new `instance_name` field to the node object  https://review.opendev.org/c/openstack/python-ironicclient/+/95510218:46
clifoverall it does seem a bit simpler than doing my own subtest dance in each test so that's good18:47
TheJuliano declared "support" past py3.10 for ddt, looks like maybe inactivity maintenance wise :\18:47
clifhrm, it does already exist in ironic's test-requirements.txt at least...18:48
TheJuliaYeah, we did a bunch of RBAC testing using it18:48
clifcardoe: the other thing about StrEnum is that py310 doesn't support it18:49
cardoeWell unless you specifically need to work with it as a str then a regular enum would work too.18:50
cardoeI just saw you using match so that helps make it exhaustive for tests and analysis.18:51
cardoeI believe ddt is the recommended thing by the OpenStack testing tools.18:52
cardoeI'm personally a pytest fan so I don't keep up with the unittest helpers that are out there.18:52
JayFddt is what nova asked me to rewrite some tests in for a follow-up19:05
JayFso if it's unsupported it's at least everyones' problem not just ours 19:05
clifwe can support it :)19:06
clifpytest is good ime19:07
JayFthere's already an ML thread about people trying to paint their test-shed that color19:08
cardoeThat's why I suggested ddt cause I saw nova suggest that to JayF 19:09
TheJuliaddt needs a new release, basically, I'm opening an issue on their GH regarding this19:14
opendevreviewMerged openstack/networking-generic-switch master: tox: Enable codespell by default and limit it to relevant files  https://review.opendev.org/c/openstack/networking-generic-switch/+/95815421:23
opendevreviewClif Houck proposed openstack/ironic master: Generate network plan based on trait based networking config  https://review.opendev.org/c/openstack/ironic/+/96489521:28
opendevreviewClif Houck proposed openstack/ironic master: Trait Based Networking Simulator  https://review.opendev.org/c/openstack/ironic/+/96620221:28
opendevreviewClif Houck proposed openstack/ironic master: WIP: Add configuration options for trait based networking  https://review.opendev.org/c/openstack/ironic/+/96805421:28
opendevreviewClif Houck proposed openstack/ironic master: Generate network plan based on trait based networking config  https://review.opendev.org/c/openstack/ironic/+/96489522:24
opendevreviewClif Houck proposed openstack/ironic master: Trait Based Networking Simulator  https://review.opendev.org/c/openstack/ironic/+/96620222:24
opendevreviewClif Houck proposed openstack/ironic master: WIP: Add configuration options for trait based networking  https://review.opendev.org/c/openstack/ironic/+/96805422:25

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