Wednesday, 2026-08-05

opendevreviewMerged openstack/ironic stable/2026.1: Portgroup shard filter bypasses scope to project  https://review.opendev.org/c/openstack/ironic/+/99976200:25
opendevreviewMerged openstack/ironic bugfix/37.0: Portgroup shard filter bypasses scope to project  https://review.opendev.org/c/openstack/ironic/+/99976400:25
opendevreviewMerged openstack/ironic bugfix/38.0: Portgroup shard filter bypasses scope to project  https://review.opendev.org/c/openstack/ironic/+/99976300:25
opendevreviewVerification of a change to openstack/ironic bugfix/34.0 failed: Portgroup shard filter bypasses scope to project  https://review.opendev.org/c/openstack/ironic/+/99976500:25
rpittaudtantsur: looks like the fix made BMO job happy :)06:42
dtantsurrpittau: nice! have you checked the logs? how much time do we lose in the end?07:31
rpittauI haven't checked the logs yet07:36
dtantsur(the revert of autodetect did not help, so your patch is all we have, I guess)07:37
rpittaufingers crossed07:38
dtantsurRetry after 7 seconds, not too bad07:40
dtantsurrpittau: can you address a small comment in the PR please? I don't want someone to remove the fix assuming it's accidental07:40
rpittauyes, I was going to do add that now actually07:41
dtantsurThanks for looking into the issue! It was a wild ride07:42
dtantsuronce you're done, anything blocking W+1 of https://review.opendev.org/c/openstack/ironic/+/999718?07:44
rpittaunope, I just didn't see Doug already +2ed that :D07:44
rpittauadded +W07:44
dtantsurthx!07:46
opendevreviewRiccardo Pittau proposed openstack/ironic master: Fix fast-track deployment failure with autodetect deploy interface  https://review.opendev.org/c/openstack/ironic/+/99969907:50
rpittaucomment added07:50
dtantsurthx!07:57
dtantsurhttps://review.opendev.org/c/openstack/ironic/+/997998 could you some input as well07:57
opendevreviewMerged openstack/ironic master: Fix Redfish inspection without Storage API  https://review.opendev.org/c/openstack/ironic/+/99971808:37
opendevreviewYorick proposed openstack/sushy master: system: guard HttpBootUri writes against iDRAC 10 removal  https://review.opendev.org/c/openstack/sushy/+/99974309:24
blanson[m]Hello ironic ! I have a bit of an issue using the hardware raid feature with redfish on HPE DL365 Gen11 with megaraid controllers. These controllers do not accept a size parameter and will throw iLO.2.30.UnsupportedOperation. I did the mess of a fix locally that just retries without the parameter, but I'm wondering if you'd have a better idea that I could put upstream.10:31
dtantsurblanson[m]: hmm, do we send the size unconditionally or is it based on API input?10:59
opendevreviewJacob Anders proposed openstack/ironic master: Refactor firmware update SimpleUpdate submission and error dispatch  https://review.opendev.org/c/openstack/ironic/+/99923011:21
opendevreviewJacob Anders proposed openstack/ironic master: Unify non-BMC firmware updates under batched state machine  https://review.opendev.org/c/openstack/ironic/+/99972011:21
opendevreviewJacob Anders proposed openstack/ironic master: Document batched firmware updates and add release note  https://review.opendev.org/c/openstack/ironic/+/99972111:21
opendevreviewDmitry Tantsur proposed openstack/ironic stable/2026.1: Fix Redfish inspection without Storage API  https://review.opendev.org/c/openstack/ironic/+/99980911:50
opendevreviewDmitry Tantsur proposed openstack/ironic bugfix/38.0: Fix Redfish inspection without Storage API  https://review.opendev.org/c/openstack/ironic/+/99981011:50
opendevreviewDmitry Tantsur proposed openstack/ironic bugfix/37.0: Fix Redfish inspection without Storage API  https://review.opendev.org/c/openstack/ironic/+/99981111:51
opendevreviewJacob Anders proposed openstack/ironic master: Unify non-BMC firmware updates under batched state machine  https://review.opendev.org/c/openstack/ironic/+/99972011:54
opendevreviewJacob Anders proposed openstack/ironic master: Document batched firmware updates and add release note  https://review.opendev.org/c/openstack/ironic/+/99972111:54
blanson[m]dtantsur: took me a little while to remember the whole thing but essentially: those megaraid controllers only support using the whole disk and don't take a size_gb value for the config. but using max we have to pass the exact disks, which we do not know because we use agent inspection for other reasons (getting switches lldp infos and such). I didn't find a way to surface the redfish disk infos in the inventory this way, so we11:59
blanson[m]have to pass the size_gb parameter and not MAX, which then breaks the megaraid controller that does not want it 11:59
blanson[m]I believe that's our whole situation11:59
dtantsurblanson[m]: would you be fine with max using all available disks?12:04
opendevreviewJacob Anders proposed openstack/ironic master: Unify non-BMC firmware updates under batched state machine  https://review.opendev.org/c/openstack/ironic/+/99972012:27
opendevreviewJacob Anders proposed openstack/ironic master: Document batched firmware updates and add release note  https://review.opendev.org/c/openstack/ironic/+/99972112:27
blanson[m]yeah, usually we only plug in 2 drives on these controllers and we expect it to use everything to make a raid 112:33
TheJuliawow, I just had someone ask me about using conductor groups downstream... for 10 nodes. :(13:36
TheJuliasounds like the validation really just needs to maybe retry "popping" out the size_gb parameter?13:38
TheJulia(this also feels semi-known why MAX even came into being, but it sounds like something changed13:38
TheJulialast call on https://review.opendev.org/c/openstack/ironic-specs/+/99765313:46
dtantsurblanson[m], TheJulia, I imagine we could relax the restriction and allow size=max without holder disks to mean "take all disks"13:55
dtantsurTheJulia: also, could you check https://review.opendev.org/c/openstack/ironic/+/999699 please? it's blocking BMO13:57
dtantsurI'm +1 on 997653 simply because I cannot find time for a full detailed read. But don't want for me in any case.13:58
TheJuliadtantsur: yeah, that is also reasonable I think14:01
TheJuliadtantsur: ack, okay14:01
TheJuliadtantsur: I'll make another cup of coffe and check that change out14:01
blanson[m]dtantsur: that was my initial though for it and then I was like "there is probbly a good reason that I don't know about for it not to be done this way", but if that sounds like a viable solution I'm all for it 14:04
dtantsurblanson[m]: I don't remember the reason, but we might have been simple cautious14:05
dtantsury'know, it's easier to relax a restriction than to introduce one..14:05
blanson[m]gotchu, thanks for looking into it ! I will try to find some time to propose something upstream then you can choose if it meets with what you had in mind :)14:07
dtantsurnice, thanks!14:09
TheJuliadtantsur: I approved and workflowed that change. It confused me greatly until I re-read the commit message.14:23
* dtantsur nods14:59
dtantsurFast track is a bit of a minefield14:59
dtantsurthank you!14:59
opendevreviewVerification of a change to openstack/ironic-specs master failed: Advance connectivity/driver modeling in NGS  https://review.opendev.org/c/openstack/ironic-specs/+/99765315:24
TheJuliadtantsur: unfortunately, yeah. See movie: dodgeball15:31
JayFlast call to get a review on [OSSA-2026-033] Ironic portgroup scoping  https://review.opendev.org/c/openstack/ossa/+/999767 before it's advisoried15:32
opendevreviewMerged openstack/ironic master: Fix fast-track deployment failure with autodetect deploy interface  https://review.opendev.org/c/openstack/ironic/+/99969915:34
opendevreviewOpenStack Release Bot proposed openstack/ironic-python-agent bugfix/12.0: Update .gitreview for bugfix/12.0  https://review.opendev.org/c/openstack/ironic-python-agent/+/99986015:35
opendevreviewDmitry Tantsur proposed openstack/ironic bugfix/38.0: Fix fast-track deployment failure with autodetect deploy interface  https://review.opendev.org/c/openstack/ironic/+/99987015:55
opendevreviewMerged openstack/ironic-python-agent bugfix/12.0: Update .gitreview for bugfix/12.0  https://review.opendev.org/c/openstack/ironic-python-agent/+/99986016:12
*** Unknown123 is now known as Mike--16:28
opendevreviewcid proposed openstack/ironic-python-agent master: Ignore the transport prefix when matching allowed containers  https://review.opendev.org/c/openstack/ironic-python-agent/+/99838116:42
opendevreviewcid proposed openstack/ironic-python-agent master: Drop the unused container_conf_file option  https://review.opendev.org/c/openstack/ironic-python-agent/+/99838216:42
opendevreviewcid proposed openstack/ironic-python-agent master: Rename container_clean_step to generic_container_step  https://review.opendev.org/c/openstack/ironic-python-agent/+/99838316:42
opendevreviewcid proposed openstack/ironic-python-agent master: Verify registry TLS by default  https://review.opendev.org/c/openstack/ironic-python-agent/+/99972516:42
opendevreviewMerged openstack/ironic-specs master: Advance connectivity/driver modeling in NGS  https://review.opendev.org/c/openstack/ironic-specs/+/99765317:48
opendevreviewSerhii Ivanov proposed openstack/sushy-tools master: Add empty BootOptions collection stub and Boot navigation links  https://review.opendev.org/c/openstack/sushy-tools/+/99191018:54
-opendevstatus- NOTICE: The Gerrit service on review.opendev.org will be offline momentarily at 20:00 UTC (an hour from now) while we restart for a patch upgrade, but should return within a few minutes.19:02
-opendevstatus- NOTICE: The Gerrit service on review.opendev.org will not be restarted at 20:00 UTC after all, due to unanticipated image build regressions, and will be rescheduled to a later time/date.19:13
opendevreviewJulia Kreger proposed openstack/ironic master: Add [api] enable_ramdisk_endpoints config option  https://review.opendev.org/c/openstack/ironic/+/99989719:44
TheJuliaJayF: ^^^ may be of interest to you19:44
JayFYou gonna write up the OSSN to pair with it?19:51
TheJuliaI guess I can if you want19:58
JayFI just don't wanna dupe work19:59
JayF90% of the work for the OSSN will be copy+paste your doc edits into a different format lol19:59
JayFjust let me know if you want me to knock it out19:59
JayFare we gonna backport 999897? 20:00
JayFTechnically it's a feature but I think it's easily justifible20:00
TheJuliayeah, that was why I was going to raise it20:01
TheJuliato your attention at least20:01
TheJuliaWe can, I don't see why not20:01
opendevreviewAdam McArthur proposed openstack/ironic-specs master: Proposal for a Bulk Node Operations API  https://review.opendev.org/c/openstack/ironic-specs/+/98318320:01
TheJuliaYeah, its a justifable backport where we just change it to a "fix" ;)20:01
cardoewho was it that picked up openstack-exporter?20:16
JayFadamcarthur5 from my team20:19
JayFHe's in the channel #openstack-exporter 20:19
opendevreviewSerhii Ivanov proposed openstack/sushy-tools master: Add one-time boot via libvirt reboot listener (SUSHY_EMULATOR_BOOT_ONCE)  https://review.opendev.org/c/openstack/sushy-tools/+/99191221:05
opendevreviewJulia Kreger proposed openstack/ironic master: Add image_server_auth_hosts to restrict credential scope  https://review.opendev.org/c/openstack/ironic/+/99974421:10
opendevreviewJulia Kreger proposed openstack/ironic master: Change image_server_auth_permit_unknown_hosts default to False  https://review.opendev.org/c/openstack/ironic/+/99990721:10
TheJuliacalling it a day folks21:11
JayFo/21:11
cardoeTheJulia: but there's so much more yaks that need shaving!21:20
TheJuliacardoe: spending half my day trying to figure out why I had an email saying "you have $4017 in 17 outstanding expenses"... meaning migraine21:43
TheJuliaso yeah, the yak's are limited today21:43
cardoeI mean that sounds like yaks of a different color21:46
TheJuliabusiness process automation yaks which required very large electric shavers [of doom]21:50
cardoeWould you guys be against me removing some behavior in inspection rules to make it more clear/declarative?21:51
cardoeRight now you can pass the values as a list OR a dict.21:51
cardoeSo if you have an operation "eq" which is defined as eq(arg1, arg2). You can pass.. {"args": {"values": [1, 2]}} or {"args": [1, 2]}21:53
cardoeBut here's the FUN part21:53
cardoe{"args": {"values": ["blue", "blue"]}} or {"args": [["blue", "blue"]]}21:54
cardoenotice the double / nested list?21:54
cardoeAnd my examples suck cause the string one is the case where it returns true21:55
cardoeCause {"args": ["blue", "blue"]} would return False 21:55
cardoeBecause it compares "b" against "l"21:56
cardoeJust feels like JayF's gonna be writing another OSSN in the future with so many different syntaxes and magical data marshalling.21:57
* TheJulia twitches and falls over21:58
cardoeI'd change it to require the "values" key always.21:58
TheJuliafile my thoughts under "just don't break things"21:58
cardoeI mean I suspect I'm the one using these things the most.21:58
TheJuliaand "keep things working in the support matrix across services"21:58
TheJuliaeh, your doing so quite aggressively, that is for sure21:59
TheJulia(and there is nothing wrong with that)21:59
JayF>  file my thoughts under "just don't break things"21:59
JayFliterally exactly what I was thinking21:59
JayFthere's so much customization and inconsistency here fixing it to be sensible while navigating stability is ... hard22:00
cardoeBasically if input isn't a dict, we make a dict. We look at the operation and grab the fields that the operation takes and make those dictionary keys.22:00
cardoeWe walk the list and assign the list values to each input parameter22:01
cardoeBut eq is actually eq(*args, optional_arg=None) syntax22:02
cardoeSo we further check if the FIRST argument isn't enumerable then pass it to *args but if it is enumerable set the parameters differently.22:02
cardoeBasically it's a nightmare of un-safety22:03
TheJuliaso, I'm good with making consistent and cleaning up, but again, version matrix of agents to ironic services is the thing to navigate22:04
TheJuliathat is if stuff needs to chagne22:04
TheJuliaand if rules need to change themselves, a path22:04
cardoeI'm 100% gonna locally patch it to get rid of all the hammering the data into the correct shape and just require people pass data correctly22:04
TheJuliaand wwarning/error/etc22:04
TheJuliathe upgrade check being the path there22:04
cardoeWell here's the even funner case since the variable gets interpolated as well... the behavior can change based on the inspected data of the hardware if you have an interpolated variable first.22:05
cardoeWe only shipped this in 2026.122:06
cardoeSo it just feels better to tighten it up now and make an upgrade note.22:06
opendevreviewDoug Goldstein proposed openstack/ironic master: Interpolate values for comparison operators in inspection rules  https://review.opendev.org/c/openstack/ironic/+/99991022:09
TheJulia++++22:16

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