Friday, 2025-03-28

freemanboss[m]<Richard-Akin> "masghar: yeah, I haven't been..." <- I think you're missing things out that's why.05:25
freemanboss[m]You can either choose to run ironic using bifrost in a test mode or prod mode (and this is when you've your physical server that you want to run it on). For now we are advised to run it in a test mode to understand how it works better and experiment with it05:25
Amarachi_OGood morning Ironic!06:08
freemanboss[m]Good morning 🌄06:43
opendevreviewMerged openstack/ironic stable/2025.1: [Trivial] Fix typo of exception error message  https://review.opendev.org/c/openstack/ironic/+/94574006:55
rpittaugood morning ironic! happy friday! o/08:26
opendevreviewRiccardo Pittau proposed openstack/ironic stable/2025.1: [2025.1 only] update devstack config  https://review.opendev.org/c/openstack/ironic/+/94579308:59
ayo_Good morning rpittau :)09:01
vsaienkoTheJulia, JayF please add to your review queue https://review.opendev.org/c/openstack/networking-generic-switch/+/945708 - adds a new job to test portgroups, and dependent patches https://review.opendev.org/c/openstack/ironic/+/940611 allows to setup environment with PG and https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/940678 adds validation for network data related to PG09:01
opendevreviewHarald JensĂĄs proposed openstack/sushy-tools master: os-vmedia: Add option to delay rebuild on eject  https://review.opendev.org/c/openstack/sushy-tools/+/94580010:06
opendevreviewMerged openstack/ironic stable/2024.1: Make floppy images more floppy  https://review.opendev.org/c/openstack/ironic/+/94504610:09
dtantsurTheJulia, JayF, let's imagine I find someone to work on standalone switch management. Do you think we can give them enough space and provide enough guidance to make it happen in the near future?11:27
dtantsurI'm trying to stay honest with expecations11:27
opendevreviewDoug Goldstein proposed openstack/networking-baremetal master: fix spelling mistakes  https://review.opendev.org/c/openstack/networking-baremetal/+/94581812:22
opendevreviewDoug Goldstein proposed openstack/networking-baremetal master: fix awkward logic that the linter didn't like  https://review.opendev.org/c/openstack/networking-baremetal/+/94581912:22
opendevreviewDoug Goldstein proposed openstack/networking-baremetal master: fix sphinx-lint issues  https://review.opendev.org/c/openstack/networking-baremetal/+/94582012:22
opendevreviewDoug Goldstein proposed openstack/networking-baremetal master: add pre-commit and adjust tox to utilize it  https://review.opendev.org/c/openstack/networking-baremetal/+/94582112:22
freemanboss[m]<rpittau> "good morning ironic! happy..." <- Happy Friday 12:43
JayFdtantsur: I will be taking a 6-week sabbatical in the middle of this cycle. So I am keeping my commitments the cycle to a minimum13:09
JayFTbh, even if not that I'm not sure I would be the best person to help someone down that path anyway13:09
dtantsurnoted, sounds fun :)13:09
dtantsurYeah, I'm looking less for personal commitments more for your takes on whether our community can absorb that13:09
TheJuliadtantsur: I'm semi-hoping we can frame things sufficient I might be able to spend time on it, but dunno. I don't know about anything at this point given the broad uncertinty13:19
dtantsurright13:23
TheJuliaWell, I do know immediate demands, and the only way I ensure appropriate tie-in is to be involved upfront13:25
TheJuliaanyway, first meeting of the day in a few minutes, coffeeeee to finish13:25
opendevreviewStephen Finucane proposed openstack/ironic master: conf: Add '[api] response_validation' option  https://review.opendev.org/c/openstack/ironic/+/94551513:57
opendevreviewStephen Finucane proposed openstack/ironic master: api: Add schema for allocations API (responses)  https://review.opendev.org/c/openstack/ironic/+/92892113:57
opendevreviewDmitry Tantsur proposed openstack/ironic master: Do not silence the actual error in prepare_instance_boot  https://review.opendev.org/c/openstack/ironic/+/94584414:52
dtantsurdid we have any reasons not to ^^^?14:52
dtantsurwe've just wasted plenty of time because of it14:52
rpittauI can't think of any15:04
UgglaHello ironic team, I have not seen any common topic between nova-ironic for the PTG. So I guess no cross team session is needed, but I'd like to confirm.15:05
rpittauUggla: hi! o/ we've discussed about topics during the last weeks and so far we have nothing with nova, I guess we'll ahve another discussion on Monday during the meeting to finalize the list, and then provide a better answer15:11
JayFWe're working on some network enhancement specs which may need some changes in the ironic nova virt driver; but I doubt we'll have details in time for a PTG cross-team15:12
JayFprobably something for next PTG if it's beefy enough for a PTG session15:12
Ugglarpittau, good. Just let me know if you need one. Same on us, the list of topics is not closed yet so I'll warn here if needed.15:13
rpittauperfect, thanks Uggla 15:14
Ugglarpittau, you're welcome.15:14
TheJuliaJayF: I glanced at the nova.virt.ironic driver code and pattern, the existing pattern matches what we would expect with the current idea on the table15:33
TheJulia(so as long as nova posts the flavor or requested flavor traits to the driver, we're good15:33
TheJulia)15:33
TheJulia(I didn't double check that, since I think you indicated it did happen now)15:36
cardoeUggla: so I've got some items around improving the nova-ironic interaction.15:53
cardoeThe error reporting is sub-optimal and that's what I think we could (and should) improve on.15:53
cardoeUggla: so like first patch I've got is... https://review.opendev.org/c/openstack/nova/+/942019 which needs to be backported now to 2025.1, 2024.2, 2024.115:54
cardoeI forget the bug off hand but nearly all build errors end up wrapped up in a block device error message even if it was a neutron network issue. So you've got to read the call stack to figure it out. But last error won't show that.15:57
cardoeThere's some in flight stuff to expose trunk ports in the metadata as well.15:58
Ugglacardoe, ok but I guess it does not required a cross team session. Am i wrong ?16:11
cardoeNo cross session necessary. I just figured I'd highlight there's some items around error handling / reporting that's rough right now.16:39
JayFTheJulia: I think the question was around ordering, and if we would need to change ordering around in how nova handles e.g. configdrive generation (or, as you proposed, move most of that code into ironic)16:47
JayFTheJulia: So I don't think we'll need surgery in the nova ironic driver, but we might need a patch or three16:47
opendevreviewSatoshi Shirosaka proposed openstack/ironic-python-agent master: WIP Implement manual cleaning for ContainerHardwareManager  https://review.opendev.org/c/openstack/ironic-python-agent/+/94586217:03
ayo_freemanboss[m]: after deploying, my state has been waiting for a call back for quite a while, I want to know if yours did the same17:16
freemanboss[m]ayo_: Call back? I don't understand 17:17
ayo_After enrolling and deploying17:17
JayFprovision_state: wait call-back17:17
ayo_I believe it’s supposed to show successful if the deployment worked17:18
ayo_My issue is my provisioning state has been on wait call-back for a while17:19
ayo_I just want to know if it normally takes a while before it works17:20
ayo_It’s been 10 minutes now17:20
ayo_Oops17:21
ayo_Never mind17:21
ayo_Just showed active, thank you17:21
ayo_I’m still a bit confused on the enroll and deploy line17:23
ayo_JayF: if you don’t mind me asking17:23
freemanboss[m]<ayo_> "My issue is my provisioning..." <- Just wait to see the final results.17:24
freemanboss[m]Immediately after the deployment is run it will be on deploying and I didn't check back on the end status 17:24
freemanboss[m]Since you mentioned it now I'll have to take note on the end status. But if it's on anything wait just wait a while if it'll change17:24
freemanboss[m]ayo_: Great you're welcome!17:25
freemanboss[m]ayo_: What's the confusion? You can ask if I may be of help 17:25
freemanboss[m]* Just wait to see the final results.17:26
freemanboss[m]Immediately after the deployment is run it will be on deploying and I didn't check back on the end status17:26
freemanboss[m]Since you mentioned it now I'll have to take note on the end status. But if it's on anything wait just wait a while it'll change17:26
ayo_I understand using the “enroll” alongside the .json package problvides detail regarding a bare metal server17:26
freemanboss[m]ayo_: Yes?17:27
ayo_The deploy stage attaches OS on the enrolled server17:27
ayo_I still don’t understand where the details on the OS is predefined17:28
ayo_freemanboss[m]: ?17:30
freemanboss[m]<ayo_> "I still don’t understand where..." <- To understand things better one has to go deep into the code and it's a matter of time. My PC is off I'd have like to explore more so that I can explain this better. But so far actually I've what I've observed too17:33
freemanboss[m]Since it's a test mode it provisioned 2 servers (testvm1 and testvm2) from the actual server you're running Bifrost/ironic on. It then made ssh connections to the 2 provisioned servers and installed the OS which I think it's Ubuntu therein. 17:36
freemanboss[m]<ayo_> "I still don’t understand where..." <- I think it's defined in the inventory file. Can you send it here?17:36
ayo_Inventory.json file?17:43
opendevreviewJulia Kreger proposed openstack/ironic master: devstack: network simulator support for sonic  https://review.opendev.org/c/openstack/ironic/+/94572618:00
freemanboss[m]<ayo_> "Inventory.json file?" <- Yes18:33
freemanboss[m]Or nodes.py18:33
freemanboss[m]Nodes json18:33
opendevreviewMerged openstack/networking-generic-switch master: docs: add additional context around Cisco nxos usage  https://review.opendev.org/c/openstack/networking-generic-switch/+/94514518:42
opendevreviewJulia Kreger proposed openstack/ironic master: devstack: network simulator support for sonic  https://review.opendev.org/c/openstack/ironic/+/94572619:01
opendevreviewJulia Kreger proposed openstack/networking-generic-switch master: Adding notes for SONiC switches  https://review.opendev.org/c/openstack/networking-generic-switch/+/94588719:01
TheJuliaJayF: metadata generation is done after vif attachments get sent from what I saw, which is also done after instance_info is supposed to be populated, so as long as we wire things together that vif attachment handling logic triggers dynamic portgroup assembly, then metadata and config drive should all be good19:31
cardoeYeah it’s definitely done after vif attachment cause that’s how MAC addresses are correct. We should update the networking doc.19:41
opendevreviewSatoshi Shirosaka proposed openstack/ironic-python-agent master: WIP Implement manual cleaning for ContainerHardwareManager  https://review.opendev.org/c/openstack/ironic-python-agent/+/94586219:43
TheJuliacardoe: I'm not sure why you note update our networking doc, is there a doc which is wrong, the thread of discussion was related to dynamic grouping20:31
cardoeI was saying if there’s confusion.20:32
cardoeSorry. Didn’t come across correctly.20:33
TheJuliano worries20:40
kulsoomsfreebossman, ps_adavize, queensly: Hi everyone, if any of you have come across and resolved this error, please let me know https://imgur.com/a/lOzDPgw22:10
opendevreviewSatoshi Shirosaka proposed openstack/ironic-python-agent master: WIP Implement manual cleaning for ContainerHardwareManager  https://review.opendev.org/c/openstack/ironic-python-agent/+/94586222:40

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!