| TheJulia | by removed, changed to another conductor group right? | 00:00 |
|---|---|---|
| NobodyCam | yes, in a different region | 00:00 |
| TheJulia | conductor_group? | 00:00 |
| NobodyCam | I am migrating nodes from ussuri to antelope | 00:00 |
| TheJulia | so the tl;dr is it is not until 2024.1 afaik that nova has a command to reset the ComputeNode Host records | 00:01 |
| TheJulia | for nova, it maps to the compute node's hostname, so sort of no solution there unless you manually want to just fix records | 00:02 |
| NobodyCam | oh yea! | 00:02 |
| NobodyCam | `openstack baremetal allocation create --node <NODE>` | 00:02 |
| TheJulia | you don't need to do that... | 00:03 |
| JayF | In fact because an allocation sets instance_uuid if you add an allocation for a node it will never be able to be deployed by Nova | 00:28 |
| NobodyCam | wait what | 00:30 |
| NobodyCam | even if the instance is deleted? | 00:30 |
| JayF | Adding an allocation implicitly sets an instance uuid on the node, until the allocation is removed | 00:39 |
| TheJulia | so you can create an allocation if the node was already deployed | 01:07 |
| TheJulia | the long story short is NobodyCam is doing a migration right now so not quite "run of the mill" processes | 01:08 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 09:04 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 09:09 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 09:52 |
| opendevreview | cid proposed openstack/networking-generic-switch master: WIP: Add VLAN and Port whitelist https://review.opendev.org/c/openstack/networking-generic-switch/+/888047 | 10:11 |
| opendevreview | cid proposed openstack/networking-generic-switch master: WIP: Add VLAN and Port whitelist https://review.opendev.org/c/openstack/networking-generic-switch/+/888047 | 10:34 |
| opendevreview | cid proposed openstack/ironic master: Flexible IPMI credential persistence method configuration https://review.opendev.org/c/openstack/ironic/+/917229 | 11:21 |
| iurygregory | good morning Ironic | 11:38 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 12:58 |
| cid | good afternoon/morning, iurygregory | 13:08 |
| iurygregory | cid, o/ | 13:09 |
| cid | iurygregory, have you got a minute. | 13:10 |
| iurygregory | cid, sure | 13:10 |
| cid | One of my changes that failed at the gate does not seem to be caused by the change (my observation, probably not the case), what do you think | 13:12 |
| cid | https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/913926 | 13:12 |
| iurygregory | looking | 13:12 |
| iurygregory | yeah, doesn't seem related since you pushed the changes on May 12 and the gate failed on May 15 | 13:14 |
| iurygregory | let me check the status for the jobs | 13:14 |
| cid | ok, by the way, I think 15th is when it was approved. So it failed to merge | 13:15 |
| iurygregory | yeah | 13:15 |
| iurygregory | so in 15th there was already a problem with the standalone job | 13:15 |
| iurygregory | and we still have the same problem going on | 13:15 |
| iurygregory | I just saw the same failure in TheJulia patch https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/920113 | 13:16 |
| iurygregory | I would avoid recheck since the job seems broken =( | 13:16 |
| cid | owhk, seems like a job | 13:16 |
| TheJulia | good morning | 13:16 |
| cid | *problem | 13:17 |
| cid | good morning, TheJulia | 13:17 |
| cid | thanks, urygregory | 13:34 |
| TheJulia | JayF: so I think https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 really just needs a reno :) | 13:39 |
| TheJulia | And well, the commit message amended | 13:39 |
| opendevreview | Jake Hutchinson proposed openstack/bifrost master: Bifrost NTP configuration https://review.opendev.org/c/openstack/bifrost/+/895691 | 13:41 |
| JayF | cid: I think that IPA patch that Julia just linked back to me is the fix for that test you're fighting with | 13:44 |
| JayF | I'll be at my desk in 15 to 30 minutes, I'll revise it and hopefully we can land it this morning | 13:45 |
| TheJulia | ++ | 13:45 |
| cid | o/ | 13:46 |
| TheJulia | Yeah, once that is merged, it should clear up the sporadic failures on the tempest jobs which will allow some test fixes to head towards landing | 13:50 |
| cid | That's good news :D | 13:53 |
| TheJulia | Yeah, it has been frustrating since in CI it has been sporatically failing and we have a specific test failure because the bug fix in the agent, when not fixed causes the agent to spin for a while | 14:01 |
| JayF | I'm hoping (I have little evidence yet) that this could have a significant impact on CI runtimes | 14:09 |
| JayF | if we're spinning 30-60s+ each IPA boot (in a VM-in-a-VM) due to this bug; that means we could gain back 5 minutes in some jobs | 14:09 |
| TheJulia | The log files are definitely smaller. | 14:12 |
| TheJulia | Actual entire job runtimes really don't seem to be be that different, but we really don't have the best data on that job | 14:15 |
| JayF | and today is the day I learn we have 0% test coverage for get_managers | 14:15 |
| TheJulia | 1-2 minutes faster than the fastest run | 14:15 |
| JayF | I went to look at the existing tests to adapt one for _detail and ... yeah nothing | 14:16 |
| TheJulia | It is in a semi-weird area to try and mock out | 14:16 |
| TheJulia | but yeah, we should have something | 14:16 |
| JayF | yeah, I can see that | 14:16 |
| JayF | oh, you'd have to mock out stevedore | 14:16 |
| JayF | that would've 100% been something we punted at the point in time get_managers was written | 14:17 |
| JayF | (aka the "get us a bare metal cloud in an impossible timeframe" portion of my contribution period here lol) | 14:17 |
| TheJulia | yeah | 14:17 |
| TheJulia | I'd sort of be okay if we didn't try to mock it all out, fwiw | 14:18 |
| JayF | https://opendev.org/openstack/ironic-python-agent/src/branch/master/ironic_python_agent/tests/unit/test_hardware.py#L94 | 14:18 |
| JayF | we have a test class that appears to be setup for it | 14:19 |
| JayF | with zero tests attached | 14:19 |
| JayF | and that class is not referenced anywhere else | 14:19 |
| * JayF going to timebox an attempt to make this work | 14:19 | |
| * JayF just learned a python trick | 14:55 | |
| JayF | type() can create on-the-fly subclasses for you. Very useful when mocking out entire HWM classes :D | 14:55 |
| JayF | I'll note that this means the tests I'm adding might not backport to earlier python versions | 15:02 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent master: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 | 15:10 |
| * JayF runs unit tests on py3.9 and crosses fingers | 15:10 | |
| JayF | I think type() doing fun stuff was added in 3.8 | 15:10 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent-builder master: Block glean-early.sh from running https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/920183 | 15:17 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent-builder master: Block cloud-init user-data loading https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/920184 | 15:17 |
| JayF | this is passing when run by itself, and failing when run with all the tests together :| | 15:25 |
| JayF | I think _global_managers is caching between tests, haha | 15:37 |
| JayF | I just made IPA tests a little slower but a lot more actual-test-y :D | 15:38 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent master: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 | 15:47 |
| JayF | that should be reviewable now | 15:47 |
| JayF | that was fun finding the fix for that test | 15:47 |
| opendevreview | Julia Kreger proposed openstack/ironic master: WIP Advanced vmedia deployment test ops https://review.opendev.org/c/openstack/ironic/+/898010 | 15:48 |
| TheJulia | LGTM, lets see what CI says | 15:53 |
| JayF | I'm going to get the first backport patch done too since it won't be clean | 15:53 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent stable/2024.1: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920136 | 15:53 |
| TheJulia | iurygregory: you around? :) | 15:53 |
| TheJulia | after the first backport, it shoudl be a clean cherry-pick | 15:54 |
| JayF | yep that's my thought | 15:55 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent stable/2024.1: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920136 | 16:01 |
| JayF | there's the clean cherry-pick, I won't put it back further until master lands, just wanted to enable others to push it back if they needed | 16:02 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent master: Enable voting for codespell CI https://review.opendev.org/c/openstack/ironic-python-agent/+/920192 | 16:03 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent master: Enable voting for codespell CI https://review.opendev.org/c/openstack/ironic-python-agent/+/920192 | 16:04 |
| iurygregory | TheJulia, I'm now | 16:25 |
| iurygregory | just finished my lunch =) | 16:25 |
| JayF | iurygregory: I suspect she wanted you to look at https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 | 16:30 |
| JayF | I know I do :D | 16:31 |
| TheJulia | Indeed | 16:31 |
| TheJulia | sorry, slightly distracted, I have to go sit down with an accountant/attorney regarding an estate matter in ~1.5 hours | 16:31 |
| iurygregory | JayF, sure, let me finish the tests I was running to update firmware in 2 machines, I will take a look today | 16:31 |
| iurygregory | TheJulia, good luck! | 16:32 |
| JayF | cool, we expect this IPA fix, when it lands to make CI more stable | 16:32 |
| TheJulia | iurygregory: thanks, it shoudl be fairly simple unless the accountant's office didn't listen to me | 16:32 |
| iurygregory | fingers crossed | 16:33 |
| iurygregory | today I've discovered that a Dell machine was complaining about missing Location header when trying to do the SimpleUpdate <facepalm> | 16:33 |
| TheJulia | on a post? | 16:34 |
| * TheJulia wonders if we need "bear metal awards" | 16:34 | |
| iurygregory | yeah | 16:35 |
| iurygregory | "Response to /redfish/v1/UpdateService/Actions/UpdateService.SimpleUpdate did not contain a Location header: sushy.exceptions.MissingHeaderError: Response to /redfish/v1/UpdateService/Actions/UpdateService.SimpleUpdate did not contain a Location header" | 16:36 |
| * TheJulia goes and double checks location does what she thinks it does | 16:36 | |
| TheJulia | "Hello, my name is Domingo Bear-Metal, you gave me a bad response...." | 16:38 |
| TheJulia | That doesn't make any sense... | 16:39 |
| iurygregory | Domingo is Sunday in Portuguese :D | 16:40 |
| TheJulia | TIL | 16:40 |
| -opendevstatus- NOTICE: There will be a short Gerrit outage while we update to the latest 3.8 release in preparation for next weeks 3.9 upgrade. | 17:01 | |
| dtantsur | folks, especially TheJulia and JayF, what do you think about rendering https://opendev.org/openstack/ironic/src/branch/master/redfish-interop-profiles/OpenStackIronicProfile.v1_0_0.json in a human-friendly form in our documentation? | 17:07 |
| dtantsur | I'm tasked with providing something we can give to partners that ask "what parts of Redfish do I need to work with you?" | 17:07 |
| JayF | I think that's what the profile is intended for; if you need to put some prose around it to satisfy your requirement, I don't have any opposition | 17:07 |
| JayF | I'd just say to ensure you datestamp it | 17:08 |
| dtantsur | I don't assume we can use it in its raw form, maybe a separate JSON document with annotations.. | 17:08 |
| dtantsur | datestamp? | 17:08 |
| JayF | so in 2030 someone isn't thinking the 2024-dated hardware requirements are accurate when we eventually forget to update it :D | 17:08 |
| dtantsur | :D | 17:08 |
| JayF | "as of Caracal; hardware working with Ironic should support []" | 17:08 |
| dtantsur | yeah, it's already outdated | 17:08 |
| TheJulia | I think *something* makes sense in term of human parsable prose | 17:08 |
| dtantsur | maybe we change ProfileVersion field to use OpenStack version instead of the current 1.0.0? | 17:09 |
| TheJulia | I suspect partners can't seem to wrap their head around the interop profiles for the dmtf tool if they are not already familiar with it | 17:09 |
| TheJulia | I've already encountered a few discussions over the past year where folks couldn't really wrap their brains around it | 17:09 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 17:09 |
| dtantsur | okay, thanks for the feedback! | 17:10 |
| dtantsur | (working on docs is my priority for the next several weeks) | 17:10 |
| JayF | maybe try and sync up with Reverbverbverb if so? | 17:12 |
| JayF | he might have enough of his research done to help point in a direction | 17:12 |
| JayF | (that's Dave, who is doing an ironic docs audit on behalf of upstream via GR-OSS) | 17:13 |
| dtantsur | ah, was about to ask :) | 17:13 |
| dtantsur | Mainly Metal3 docs for me, but I'll probably do a share of fixes/additions in the Ironic docs as needed | 17:13 |
| JayF | dtantsur: https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 might be interesting to you, we could use a core review and tbh you might just find it generally interesting | 17:14 |
| dtantsur | putting on my queue | 17:14 |
| TheJulia | stepping out to go into town | 17:20 |
| opendevreview | cid proposed openstack/ironic master: wip: Provision aarch64 fake-bare-metal-vms https://review.opendev.org/c/openstack/ironic/+/915441 | 17:42 |
| iurygregory | From what I checked in openstacksdk we are missing 1- https://docs.openstack.org/api-ref/baremetal/#history-of-nodes 2- https://docs.openstack.org/api-ref/baremetal/#node-firmware-nodes 3-https://docs.openstack.org/api-ref/baremetal/#attach-detach-virtual-media-nodes | 18:30 |
| iurygregory | does anyone know about the magic with openstack client? if we add things to openstacksdk do we need to do something in the python-openstackclient ? | 18:31 |
| iurygregory | Tomorrow I have a meeting with 5 mentees from CMU Qatar , so I wanted to be sure to not say something wrong :D | 18:32 |
| dtantsur | iurygregory: yes, openstackclient (or rather: our plugin) requires explicit code | 18:54 |
| dtantsur | furthermore, our OSC plugin does not quite use openstacksdk yet :( | 18:54 |
| dtantsur | realistically, that could be a project of its own | 18:54 |
| dtantsur | iurygregory: with attach/detach, please sync with tdomnescu, at some point he was going to look into them | 18:55 |
| * dtantsur blends into background again | 18:55 | |
| JayF | iurygregory: if you're looking for projects, I believe gouthamr and gtema are doing some stuff related to openapi specs for various services, and integrating them into repos. Probably some straightforward work there for Ironic if you need to dig something up | 19:07 |
| TheJulia | https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 is green \o/ | 19:14 |
| iurygregory | dtantsur, when you say our plugin you mean https://github.com/openstack/python-ironicclient/tree/master/ironicclient/osc/v1 ? | 19:15 |
| iurygregory | JayF, ack, this was the topic we had during the PTG if I recall? | 19:16 |
| JayF | TC PTG iirc | 19:16 |
| JayF | It all sorta mixes together for me at this point, the etherpad remembers :D | 19:16 |
| JayF | dtantsur: have you booted a machine with the arm64 images for ipa-b recently? | 19:36 |
| JayF | dtantsur: cid's patch has a really weird error; it could be caused by a bad image build -- basically kernel is unable to unpack initramfs, so it can't mount root | 19:36 |
| TheJulia | iurygregory: https://review.opendev.org/c/openstack/ironic-python-agent/+/920136 is also up in review | 19:47 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent stable/2023.2: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920226 | 20:22 |
| JayF | please don't land this one until the master version lands :P | 20:23 |
| JayF | lest we get a ಠ_ಠ from folks who care about stable policy | 20:24 |
| TheJulia | of course ;) | 20:37 |
| * TheJulia gets out the anti-ಠ_ಠ sunscreen | 20:37 | |
| JayF | you just need the dark glasses from MIB | 20:39 |
| opendevreview | Merged openstack/ironic-python-agent master: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920153 | 20:50 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent-builder master: Block glean-early.sh from running https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/920183 | 21:31 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent-builder master: Block cloud-init user-data loading https://review.opendev.org/c/openstack/ironic-python-agent-builder/+/920184 | 21:31 |
| opendevreview | Merged openstack/ironic-python-agent stable/2024.1: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920136 | 21:34 |
| opendevreview | Jay Faulkner proposed openstack/ironic-python-agent stable/2023.1: Call evaluate_hardware_support exactly once per hwm https://review.opendev.org/c/openstack/ironic-python-agent/+/920216 | 21:54 |
| JayF | we have to land https://review.opendev.org/c/openstack/ironic-python-agent/+/912979 before I can go back to xena | 21:58 |
| JayF | but otherwise I have backports in place back to yoga | 21:58 |
| JayF | review at your leisure o/ | 21:58 |
| iurygregory | things I love about hardware, seems like the problem with location header is only in R6525 , R640 works like a charm | 22:50 |
| * iurygregory takes note to see how to handle location header for firmware update ... | 22:51 | |
| TheJulia | iurygregory: AMD vs Intel? | 22:52 |
| iurygregory | TheJulia, let me check the processor in the machine.. | 22:52 |
| iurygregory | TheJulia, you are correct that they have AMD (R6525) and Intel (R640) | 22:55 |
| iurygregory | but is a bit crazy they are providing firmware with different redfish implementations I would say... | 22:56 |
| iurygregory | or maybe is not yet in R640... | 22:56 |
| TheJulia | yeah, check versions | 22:56 |
| iurygregory | I asked the QE to not attempt to upgrade to the 7.10.30.00 | 22:57 |
| iurygregory | now time to grab dinner before playing with the HPE machines and their timeouts... | 23:00 |
| TheJulia | The downstream report I have of the HPE issue I was talking about, they ahven't yet filed a bug yet nor gotten me logs | 23:01 |
| JayF | Use your powers of ESP to divine the problem remotely | 23:03 |
| TheJulia | The energy required to tunnel through space time to divine such problems is better leveraged for creating the technology of star trek | 23:04 |
| opendevreview | Julia Kreger proposed openstack/ironic-specs master: Intermediate networking service https://review.opendev.org/c/openstack/ironic-specs/+/916126 | 23:21 |
| TheJulia | Any thoughts on if we made an httpboot interface which just did the network dhcp flow for an ISO ? | 23:39 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!