| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: don't use paramiko's get_fingerprint (md5) https://review.opendev.org/c/openstack/networking-generic-switch/+/942160 | 00:26 |
|---|---|---|
| opendevreview | Julia Kreger proposed openstack/networking-generic-switch master: don't use paramiko's get_fingerprint (md5) https://review.opendev.org/c/openstack/networking-generic-switch/+/942160 | 00:47 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 01:36 |
| opendevreview | Merged openstack/ironic bugfix/26.0: Pass agent token to get command results https://review.opendev.org/c/openstack/ironic/+/942062 | 01:48 |
| cardoe | There’s no way to have some kind of nova flavor with properties that users cannot control? Just thinking we might want to | 01:53 |
| cardoe | Somehow let the flavor select the deploy interface. | 01:54 |
| cardoe | Uhhh. This is why I need to walk away when it’s late. It’s a flavor duh. | 02:17 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add ironic-novncproxy service https://review.opendev.org/c/openstack/ironic/+/939191 | 03:06 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add systemd provider for console containers https://review.opendev.org/c/openstack/ironic/+/941614 | 03:06 |
| opendevreview | Steve Baker proposed openstack/ironic master: Implement drivers redfish-graphical, fake-graphical https://review.opendev.org/c/openstack/ironic/+/941615 | 03:06 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add vnc-container image build https://review.opendev.org/c/openstack/ironic/+/942017 | 03:06 |
| cardoe | What the heck.... these tests... behaving differently between machines. | 03:44 |
| TheJulia | cardoe: sushy? | 04:16 |
| cardoe | Yeah | 04:16 |
| TheJulia | cardoe: I've too been thinking about a deploy interface switching based upon flavor | 04:16 |
| cardoe | Okay so not a crazy idea | 04:17 |
| TheJulia | no, maybe not that | 04:19 |
| TheJulia | maybe ironic can swap based upon a property | 04:19 |
| TheJulia | I'd have to look at the logic around instance_info field population from nova | 04:20 |
| TheJulia | Might need to be a plugin interface because it could also quickly become business logic | 04:20 |
| * TheJulia remembers the operators agreeing that they will never agree about traits from the Sydney Summit | 04:21 | |
| cardoe | It's pretty basic. | 04:28 |
| cardoe | Okay. Now I can get different results between runs | 04:28 |
| cardoe | This teardown between these oem extension loading isn't working. The dell module is appearing in there. | 04:37 |
| opendevreview | Merged openstack/ironic master: trivial: lock dnsmasq check to ubuntu https://review.opendev.org/c/openstack/ironic/+/941710 | 04:37 |
| cardoe | okay I need another pair of eyes... | 04:48 |
| cardoe | https://opendev.org/openstack/sushy/src/commit/d0752a5598170c0ad4358f154dad24ef9ef7b938/sushy/tests/unit/resources/oem/test_common.py#L108 | 04:48 |
| cardoe | https://opendev.org/openstack/sushy/src/commit/d0752a5598170c0ad4358f154dad24ef9ef7b938/sushy/tests/unit/resources/oem/test_common.py#L143 | 04:48 |
| cardoe | It's not the dell code but it's the tests and the mocks. | 04:49 |
| cardoe | So basically if ANYTHING provides "sushy.resources.manager.oems" it's loading that rather than calling the ExtensionManager_mock a 2nd time. | 04:50 |
| cardoe | Wait a minute... are we parallelizing these tests with stestr ? | 04:52 |
| cardoe | Yes.... I just threw a clearing of that global in setUp() and now a bunch of different tests fail. | 04:53 |
| cardoe | The only reason this all passes is because the only thing that sushy ships with internally is 1 extension in "sushy.resources.system.oems" called "contoso". Which is what the VERY first mock always provides. | 04:55 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 04:55 |
| cardoe | Installing a vendor called "bob" and writing a test that loads it causes the tests to fail. | 04:56 |
| cardoe | I dunno how to fix this. | 04:56 |
| cardoe | But that's the blocker on merging sushy-oem-idrac | 04:56 |
| opendevreview | Doug Goldstein proposed openstack/sushy master: migrate sushy_oem_idrac to sushy https://review.opendev.org/c/openstack/sushy/+/940557 | 05:16 |
| opendevreview | Doug Goldstein proposed openstack/sushy master: explicit oem extn names and fix parallel tests https://review.opendev.org/c/openstack/sushy/+/942166 | 05:16 |
| JayF | Look at the hardware manager mocking in IPA, assuming y'all are using stevedore to load OEM stuff | 05:17 |
| opendevreview | Verification of a change to openstack/ironic master failed: Add ironic-novncproxy service https://review.opendev.org/c/openstack/ironic/+/939191 | 06:38 |
| opendevreview | Merged openstack/ironic master: deprecate shellinabox https://review.opendev.org/c/openstack/ironic/+/942124 | 08:03 |
| rpittau | good morning ironic! o/ | 08:52 |
| Sandzwerg[m] | Morning ironic | 12:35 |
| cardoe | Well my yolo change to Sushy passed last night. | 14:03 |
| cardoe | JayF: good call. The whole issue is that the test relies on a global of loaded stevedore extensions. But can only ever have 1 which matches name wise with the one that Sushy ships with. As soon as you add more. Sushy loads them into that global when called and stevedore isn’t mocked out. Stevedore isn’t mocked in many other tests. So the tests passed since the real one had the same name as the mocked one. But now we had | 14:07 |
| cardoe | more. | 14:07 |
| cardoe | rpittau: the Sushy bits are ready for review. I need to know how to | 14:08 |
| cardoe | Make an install conflict on Sushy-oem-idrac | 14:08 |
| rpittau | cardoe: thanks! I will check that after the current meeting | 14:08 |
| TheJulia | good morning | 14:16 |
| * TheJulia attempts to wake up | 14:23 | |
| TheJulia | Looks like we need to take a closer look at https://zuul.opendev.org/t/openstack/builds?job_name=ironic-standalone-ipa-src&project=openstack/ironic-python-agent | 14:46 |
| opendevreview | Julia Kreger proposed openstack/ironic master: ci: focus ironic-tempest-bios-ipmi-direct-tinyipa https://review.opendev.org/c/openstack/ironic/+/942204 | 15:00 |
| TheJulia | ^^^ maybe presently has a 60% pass rate, that is horrible, so hopefully that will improve the odds for us | 15:01 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 15:03 |
| TheJulia | Another job trying to run all the tests: https://zuul.opendev.org/t/openstack/build/4ebcbddd7d314ab59cf035de7b8bd5e6 :( | 15:11 |
| TheJulia | the results are so long we're not getting real results | 15:11 |
| cardoe | JayF: speaking of IPA... skrobul and I are debugging something weird on a branch new IPA we built from debian. | 15:13 |
| cardoe | dhcp-all-interfaces is creating a rule to DHCP. But then systemd-networkd is also doing DHCP on that interface. | 15:14 |
| cardoe | and oddly with the dnsmasq backend (not neutron) it gets 2 IPs sometimes | 15:15 |
| cardoe | Do we have something to block between the two? | 15:18 |
| JayF | That sounds like either a minor dib bug or an error in your dib configuration | 15:21 |
| cardoe | https://github.com/rackerlabs/understack/blob/main/ironic-images/ipa-debian-bookworm.yaml that's what I'm using presently. | 15:25 |
| JayF | My suspicion is that with the new Debian release the defaults in Debian changed and no one updated dib. | 15:27 |
| cardoe | Thats | 15:28 |
| cardoe | That's what I was thinking. | 15:28 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - Raise SYS518 server.task_state != None https://review.opendev.org/c/openstack/sushy-tools/+/942209 | 15:37 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: OS vmedia - post rebuild ACTIVE + SHUTOFF is ok https://review.opendev.org/c/openstack/sushy-tools/+/942210 | 15:37 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get_boot_mode support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942211 | 15:37 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get/set_boot_image support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942212 | 15:37 |
| rpittau | cardoe: +2 on the sushy-oem-idrac migration, thanks! | 15:40 |
| opendevreview | Kaifeng Wang proposed openstack/ironic master: Remove gmr from the docs https://review.opendev.org/c/openstack/ironic/+/942213 | 15:41 |
| cardoe | rpittau: how can I express a "this package should not be installed with any version of sushy-oem-idrac" | 15:42 |
| dtantsur | RPM has such a feature but I doubt pip does | 15:56 |
| rpittau | at least not qith requirements | 15:56 |
| rpittau | there might be something using pyproject.toml | 15:57 |
| cardoe | There are, but this is gonna be a pbr thing. | 15:57 |
| cardoe | https://review.opendev.org/c/openstack/diskimage-builder/+/942215 I dunno. Something like that. I feel like if ya got systemd-networkd that's setup for DHCP then dhcp-all-interfaces should be a no-op. | 15:58 |
| rpittau | cardoe: ideally we should deprecate sushy-oem-idrac after we merge that | 15:59 |
| shermanm | https://www.irccloud.com/pastebin/xOJkpHcd/ | 16:01 |
| shermanm | crud, wrong button, editing... | 16:02 |
| shermanm | I got a timeout for `ironic-tempest-ipa-wholedisk-direct-tinyipa-multinode` on https://review.opendev.org/c/openstack/networking-generic-switch/+/941172, but not totally sure what to do about that | 16:03 |
| TheJulia | shermanm: so, yeah | 16:03 |
| TheJulia | I noticed similar failures earlier | 16:03 |
| TheJulia | I'm trying to figure out what is actually failing there, but it is also timing out | 16:03 |
| TheJulia | the job, itself, is trying to do everything under the sun | 16:04 |
| TheJulia | and that takes a while and can timeout | 16:04 |
| TheJulia | or fail for unrelated reasons | 16:04 |
| TheJulia | similarly, that resulted in https://review.opendev.org/c/openstack/ironic/+/942204 for a different job, haven't quite figured out what to do with the multinode version yet | 16:05 |
| shermanm | is this something where I should try `recheck`, or just hold off until multinode seems more stabilized? Or something else? | 16:06 |
| JayF | cardoe: we could potentially go the other direction, release a final sushy-oem-drac that explicitly depends on sushy old enough to run it | 16:07 |
| cardoe | That's a good idea | 16:07 |
| TheJulia | shermanm: There are a few different variables at play, but mostly overall cloud load is one of those variables | 16:10 |
| TheJulia | shermanm: youc cnn recheck, but odds are it won't pass or may fail on a gate run, we need to figure out what is actually failing to limit scope of any changes | 16:10 |
| TheJulia | s/youc cnn/you can/ | 16:11 |
| opendevreview | cid proposed openstack/ironic master: Follow-up: Apply Inspection Rules https://review.opendev.org/c/openstack/ironic/+/942112 | 16:12 |
| opendevreview | cid proposed openstack/ironic master: API/Testing: Inspection rules migration https://review.opendev.org/c/openstack/ironic/+/939217 | 16:12 |
| shermanm | and separately, I'd be happy to take over https://review.opendev.org/c/openstack/networking-generic-switch/+/868318 , we're heavy users of both switch models, and I see that it's sat since 2023 | 16:14 |
| TheJulia | okay, so we're hitting logging issues because we're generating *too many* messages | 16:25 |
| TheJulia | the log content exceeds 4M | 16:26 |
| TheJulia | which is the arbitrary maximum in subunit | 16:26 |
| TheJulia | shermanm: I'd say that is entirely fair, most folks don't expect you to ask though ;) | 16:26 |
| TheJulia | well, most folks don't expect people to ask | 16:26 |
| TheJulia | :) | 16:26 |
| shermanm | thejulia: gotcha, just figured I'd ask while I'm learning the dev patterns | 16:28 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: ci: dial back check intervals https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/942220 | 16:42 |
| TheJulia | That... should help some of the lack of clarity | 16:42 |
| opendevreview | Julia Kreger proposed openstack/ironic master: ci: restrict multinode job to multinode tests https://review.opendev.org/c/openstack/ironic/+/942223 | 16:49 |
| TheJulia | shermanm: https://review.opendev.org/c/openstack/ironic/+/942223 | 16:51 |
| rpittau | good night! o/ | 16:53 |
| opendevreview | Merged openstack/ironic master: Remove gmr from the docs https://review.opendev.org/c/openstack/ironic/+/942213 | 17:34 |
| JayF | Noting for anyone in earshot: TC is still short one candidate | 17:49 |
| jrosser | what should i do as a maintainer of one of the community deployment tools (openstack-ansible) regarding the new ironic-novncproxy stuff? | 17:57 |
| jrosser | it seems really quite opinionated and by admission won't work with an already-containerised deployment tool | 17:58 |
| JayF | I assume stevebaker[m] will be providing some documentation on how to do it yourself; or if that's not already planned perhaps it should be :D | 17:58 |
| jrosser | i just observe that somehow it is backwards compared to almost all the rest of openstack | 17:59 |
| JayF | I agree that BMC console interfaces are extremely backwards ;) | 17:59 |
| jrosser | like python service -> whichever uwsgi or whatever you want -> venv if you want -> container (choose your own runtime) -> k8s if you want | 17:59 |
| jrosser | ^ this is "normal" kind of evolution from the source code repo | 18:00 |
| jrosser | but somehow it feels like this is running that path in reverse | 18:00 |
| JayF | It's a little bit weird because we're building a bridge/proxy that needs more control. I'm sure there's a path to making all parties happy but I don't fault Steve for trying to get something working as quickly as possible :) | 18:01 |
| JayF | I suggest a post to the mailing list about this or maybe even a bug against the feature talking about this and asking for a solution | 18:02 |
| jrosser | is there a spec? | 18:02 |
| JayF | my downstream will probably end up consuming whatever openstack-ansible builds for it anyway :D | 18:02 |
| jrosser | like we have a no-containers-whatsoever mode also that some people really like | 18:02 |
| JayF | https://review.opendev.org/c/openstack/ironic-specs/+/938526 / https://specs.openstack.org/openstack/ironic-specs/specs/not-implemented/graphical-console.html | 18:03 |
| jrosser | and i have no idea where to even start for that | 18:03 |
| stevebaker[m] | jrosser: hey | 18:03 |
| jrosser | hello o/ | 18:03 |
| stevebaker[m] | I'll be writing a container provider for a kubernetes operator which manages ironic (well, openshift) | 18:03 |
| stevebaker[m] | there is likely not a no-container option which is sufficiently secure. xvfb + a browser + x11vnc really needs a sandbox | 18:05 |
| JayF | There is an implied "we really wish the BMC interfaces didn't require us to engineer a whole $#%#$'n browser+vnc stack" in that statement ;) | 18:05 |
| JayF | and by "us" I mean steve :D | 18:06 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 18:07 |
| stevebaker[m] | I'll include docs on the requirement for building a custom console container provider, but I'd also be happy to advise specifically on openstack-ansible jrosser | 18:08 |
| jrosser | one that uses no containerswould be a great start | 18:08 |
| jrosser | as i say we have a "bare metal" mode (as would also be the case for the debian packaging people) | 18:09 |
| JayF | jrosser: I cannot see a future where Ironic would support/encourage the deployment of that console shim without signficiant sandboxing. It should not be colocated on a machine with *anything*. | 18:09 |
| JayF | jrosser: Especially if it's a bare-metal mode, and not even utilizing VMs, there's not really a safe path unless you are using some kinda sandboxing tech | 18:09 |
| jrosser | yes, so in openstack-ansible we have LXC available for that, which you can pretty much equate to a VM | 18:10 |
| JayF | I honestly dislike the containerization of the world, but this is a case where we *need* that featureset; it's not being made for convienience | 18:10 |
| jrosser | but you are in a nested containers world if then we need podman (or other) inside that, and thats a world of painfulness | 18:10 |
| JayF | Yeah I think there's a path to making something like that work | 18:11 |
| jrosser | and i guess the question is if the console shims need isolating from each other, or from everything else | 18:11 |
| JayF | almost certainly will need to be isolated from each other, but Steve would know for sure | 18:12 |
| stevebaker[m] | both really | 18:12 |
| JayF | ^ is what I meant | 18:12 |
| jrosser | anyway - i was just kind of surprised to see that all the community deploy tools will struggle | 18:12 |
| JayF | jrosser: just know this isn't like, Ironic trying to make operational decisions so much as seeing this as basically the only route to meeting the requirements with the "raw material" the BMCs give up | 18:13 |
| jrosser | perhaps less so the k8s based ones, but thats really not my domain of knowledge at all | 18:13 |
| jrosser | cool ok | 18:13 |
| JayF | iLO basically you can't see the console in any kinda native/nonproxied way without ALSO getting access to the full BMC control panel | 18:13 |
| JayF | and that's before we even get to the "needs to support read only" requirement | 18:14 |
| TheJulia | afaik, iLO has a special console only url | 18:14 |
| stevebaker[m] | it does! | 18:14 |
| JayF | Interesting, that's not what I took away from reading the spec and the comments on it | 18:14 |
| JayF | but still we'd need it for read-only | 18:14 |
| stevebaker[m] | I might have a chance to implement read-only today | 18:15 |
| JayF | that feature would be awesome to see a demo of | 18:15 |
| JayF | even if just a screenrecording | 18:15 |
| stevebaker[m] | I'll definitely be putting together a recorded demo at some point | 18:16 |
| TheJulia | always fun when I have a quick meeting and it turns into rfes | 19:04 |
| TheJulia | Going back to what jrosser should do regarding openstack-ansible, I think he is doing it, communicating which is the first step. The reality is opinionated models are also super difficult to deal with and there is a two way street of communication required. The next step is the hardest, wait for things to reach a minimum viable usable state :) | 19:11 |
| JayF | kolla-ansible probably should have a heads up too | 19:11 |
| TheJulia | "heads up, we're adding a new service. Don't expect it to be usable 'right now', its going to take time to polish it, go read x,y,z" | 19:12 |
| opendevreview | Verification of a change to openstack/ironic master failed: Add ironic-novncproxy service https://review.opendev.org/c/openstack/ironic/+/939191 | 19:14 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get_boot_mode support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942211 | 19:28 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get/set_boot_image support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942212 | 19:28 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get_boot_mode support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942211 | 19:55 |
| opendevreview | Harald Jensås proposed openstack/sushy-tools master: Openstack - get/set_boot_image support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942212 | 19:55 |
| stevebaker[m] | jrosser: the general approach in all cases to start a console will be: conductor writes out a file describing desired state then invokes some external tool which is responsible for making it happen. External tool starts the console proxy and returns an ip & port for the VNC connection. If the deployment tool has no always-running external tool to react to this then a lightweight service will need to be written | 20:11 |
| jrosser | i left a couple of after-the-fact comments on the spec | 20:14 |
| JayF | (lightweight service could even be a triggered systemd service, I suspect) | 20:15 |
| JayF | e.g. ConditionPathExists=/opt/ironic/do-the-thing ; Exec=lxc vnc-proxy-container | 20:15 |
| JayF | (as you can tell, I'm an expert on running lxc containers /s ) | 20:15 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add ironic-novncproxy service https://review.opendev.org/c/openstack/ironic/+/939191 | 20:18 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add systemd provider for console containers https://review.opendev.org/c/openstack/ironic/+/941614 | 20:18 |
| opendevreview | Steve Baker proposed openstack/ironic master: Implement drivers redfish-graphical, fake-graphical https://review.opendev.org/c/openstack/ironic/+/941615 | 20:18 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add vnc-container image build https://review.opendev.org/c/openstack/ironic/+/942017 | 20:18 |
| frickler | TheJulia: the multitenancy docs cleanup has left some weird references, like "See <no title> ..." in https://docs.openstack.org/ironic/latest/install/enrollment.html . are there plans already to further clean this up? | 20:25 |
| JayF | frickler: TheJulia: assuming it's not already in a followup, if someone can document this in a bug, we have folks who will solve those low-hanging-fruit now | 20:27 |
| TheJulia | yeah, seems like might just be a syntax issue | 20:27 |
| TheJulia | But not one I'm aware of (at least before) and looking at right now | 20:28 |
| frickler | ok, I'll do a quick bug report for reference | 20:30 |
| frickler | https://bugs.launchpad.net/ironic/+bug/2098895 | 20:33 |
| JayF | Thanks; Julia if you don't get it fixed let me know and I'll see if I can get someone to take a look | 20:38 |
| stevebaker[m] | jrosser: I replied to your comments | 20:40 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 20:42 |
| opendevreview | Verification of a change to openstack/ironic-python-agent master failed: Lockout agent command results if a token is received https://review.opendev.org/c/openstack/ironic-python-agent/+/942010 | 20:43 |
| TheJulia | JayF: replied on https://review.opendev.org/c/openstack/networking-generic-switch/+/942160 | 20:59 |
| TheJulia | Any chance I can get another reviewer to take a look at https://review.opendev.org/c/openstack/ironic/+/942204 | 21:00 |
| opendevreview | Merged openstack/sushy-tools master: Openstack - Raise SYS518 server.task_state != None https://review.opendev.org/c/openstack/sushy-tools/+/942209 | 21:01 |
| opendevreview | cid proposed openstack/ironic master: Follow-up: Apply Inspection Rules https://review.opendev.org/c/openstack/ironic/+/942112 | 21:01 |
| opendevreview | cid proposed openstack/ironic master: API/Testing: Inspection rules migration https://review.opendev.org/c/openstack/ironic/+/939217 | 21:01 |
| JayF | TheJulia: replied back, I'm still wondering in the fog'o'code | 21:02 |
| opendevreview | Verification of a change to openstack/ironic master failed: Fix devstack plugin for centos https://review.opendev.org/c/openstack/ironic/+/941720 | 21:46 |
| TheJulia | JayF: now worries, replied again :) | 21:53 |
| JayF | TheJulia: I think my impressions of how this /actually works/ have been muddied by eventlet being so ornery if you import **anything** it patches beforehand | 21:54 |
| JayF | I +2'd it but honestly I still don't fully understand why what you did is enough; but I also trust you have actually tested FIPS and I can't tell you what it stands for ;) | 21:55 |
| JayF | federal something something no-md5, right? /s | 21:55 |
| TheJulia | Federal Information Protection Standards | 21:59 |
| TheJulia | not that NIST will continue to exist much longer | 21:59 |
| TheJulia | but that is a different lolsob | 21:59 |
| TheJulia | yeah, its sort of a similar concept and eventlet has to do it so early on because it knows what it patches | 21:59 |
| TheJulia | and it manages all that too | 22:00 |
| TheJulia | (to hide the horrible details) | 22:00 |
| TheJulia | anyone else finding the formatting changes now think spurious tabs exist and then it reformats things actually incorrectly? | 22:01 |
| TheJulia | like, I checked them for tabs already | 22:01 |
| TheJulia | unless vi is actually vigor and it is trying to help me with all the evil of clippy | 22:02 |
| JayF | I havedn't seen that. | 22:11 |
| TheJulia | It must be tabs sneaking in anyway | 22:12 |
| opendevreview | Julia Kreger proposed openstack/ironic master: WIP: hooking in an external network simulator https://review.opendev.org/c/openstack/ironic/+/942298 | 22:13 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add systemd provider for console containers https://review.opendev.org/c/openstack/ironic/+/941614 | 22:29 |
| opendevreview | Steve Baker proposed openstack/ironic master: Implement drivers redfish-graphical, fake-graphical https://review.opendev.org/c/openstack/ironic/+/941615 | 22:29 |
| opendevreview | Steve Baker proposed openstack/ironic master: Add vnc-container image build https://review.opendev.org/c/openstack/ironic/+/942017 | 22:29 |
| opendevreview | Steve Baker proposed openstack/ironic master: Implement graphical console read-only support https://review.opendev.org/c/openstack/ironic/+/942300 | 22:29 |
| stevebaker[m] | JayF: read only support, boom | 22:29 |
| opendevreview | Merged openstack/ironic master: Mention removing codespell job in bugfix branches https://review.opendev.org/c/openstack/ironic/+/942067 | 22:30 |
| TheJulia | nice! | 22:31 |
| JayF | Julia beat me to the nice because I was too busy bragging about your good work in downstream slack :D | 22:31 |
| TheJulia | I'm going to call it a day shortly and head down the mountain. I'll review the change series in the morning, but ironic-tempest-plugin and ci job changes for ironic could use some reviews/rechecks to reduce the overall failure rate on some of the jobs and hopefully get us to a slightly happier CI state | 22:32 |
| JayF | I have been trying to finish the draft of my talk slides today and not rabbithole, but I still spent some time reviewing | 22:33 |
| JayF | and I saw a *lot* of ironic-standalone-* timeouts :( | 22:33 |
| JayF | all of them not "tinyipa build failed" but more shaped like "the ping check at the end failed" | 22:33 |
| JayF | I didn't dig for previously mentioned trying-to-do-something-else reasons :| | 22:34 |
| cardoe | I almost feel like my write up is gonna become a talk. | 22:34 |
| TheJulia | I saw one standalone job failure, and the standalone jobs do a ton of work | 22:35 |
| TheJulia | This is always an awful time for jobs anyhow | 22:35 |
| JayF | yeah, I assumed that was likely what was going on | 22:35 |
| JayF | and didn't recheck them for that reason | 22:35 |
| JayF | I think I have +2s on all your i-t-p improvements, yeah? | 22:36 |
| TheJulia | I think so yeah | 22:36 |
| TheJulia | largely those are just dialing back the aggressiveness so we might actuallly have a chance to get usable logs | 22:36 |
| TheJulia | to figure out what is failing | 22:36 |
| JayF | cid also has picked up the "make tinyipa builds less breaky" ticket I filed; I think we're mainly going to break out the download stage from the build stage so we can trivially add retries to the downloads | 22:37 |
| JayF | that should help some of hte failures too | 22:37 |
| opendevreview | Satoshi Shirosaka proposed openstack/ironic master: Remve multinancy page https://review.opendev.org/c/openstack/ironic/+/942306 | 23:28 |
| opendevreview | Merged openstack/sushy-tools master: OS vmedia - post rebuild ACTIVE + SHUTOFF is ok https://review.opendev.org/c/openstack/sushy-tools/+/942210 | 23:34 |
| opendevreview | Merged openstack/ironic-python-agent master: oci: permit an 'unknown' but valid image https://review.opendev.org/c/openstack/ironic-python-agent/+/941494 | 23:36 |
| opendevreview | Satoshi Shirosaka proposed openstack/ironic master: Remve multinancy page https://review.opendev.org/c/openstack/ironic/+/942306 | 23:40 |
| opendevreview | Satoshi Shirosaka proposed openstack/ironic master: Remve multinancy page https://review.opendev.org/c/openstack/ironic/+/942306 | 23:41 |
| opendevreview | Merged openstack/sushy-tools master: Openstack - get_boot_mode support volume boot https://review.opendev.org/c/openstack/sushy-tools/+/942211 | 23:54 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!