| opendevreview | Steve Baker proposed openstack/bifrost master: Use preferred URL for upper constraints https://review.opendev.org/c/openstack/bifrost/+/897877 | 02:27 |
|---|---|---|
| opendevreview | Steve Baker proposed openstack/bifrost stable/2023.2: Use preferred URL for upper constraints https://review.opendev.org/c/openstack/bifrost/+/897878 | 02:28 |
| opendevreview | Steve Baker proposed openstack/bifrost stable/2023.1: Use preferred URL for upper constraints https://review.opendev.org/c/openstack/bifrost/+/897879 | 02:36 |
| opendevreview | Steve Baker proposed openstack/bifrost stable/zed: Use preferred URL for upper constraints https://review.opendev.org/c/openstack/bifrost/+/897880 | 02:37 |
| stevebaker[m] | I'm assuming CI on stable bifrost branches are broken without the above changes, but I don't have any review data to confirm that | 02:38 |
| opendevreview | Merged openstack/ironic-python-agent master: Retry on checksum failures https://review.opendev.org/c/openstack/ironic-python-agent/+/897853 | 03:22 |
| opendevreview | Steve Baker proposed openstack/bifrost stable/2023.2: [DNM] check if stable CI is green https://review.opendev.org/c/openstack/bifrost/+/897917 | 03:27 |
| opendevreview | Steve Baker proposed openstack/bifrost stable/zed: Use preferred URL for upper constraints https://review.opendev.org/c/openstack/bifrost/+/897880 | 03:47 |
| stevebaker[m] | eh, it was a transient 404, those upper URLs are back to returning 301. I'll abandon the above changes | 03:54 |
| TheJulia | Wheeeee | 03:58 |
| stevebaker[m] | no, I figured it out, its wrong in 2023.2, one fix incoming | 03:59 |
| TheJulia | Ahh | 03:59 |
| * TheJulia tries to get some sleep | 03:59 | |
| opendevreview | Steve Baker proposed openstack/bifrost stable/2023.2: Fix BRANCH for install-deps.sh https://review.opendev.org/c/openstack/bifrost/+/897878 | 04:01 |
| * stevebaker[m] waves goodnight | 04:01 | |
| opendevreview | Verification of a change to openstack/ironic master failed: Refactor publishing images into a new module https://review.opendev.org/c/openstack/ironic/+/897675 | 04:17 |
| opendevreview | Verification of a change to openstack/ironic master failed: Add inspection hooks https://review.opendev.org/c/openstack/ironic/+/896355 | 05:31 |
| rpittau | good morning ironic! o/ | 06:35 |
| opendevreview | Verification of a change to openstack/ironic master failed: Refactor publishing images into a new module https://review.opendev.org/c/openstack/ironic/+/897675 | 07:27 |
| opendevreview | Verification of a change to openstack/ironic master failed: Add inspection hooks https://review.opendev.org/c/openstack/ironic/+/896355 | 07:51 |
| dtantsur | wow, CI is really unwell Oo | 08:09 |
| rpittau | nah just arm64 fake | 08:10 |
| dtantsur | mmm? | 08:10 |
| dtantsur | ah, damn, you're right | 08:10 |
| rpittau | the latest 2 msgs are arm64 jobs succeding | 08:10 |
| rpittau | :D | 08:10 |
| rpittau | it fooled me too | 08:11 |
| dtantsur | I used to keep that in mind, but this time it got me | 08:11 |
| haozhou | Hello everyone, our proposal in openshift/enhancements (https://github.com/openshift/enhancements/pull/1395) has been merged. Within it, there's a section that requires changes to "ironic". Previously, it was mentioned that changes to this section would be discussed in the ironic community. We have a demo of the modified code. | 08:26 |
| haozhou | I'd like to ask, What's the best way to bring up this discussion? Should I create a story and then submit the code to the community? Or should we discuss it directly in this channel? | 08:26 |
| dtantsur | haozhou, hi and welcome! Please create an RFE in launchpad with a detailed design (just linking to openshift is not enough, please extract the parts relevant to ironic). | 08:29 |
| dtantsur | Then we can discuss it here or on the next weekly meeting | 08:29 |
| haozhou | ok, got it, thanks! | 08:31 |
| opendevreview | Merged openstack/ironic master: Add inspection hooks https://review.opendev.org/c/openstack/ironic/+/896355 | 09:47 |
| opendevreview | Merged openstack/ironic master: Refactor publishing images into a new module https://review.opendev.org/c/openstack/ironic/+/897675 | 09:53 |
| opendevreview | Merged openstack/ironic master: Extract generic image publishing code from image_utils https://review.opendev.org/c/openstack/ironic/+/897681 | 09:56 |
| dtantsur | masghar: I guess what is missing in terms of hooks is only the LLDP stuff? | 10:48 |
| dtantsur | so, https://review.opendev.org/c/openstack/ironic/+/892661 (now has conflicts) | 10:49 |
| iurygregory | good morning Ironic | 11:56 |
| iurygregory | dtantsur, fyi I've created https://bugs.launchpad.net/ironic/+bug/2039042 | 12:22 |
| iurygregory | should I create new downstream bugs or just track in the ones we have for the no suitable vmedia device? | 12:22 |
| dtantsur | I think the existing downstream one now FailedQA, so it can/should be reused? | 12:24 |
| iurygregory | yeah, that's what I was thinking | 12:29 |
| TheJulia | good morning | 13:15 |
| TheJulia | haozhou: Also, add an entry to https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting under RFE review | 13:16 |
| iurygregory | good morning TheJulia =) | 13:32 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent stable/2023.2: Retry on checksum failures https://review.opendev.org/c/openstack/ironic-python-agent/+/897961 | 13:33 |
| opendevreview | Julia Kreger proposed openstack/ironic-python-agent stable/2023.1: Retry on checksum failures https://review.opendev.org/c/openstack/ironic-python-agent/+/897950 | 13:43 |
| TheJulia | woot, we have a CI job running with OVN now | 14:04 |
| * TheJulia dances | 14:04 | |
| iurygregory | dtantsur, in case we don't pass the boot_device, shouldn't we try to eject media from all devices we find media inserted? thinking a bit here while I was adding tests to the eject media | 14:22 |
| dtantsur | TheJulia, congrats! | 14:22 |
| dtantsur | iurygregory, is it about the new API or the bug? | 14:22 |
| iurygregory | the bug | 14:22 |
| TheJulia | iurygregory: I would try to eject everything we don't know about/expect | 14:22 |
| TheJulia | personally | 14:23 |
| dtantsur | I"m worried about failing on something exotic, but maybe I'm overthinking it | 14:23 |
| TheJulia | "why is it attached? what did we do wrong?" are the questions I would ask, and then that would jump to "did someone do this outside of ironic? why? is this a security incident?!?!" | 14:23 |
| dtantsur | :D | 14:24 |
| dtantsur | yeah, maybe we should just unconditionally remove everything | 14:24 |
| TheJulia | Yeah, we manage the hardware, if people are doing things directly, we really want to know what/why/etc and also to drive our feedback loop | 14:25 |
| TheJulia | and one offs are a thing, just it likely got forgotten | 14:25 |
| * TheJulia has sure left forgotten vmedia config in her supermicro box | 14:25 | |
| TheJulia | JayF: low hanging fruit bug for ipa docs opened | 14:28 |
| TheJulia | in ironic... mainly just for visibility purposes | 14:29 |
| JayF | https://infosec.exchange/@tychotithonus/111213832445211051 for the recent curl socks vulnerability, should we put out an ossn advising IPA users to rebuild their ramdisk images? | 14:29 |
| dtantsur | do we use curl? or socks? | 14:30 |
| TheJulia | vulnerabilities, more than one?! | 14:30 |
| TheJulia | we *do* support socks for the agent | 14:30 |
| TheJulia | *but* hostnames longer than 255 characters should also be invalid, and I don't think we bounds check them anywhere | 14:30 |
| JayF | I don't know why I thought we used curl... | 14:30 |
| dtantsur | It's possible that we used to back when we had more bash | 14:31 |
| TheJulia | we use requests, which uses urllib3 | 14:31 |
| iurygregory | I'm really puzzled on how our code is working in this test https://github.com/openstack/ironic/blob/master/ironic/tests/unit/drivers/modules/redfish/test_boot.py#L1485 we call eject_vmedia that will call _eject_vmedia but we don't pass any boot_device.. so I don't think it would eject the media, or I'm reading things wrong O.o https://github.com/openstack/ironic/blob/master/ironic/drivers/modules/redfish/boot.py#L253 | 14:31 |
| JayF | I think somehow I was thinking IPA used it when we use it during the IPA build | 14:31 |
| TheJulia | ... which I thought uses libcurl | 14:31 |
| JayF | But that is a good point that there is value in us checking the hostname length | 14:31 |
| TheJulia | JayF: IPA used it *ages* ago back when write_image.sh existed | 14:31 |
| dtantsur | I thought urllib3 was written ~ from scratch | 14:32 |
| TheJulia | ... looking! | 14:32 |
| TheJulia | I know it has a "split me out a curl command" log feature | 14:33 |
| TheJulia | or maybe that is requests | 14:34 |
| TheJulia | it is clean | 14:34 |
| TheJulia | so yeah, no need to tell folks to rebuild their ramdisks unless they have code directly invoking curl and they are concerned about the risk, in the mean time it wouldn't be a bad thing to create and backport a bounds check anyway | 14:35 |
| JayF | So it sounds like no action for us then. Good stuff. | 14:35 |
| JayF | I'd check to see if urllib has a check before we add an additional | 14:35 |
| JayF | Honestly I would probably even try to propose it there if it doesn't have one that way we fix all of openstack and Python.... But I suspect it already checks | 14:36 |
| TheJulia | I think the longest host name I’ve seen in reality was 80 something characters and that was Steve’s lab | 14:36 |
| TheJulia | … where we found idracs can’t handle more than 64 characters | 14:36 |
| JayF | 😂 that's what you call being four times more secure than curl | 14:37 |
| TheJulia | We should regardless since hardware managers can grab parameters and it is third party plugin code at that point | 14:38 |
| JayF | I can see that argument. In reality, I personally never seen a third party hardware manager that made a network call. | 14:39 |
| TheJulia | That is a good point | 14:39 |
| JayF | Most integrations of that style were done as patches on the conductor because you would need to have credentials to do a network call | 14:39 |
| TheJulia | I could see grabbing artifacts, and I could see firmware urls in steps | 14:40 |
| TheJulia | But that is all python code | 14:40 |
| TheJulia | And if BMC vendors are using curl in their BMCs… we’ll… sigh?! | 14:41 |
| JayF | And in any event, the ability to get an oversized host name into any of those places would require the same kind of access that you require to run whatever ramdisk you want to run on the machine anyway | 14:41 |
| JayF | At least generally speaking ... I'm sure I can devise a bad design of a third party hardware manager that would violate that, but I would have to try | 14:42 |
| TheJulia | Yeah | 14:42 |
| * TheJulia tries to get her brain back to what she was going to start this morning | 14:45 | |
| opendevreview | Riccardo Pittau proposed openstack/ironic-python-agent bugfix/9.6: Retry on checksum failures https://review.opendev.org/c/openstack/ironic-python-agent/+/897968 | 14:53 |
| TheJulia | rpittau: fwiw, I'm not backporting that past 2023.1 unless we really see a need for it | 14:55 |
| rpittau | 9.6 is after 2023.1 | 14:56 |
| TheJulia | ack | 14:56 |
| rpittau | I'll skip 9.5 as it's almost expired | 14:57 |
| TheJulia | hmmmm dhcp-less, you need a tempest test | 15:04 |
| rpittau | good night! o/ | 16:07 |
| opendevreview | Verification of a change to openstack/sushy master failed: Fix wrong _get_registry logic in ResourceBase https://review.opendev.org/c/openstack/sushy/+/892261 | 16:40 |
| opendevreview | Verification of a change to openstack/ironic master failed: CI: Fix our internal MTU settings https://review.opendev.org/c/openstack/ironic/+/893112 | 16:41 |
| opendevreview | Verification of a change to openstack/ironic master failed: Introduce conductor touch while offline https://review.opendev.org/c/openstack/ironic/+/897375 | 17:00 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Make sure we eject media from DVD when CD is requested https://review.opendev.org/c/openstack/ironic/+/897989 | 17:05 |
| opendevreview | Verification of a change to openstack/ironic stable/yoga failed: redfish_address - wrap_ipv6 address https://review.opendev.org/c/openstack/ironic/+/896209 | 17:16 |
| opendevreview | Verification of a change to openstack/ironic master failed: CI: Fix our internal MTU settings https://review.opendev.org/c/openstack/ironic/+/893112 | 17:18 |
| TheJulia | Well, this is awkward. http://download.cirros-cloud.net/ | 17:27 |
| opendevreview | Verification of a change to openstack/ironic stable/xena failed: redfish_address - wrap_ipv6 address https://review.opendev.org/c/openstack/ironic/+/896210 | 17:28 |
| clarkb | frickler has been helping out with cirros things and may know what happened? | 17:32 |
| frickler | what may have happened is that another year has passed and the domain expired. I'll ping smoser | 17:39 |
| TheJulia | Is there anything we can do to assist? | 17:39 |
| opendevreview | Merged openstack/bifrost stable/2023.2: Fix BRANCH for install-deps.sh https://review.opendev.org/c/openstack/bifrost/+/897878 | 17:42 |
| frickler | well since this has been happening a couple of times now, we could either decide to consume cirros from github releases or to mirror it. note that CI images already contain the version(s) of cirros that devstack specifies, maybe you can use those directly, too, avoiding any need for a download? | 17:46 |
| TheJulia | We download a different image component that afaik is not mirrored | 17:47 |
| TheJulia | There is a third option, we could potentially figure out what it would take to fund it if it is just a pay the hosting bill thing | 17:48 |
| frickler | so far smoser has claimed that funding is not an issue, clicking the "renew" button in time is | 17:48 |
| TheJulia | Ahh, that can be a challenge | 17:49 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Make sure we eject media from DVD when CD is requested https://review.opendev.org/c/openstack/ironic/+/897989 | 17:52 |
| frickler | according to my logs, smoser has renewed the domain on 2022-08-04 for two years, so might be some other issue now | 17:55 |
| opendevreview | Verification of a change to openstack/ironic stable/wallaby failed: redfish_address - wrap_ipv6 address https://review.opendev.org/c/openstack/ironic/+/896376 | 18:00 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: WIP: Add test for dhcp-less vmedia based deployment https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/898006 | 19:42 |
| iurygregory | if anyone has some free time for reviews https://review.opendev.org/c/openstack/ironic/+/897989 a quick fix for the https://bugs.launchpad.net/ironic/+bug/2039042 we forgot to handle the case when we were fixing https://bugs.launchpad.net/ironic/+bug/2031595 =) | 19:45 |
| TheJulia | joy, do we now unattach everything? | 19:54 |
| iurygregory | if we don't pass the boot_device, yes | 19:55 |
| iurygregory | if we pass we try to respect it (unless in the case we request as CD and the hardware only has DVD support | 19:56 |
| opendevreview | Julia Kreger proposed openstack/ironic master: WIP/DNM: Advanced vmedia deployment test ops https://review.opendev.org/c/openstack/ironic/+/898010 | 19:59 |
| TheJulia | that makes sense | 19:59 |
| TheJulia | oh neat, the cirros link is back! Thanks frickler and smoser! | 19:59 |
| iurygregory | Magic \o/ | 20:00 |
| TheJulia | doh, my test needs to ensure there is a config drive | 20:00 |
| TheJulia | blah | 20:00 |
| iurygregory | it happens | 20:06 |
| TheJulia | yeah | 20:09 |
| TheJulia | it is a complex scenario I'm trying to model in one of our more advanced features, so bound to happen | 20:10 |
| opendevreview | Julia Kreger proposed openstack/ironic-tempest-plugin master: WIP: Add test for dhcp-less vmedia based deployment https://review.opendev.org/c/openstack/ironic-tempest-plugin/+/898006 | 21:51 |
| * TheJulia crosses her fingers hoping | 21:52 | |
| iurygregory | I will love if by only running "sudo apt update && sudo apt -y upgrade" before installing devstack it will solve the problem in gophercloud lol | 22:36 |
| iurygregory | normally the job would fail before 5min, 7min and counting \o/ | 22:37 |
| iurygregory | yay GREEN | 23:01 |
| iurygregory | <insert happy dance gif> | 23:01 |
| NobodyCam | Good Afternoon folks | 23:08 |
| iurygregory | hey hey NobodyCam o/ | 23:08 |
| NobodyCam | howdy iurygregory o/ | 23:09 |
| opendevreview | Julia Kreger proposed openstack/ironic master: WIP/DNM: Advanced vmedia deployment test ops https://review.opendev.org/c/openstack/ironic/+/898010 | 23:44 |
| -opendevstatus- NOTICE: Another short Gerrit outage for updates on review.opendev.org. This update ensures we are using the current versions of all Gerrit plugins. | 23:46 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!