| TheJulia | https://cb284c8f86426c242a7c-85249f43a4823dd63eb58b0df16bb6f5.ssl.cf2.rackcdn.com/835917/12/check/ironic-standalone-redfish/08552ab/controller/logs/ironic-bm-logs/node-0_no_ansi_2022-08-10-23%3A41%3A27_log.txt | 00:03 |
|---|---|---|
| TheJulia | it is *always* dns | 00:03 |
| iurygregory | DNS <3 | 00:39 |
| * iurygregory is being sarcastic | 00:40 | |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/yoga: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852797 | 01:38 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/yoga: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852797 | 05:02 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/yoga: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852797 | 05:11 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/xena: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852802 | 06:26 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/xena: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852802 | 06:34 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/wallaby: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852803 | 06:36 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 06:41 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/ussuri: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852805 | 06:45 |
| opendevreview | wangjiaqi proposed openstack/ironic-ui master: remove unicode from code https://review.opendev.org/c/openstack/ironic-ui/+/852618 | 06:48 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/train: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852806 | 06:49 |
| opendevreview | wangjiaqi proposed openstack/sushy-tools master: remove unicode from code https://review.opendev.org/c/openstack/sushy-tools/+/852622 | 07:02 |
| opendevreview | wangjiaqi proposed openstack/tenks master: remove unicode from code https://review.opendev.org/c/openstack/tenks/+/852623 | 07:05 |
| rpittau | good morning ironic! o/ | 07:45 |
| dtantsur | good morning ajya! I have an iDRAC question which you could maybe pass on: | 08:41 |
| dtantsur | if I request to enable secure boot and to boot from virtual media in the same reboot, will it cause any issues? | 08:42 |
| dtantsur | context is https://bugzilla.redhat.com/show_bug.cgi?id=2093511 | 08:42 |
| ajya | good morning dtantsur , what's ZTP in this context? | 08:51 |
| dtantsur | ajya: it's a name of one product on top of openshift | 08:51 |
| dtantsur | what matters is that they use ironic to connect their installer ISO | 08:51 |
| dtantsur | (so, the "ramdisk" deploy, not a normal one) | 08:52 |
| ajya | ok, in ironic terms what's the workflow - set secure boot and then ramdisk deploy? | 08:52 |
| ajya | dtantsur: there was testing done with enabling secure boot and deploying, but that was with ipxe. I'll check if same works with virtual media boot | 09:02 |
| opendevreview | Verification of a change to openstack/tenks master failed: remove unicode from code https://review.opendev.org/c/openstack/tenks/+/852623 | 09:21 |
| opendevreview | Jakub Jelinek proposed openstack/ironic-python-agent master: Enable skipping disks for cleaning https://review.opendev.org/c/openstack/ironic-python-agent/+/850861 | 09:30 |
| dtantsur | ajya: yeah, the workflow was like that | 09:30 |
| *** tosky_ is now known as tosky | 09:33 | |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/xena: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852802 | 09:39 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/wallaby: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852803 | 09:46 |
| opendevreview | Verification of a change to openstack/tenks master failed: remove unicode from code https://review.opendev.org/c/openstack/tenks/+/852623 | 10:03 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 10:42 |
| opendevreview | Verification of a change to openstack/tenks master failed: remove unicode from code https://review.opendev.org/c/openstack/tenks/+/852623 | 10:59 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 11:12 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/ussuri: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852805 | 11:15 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/train: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852806 | 11:27 |
| iurygregory | good morning Ironic | 11:28 |
| cbouchar | @TheJulia: I tried to look at the link you provided and got 'not found'. BTW I was pushed back onto my old project for awhile but my peer has still been using Ironic. I'll wedge this into my work. I wanted to pull in your new patch into our area but got derailed. | 11:44 |
| iurygregory | cbouchar, I got not found from the CI links also, I had to open in firefox private mode to work | 11:52 |
| cbouchar | iurygregory: That worked! TY! | 12:09 |
| iurygregory | cbouchar, yw =) | 12:09 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/xena: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852802 | 12:20 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/wallaby: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852803 | 12:26 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/victoria: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852804 | 12:28 |
| rpittau | iurygregory: not sure if it was already discussed or not but I find adding a parameter do driver_info that has the same exact name as a config option quite confusing, especially because one overrides the other :) | 12:30 |
| iurygregory | rpittau, there was a previously discussion around this let me see if I can find the link | 12:31 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/ussuri: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852805 | 12:32 |
| opendevreview | Vanou Ishii proposed openstack/ironic stable/train: Fix iRMC driver to use certification file in HTTPS https://review.opendev.org/c/openstack/ironic/+/852806 | 12:33 |
| kubajj | I am a bit confused by the documentation for RAID. In the description of the number_of_physical_disks it says that software raid always spans all the drives, but then there is the physical_disks field? Does it limit the drives it uses? Does the _do_create_configuration function ignore any hardware raid? | 12:40 |
| dtantsur | kubajj: I think we haven't fixed the docs when introduced physical_disks | 13:02 |
| dtantsur | hardware RAID is normally transparent for the userspace | 13:03 |
| kubajj | dtantsur: I am not really sure what you mean by transparent for the userspace, sorry | 13:06 |
| TheJulia | kubajj: you can't actually see the disks for hardware raid, in most cases | 13:14 |
| TheJulia | and you cannot interact with them directly | 13:14 |
| TheJulia | Also, good morning | 13:14 |
| kubajj | Good morning TheJulia | 13:14 |
| TheJulia | cbouchar: so got further, looks like my final issue is I need to fix dns resolution | 13:16 |
| TheJulia | since I pass in an FQDN, which apparently cannot be resolved | 13:16 |
| kubajj | So they are described in the target_raid_config, but not created during the apply/create_configuration. Are they deleted/erased during delete_configuration? (I am trying to extend the skip list to RAID devices too, but it is my first time interacting with RAID, so I'm sorry if my questions are trivial) | 13:17 |
| TheJulia | kubajj: for hardware raid, I don't think the idea would really work unless you are matching the *entire* device | 13:20 |
| TheJulia | and ultimately, they are described in the target_raid_config, and get created/changed/updated during apply/create config time | 13:20 |
| TheJulia | generally that is either a specialized hardware manager (uncommon these days), OR directly with a BMC that knows how to send the request to the raid controller. It varies by raid controller as to if the contents are deleted or not. | 13:21 |
| kubajj | TheJulia: If I understood it correctly, matching the entire device would be what we want to do, right? | 13:24 |
| kubajj | Because we can't skip deletion of just half of an array | 13:24 |
| TheJulia | kubajj: likely, although I would approach it as disjointed from hardware raid | 13:29 |
| TheJulia | at least initially | 13:29 |
| kubajj | TheJulia: Ok, my initial idea is to go for name of the logical disk in the skip list and remove them from raid_devices. Probably in the _scan_raids function. Then during create/apply_configuration check raid_config, which should still have the logical disk that we did not delete and just skip it so that it is not duplicated. Does this seem reasonable? | 13:33 |
| kubajj | I just realised that raid_config probably does not get updated from IPA as it is handled by the conductor/api, is it? | 13:38 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Document driver_info external_http_url https://review.opendev.org/c/openstack/ironic/+/852856 | 13:38 |
| iurygregory | good morning TheJulia =) | 13:39 |
| iurygregory | just fyi I've pushed the docs in a separate patch (so we don't have to run all functional jobs again) | 13:39 |
| TheJulia | iurygregory: can you point me to the patch. I may be in a zombie state right now | 13:41 |
| TheJulia | kubajj: your realization is correct | 13:41 |
| iurygregory | TheJulia, https://review.opendev.org/c/openstack/ironic/+/852856 =) I added in the chain (for the main one) | 13:41 |
| TheJulia | I'd keep your work, for now, constrained to the agent | 13:41 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Document driver_info external_http_url https://review.opendev.org/c/openstack/ironic/+/852856 | 13:45 |
| kubajj | TheJulia: I think I could call list_block_devices during the apply/create step and it should still have the logical disk if I don't delete it before. Is this correct? | 13:47 |
| kubajj | Also, do you think it is reasonable to operate on RAIDs using their name? I think it would make the process a lot easier, but would require users to either specify the name in the target_raid_config or figure it out somehow after their creation (is this even possible?) | 13:49 |
| *** rcastillo is now known as rcastillo|rover | 13:57 | |
| opendevreview | Ruby Loo proposed openstack/ironic master: anaconda: ks liveimg = instance_info/image_url https://review.opendev.org/c/openstack/ironic/+/852823 | 14:26 |
| TheJulia | kubajj: that is correct, if before deletion | 14:33 |
| TheJulia | kubajj: is it possible to figure out, not in host OS terms most likely | 14:36 |
| TheJulia | kubajj: The name... is likely a good idea | 14:36 |
| rloo | TheJulia: Wondering if we might want to support an instance_info/liveimg_url too. But for now... https://review.opendev.org/c/openstack/ironic/+/852823 | 14:37 |
| TheJulia | is liveimg_url explicitly the full url path so the manfiest gets read? | 14:37 |
| rloo | i believe it is the full url path. we point it to our OS tarball. | 14:38 |
| rloo | via nova... instance_info/image_source I believe | 14:38 |
| rloo | (standalone vs via nova... difficult to make consistent/simple) | 14:40 |
| TheJulia | okay, yeah | 14:42 |
| TheJulia | so I think I had gotten a -1 on exposing image_url in the past | 14:42 |
| TheJulia | fwiw | 14:42 |
| rloo | well, we can fix. just that for now -- that's what the code does to get the url for the ks liveimg command. I'm assuming that a user can set that image_url value. | 14:43 |
| rloo | if the user sets instance_info/image_source, will ironic under-the-hood set image_url based on that? Sorry, I was only updating the docs, didn't actually spend time thinking about the ironic code flow. | 14:44 |
| TheJulia | rloo: yes, it does afaik | 14:52 |
| rloo | TheJulia: I'll change it to image_info then. | 14:53 |
| TheJulia | in part, it is why I ended up doing the is source a path stuff the way I did | 14:53 |
| rloo | ah... | 14:53 |
| * TheJulia is trying to figure out why dns no work | 14:55 | |
| opendevreview | Ebbex proposed openstack/bifrost master: make sure we quote extra_kernel_options https://review.opendev.org/c/openstack/bifrost/+/852877 | 14:56 |
| TheJulia | heh, no dns on the private subnet | 14:56 |
| opendevreview | Ruby Loo proposed openstack/ironic master: anaconda: ks liveimg = instance_info/image_info https://review.opendev.org/c/openstack/ironic/+/852823 | 14:57 |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Anaconda CI changes https://review.opendev.org/c/openstack/ironic/+/849587 | 15:17 |
| * TheJulia crosses her fingers hoping things just "work" | 15:23 | |
| rpittau | bye everyone, see you next wednesday! o/ | 15:28 |
| JayF | We are not permitted to update the driver-requirements.txt for backports, are we? | 16:01 |
| JayF | looking at https://review.opendev.org/c/openstack/ironic/+/852797/3/driver-requirements.txt#b9 | 16:01 |
| dtantsur | Ideally no | 16:05 |
| * dtantsur finds another nasty issue in fast-track, SIGH | 16:05 | |
| TheJulia | ?another?! | 16:07 |
| TheJulia | sigh, I've told them they need to pin the version appropriately and I believe they have even posted fixed versions | 16:07 |
| dtantsur | TheJulia: fun? https://bugzilla.redhat.com/show_bug.cgi?id=2106378#c9 | 16:08 |
| JayF | TheJulia: Are you talking about my thing? | 16:09 |
| JayF | nice find on that dtantsur | 16:10 |
| TheJulia | JayF: yeah | 16:10 |
| TheJulia | JayF: another was meant for dtantsur | 16:10 |
| JayF | TheJulia: What does "pin the version appropriately" mean? Did they release a stable version of the earlier clients and just need to bump 'em to that? | 16:10 |
| dtantsur | ajya: if you by chance still online: wdyt https://bugzilla.redhat.com/show_bug.cgi?id=2106378#c9? I think it's an actual bug in the interaction between fast-track and the way we do post-clean-step reboots. | 16:10 |
| TheJulia | JayF: fixes on stable branches of python-scciclient | 16:10 |
| JayF | ack, makes sense | 16:10 |
| TheJulia | with appropriate version releases | 16:10 |
| JayF | I'll go make my comment more sensible in that context | 16:11 |
| dtantsur | I wonder what is better. 1) Use permanent boot with fast-track, 2) Power off the node before prepare_ramdisk, 3) Restore ISO on every reboot. | 16:11 |
| JayF | What is the use-case for a one time boot configuration? | 16:12 |
| JayF | Like, why is that virtuous to ever do? | 16:12 |
| TheJulia | dtantsur: I've not fully digested yet | 16:12 |
| TheJulia | ohhhh | 16:13 |
| TheJulia | so.. I don't think 1 would work with vmedia | 16:13 |
| dtantsur | JayF: "we've always done it like that"? | 16:13 |
| dtantsur | i.e. I don't remember | 16:13 |
| TheJulia | we've seen some BMCs reject such since it is not a permeenant device | 16:13 |
| dtantsur | yeah, permanent boot does not work everywhere | 16:14 |
| TheJulia | hmm... 2 perhaps might work | 16:14 |
| JayF | got it, permanent boot is literally a BMC setting on Redfish, then? | 16:14 |
| TheJulia | might | 16:14 |
| TheJulia | hmmm | 16:14 |
| TheJulia | JayF: yeah | 16:14 |
| * JayF only missing years of context to catch up on, no biggie | 16:14 | |
| TheJulia | dtantsur: maybe 3 might seem more elegant ? I'm just not sure. i suspect this is one of those "got to try first" things | 16:15 |
| * TheJulia needs a coffee IV this morning | 16:16 | |
| dtantsur | I have mixed feelings. 3 required the power interface to understand ISOs and fast-track. | 16:18 |
| dtantsur | I kinda feel like trying 2 and seeing where it leads me | 16:18 |
| iurygregory | I think we discussed about rising the min-version (I don't think it would make sense...) re python-scciclient | 16:18 |
| TheJulia | dtantsur: well... the conductor power handler *is* handed a task and has logic built in for smartnics | 16:27 |
| TheJulia | iurygregory: so, we *did* also catch breaking changes which required them to fix branches | 16:27 |
| TheJulia | I've not re-checked them yet | 16:28 |
| dtantsur | TheJulia: yeah, but it knows nothing about ISOs. All it can do is to call prepare_ramdisk, which helpfully no-ops in fast-track :) | 16:28 |
| JayF | I guess it's hard to trigger the right behavior without mixing interfaces | 16:35 |
| TheJulia | yeah | 16:41 |
| dtantsur | actually, there is 4. Power off before operations that will result in a reboot, like BIOS or RAID settings. | 16:44 |
| dtantsur | I wonder what BMCs think about it... | 16:44 |
| TheJulia | I like 4 | 16:45 |
| TheJulia | but... the BMC needs to be "done done" | 16:45 |
| dtantsur | sorry? | 16:45 |
| dtantsur | I mean, do power off as the first action in e.g. apply_configuration. then do the actual application. | 16:46 |
| dtantsur | I guess I should sync with ajya on whether changing stuff whiled powered off is problematic. | 16:46 |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Anaconda CI changes https://review.opendev.org/c/openstack/ironic/+/849587 | 16:47 |
| iurygregory | TheJulia, I'm going to take a closer look at the patches after finishing other things here | 16:49 |
| dtantsur | folks, a corporate holiday tomorrow, so see you on Monday o/ | 16:57 |
| TheJulia | dtantsur: o/ | 16:59 |
| TheJulia | JayF: the -1'ed patches on https://review.opendev.org/q/I4bd8b1b4ea974834f1149fcaa79de85d24f5f7d1 should pass now, recheck they were all dependent upon another fix to merge first. | 17:00 |
| TheJulia | JayF: I issued rechecks accordingly | 17:00 |
| JayF | aha, good stuff | 17:00 |
| opendevreview | Julia Kreger proposed openstack/ironic master: Redfish: Consider password part of the session cache https://review.opendev.org/c/openstack/ironic/+/850259 | 17:09 |
| TheJulia | hjensas: quick question on https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 17:45 |
| TheJulia | context is getting removed.... o.O | 17:46 |
| hjensas | TheJulia: looks like I tought context would be required there in a previous change where delete_pre_conf_aggregate was a "noop". I guess it would be good to clean that up in the parent and rebase. | 17:53 |
| TheJulia | has the parent that introduced it been merged yet? | 17:59 |
| hjensas | no | 17:59 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 18:07 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 18:07 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 18:07 |
| hjensas | TheJulia: ignore ^, looks like I made a mistake in the rebasing. | 18:09 |
| * hjensas is checking. | 18:10 | |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 18:14 |
| *** tosky_ is now known as tosky | 18:16 | |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 18:28 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 18:28 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 18:28 |
| opendevreview | Merged openstack/ironic stable/yoga: Update raid_type handling for Redfish raid_config https://review.opendev.org/c/openstack/ironic/+/852639 | 18:31 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add netconf-openconfig device driver https://review.opendev.org/c/openstack/networking-baremetal/+/835324 | 18:40 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 18:40 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 18:40 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add LACP support to Netconf OpenConfig driver https://review.opendev.org/c/openstack/networking-baremetal/+/835378 | 18:49 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Add support for pre-configured link aggregates https://review.opendev.org/c/openstack/networking-baremetal/+/840533 | 18:49 |
| hjensas | TheJulia: I realized some moving around code etc in the chain that made it more painful for reviews. I did some cleanup ... | 18:55 |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Anaconda CI changes https://review.opendev.org/c/openstack/ironic/+/849587 | 19:06 |
| opendevreview | Merged openstack/networking-baremetal master: Device management driver iface https://review.opendev.org/c/openstack/networking-baremetal/+/835170 | 19:50 |
| TheJulia | cbouchar: is there like a realistic minimum for VM based testing with anaconda? 2GB VMs don't seem like they are enough | 20:55 |
| TheJulia | oh, helps if I change the correct job config | 21:02 |
| opendevreview | Julia Kreger proposed openstack/ironic master: DNM: Anaconda CI changes https://review.opendev.org/c/openstack/ironic/+/849587 | 21:02 |
| iurygregory | bye everyone! have a great weekend o/ | 21:44 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Doc - network device configuration capabilities https://review.opendev.org/c/openstack/networking-baremetal/+/852924 | 22:55 |
| opendevreview | Harald Jensås proposed openstack/networking-baremetal master: Doc - network device configuration capabilities https://review.opendev.org/c/openstack/networking-baremetal/+/852924 | 23:36 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!