| *** pmannidi is now known as pmannidi|Lunch | 01:41 | |
| *** pmannidi|Lunch is now known as pmannidi | 03:04 | |
| iurygregory | good morning Ironic | 05:36 |
|---|---|---|
| stendulker_ | iurygregory: o/ | 05:45 |
| iurygregory | stendulker_, o/ | 05:47 |
| stendulker_ | iurygregory: One question.. | 05:47 |
| iurygregory | stendulker_, sure | 05:47 |
| stendulker_ | in spec you have mentioned bmc events would be management interface, but for redfish its implemented as vendor_passthru | 05:48 |
| iurygregory | stendulker_, we dropped the spec :D | 05:48 |
| stendulker_ | is it ? :) | 05:48 |
| iurygregory | it will be just an idea because there will be new fields required in the newest versions of redfish | 05:49 |
| stendulker_ | you just updated it yday... | 05:49 |
| iurygregory | we moved to Ideas folder =) | 05:49 |
| iurygregory | ops backlog | 05:49 |
| stendulker_ | ok, so vendor_passthru is the way to go.. | 05:49 |
| iurygregory | yeah for now =) | 05:49 |
| stendulker_ | ok, thanks :) | 05:49 |
| iurygregory | np =) | 05:49 |
| iurygregory | maybe in the future we will provide as an API (we don't want to add and API and say that some options on it are deprecated) | 05:50 |
| iurygregory | stendulker_, there is a releasenote .-. | 06:00 |
| iurygregory | https://review.opendev.org/c/openstack/ironic/+/801064/14/releasenotes/notes/vendor-passthru-subscription-5d28a2420e2af111.yaml | 06:00 |
| stendulker_ | iurygregory: oops, i was comparing against older patch... my problem | 06:01 |
| iurygregory | no worries =) | 06:01 |
| stendulker_ | iurygregory: have one comment on releasenotes. have a look if you want to address it.. | 06:04 |
| iurygregory | stendulker_, can I address in a follow-up? | 06:07 |
| iurygregory | so we don't run functional jobs again =) | 06:07 |
| iurygregory | ie any ironic-base jobs https://github.com/openstack/ironic/blob/master/zuul.d/ironic-jobs.yaml#L23 | 06:07 |
| stendulker_ | iurygregory: fine | 06:07 |
| iurygregory | going to drink more coffee and I will push the follow-up =) | 06:08 |
| opendevreview | Iury Gregory Melo Ferreira proposed openstack/ironic master: Update vendor_passthru subscriptions releasenote https://review.opendev.org/c/openstack/ironic/+/801705 | 06:19 |
| cenne | Mornings iurygregory, stendulker_ | 06:40 |
| stendulker_ | cenne o/ | 06:40 |
| iurygregory | hey cenne o/ | 06:41 |
| cenne | hey o/ | 06:43 |
| *** pmannidi is now known as pmannidi|brb | 07:08 | |
| opendevreview | Merged openstack/ironic master: Add vendor_passthru method for subscriptions https://review.opendev.org/c/openstack/ironic/+/801064 | 07:46 |
| iurygregory | \o/ YAY \o/ | 07:46 |
| iurygregory | rpittau|afk, fyi ^ | 07:46 |
| *** rpittau|afk is now known as rpittau | 07:49 | |
| rpittau | gooooood morning ironic! o/ | 07:49 |
| iurygregory | morning rpittau o/ | 07:49 |
| rpittau | woohoo that's awesome iurygregory :) | 07:50 |
| rpittau | going to ask for the release now | 07:50 |
| iurygregory | rpittau, maybe we can just approve the release note fix https://review.opendev.org/c/openstack/ironic/+/801705 | 07:50 |
| iurygregory | should merge fast since is just a releasenote change | 07:50 |
| rpittau | done | 07:51 |
| iurygregory | ty | 07:51 |
| iurygregory | do we need anything else for bifrost or we are good? | 07:51 |
| rpittau | well I'm going to submit one change now | 07:53 |
| rpittau | a corrrection | 07:53 |
| rpittau | if we can merge that also would be great | 07:53 |
| rpittau | then we should be good | 07:53 |
| opendevreview | Riccardo Pittau proposed openstack/bifrost master: Show debug messages during upgrade ci job https://review.opendev.org/c/openstack/bifrost/+/801608 | 07:54 |
| rpittau | there it is ^ | 07:54 |
| iurygregory | rpittau, ack =) | 07:54 |
| opendevreview | Merged openstack/ironic master: Update vendor_passthru subscriptions releasenote https://review.opendev.org/c/openstack/ironic/+/801705 | 08:14 |
| iurygregory | rpittau, +1 on the cut for the ironic bugfix, now let's hope release team will take care of it today :D | 08:48 |
| rpittau | heh fingers crossed also for inspector and ipa | 08:48 |
| iurygregory | yeah | 08:49 |
| * cenne intimidated by mocking (management interface etc) for unit-tests | 09:01 | |
| rpittau | btw the lower-constraints job seems working fine in ironic https://review.opendev.org/c/openstack/ironic/+/801350 | 09:05 |
| janders__ | good morning Ironic o/ | 09:08 |
| cenne | good morning janders__ o/ | 09:09 |
| iurygregory | hey janders__ o/ | 09:20 |
| iurygregory | rpittau, oh WOW | 09:20 |
| dtantsur | good late morning ironic | 10:12 |
| rpittau | hey dtantsur :) | 10:12 |
| dtantsur | I see we're ready to proceed with ironic and bifrost releases? good | 10:15 |
| rpittau | for bifrost, I'm waiting for https://review.opendev.org/c/openstack/bifrost/+/801608 to merge, dtantsur wdyt? | 10:22 |
| dtantsur | agreed, lemme check | 10:36 |
| rpittau | thanks | 10:36 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-python-agent bugfix/8.1: Update .gitreview for bugfix/8.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/801762 | 10:37 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic-inspector bugfix/10.7: Update .gitreview for bugfix/10.7 https://review.opendev.org/c/openstack/ironic-inspector/+/801764 | 10:37 |
| dtantsur | This time, for a change, I'm considering actually making the CI working for the new branches :) | 10:41 |
| opendevreview | Merged openstack/ironic-inspector bugfix/10.7: Update .gitreview for bugfix/10.7 https://review.opendev.org/c/openstack/ironic-inspector/+/801764 | 11:02 |
| opendevreview | Merged openstack/ironic-python-agent bugfix/8.1: Update .gitreview for bugfix/8.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/801762 | 11:02 |
| opendevreview | Merged openstack/bifrost master: Show debug messages during upgrade ci job https://review.opendev.org/c/openstack/bifrost/+/801608 | 11:43 |
| timeu | Hi all!, So I am getting closer to deploying Lenovo nodes with virtual media. However for some reason the deploy iso (ESP + IPA) can't load the ramdisk and then the kernel panics because it can't mount the root file system. The error is can't allocate initrd. Anybody has an idea what might be causing this ? | 11:49 |
| iurygregory | I do remember someone having trouble with initrd... if I recall it was a bad image? | 11:54 |
| * iurygregory checks openstack-discuss to find the thread | 11:54 | |
| iurygregory | ok the thread starts in http://lists.openstack.org/pipermail/openstack-discuss/2021-April/021512.html | 11:54 |
| timeu | iurygregory: Thanks will check it. FYI: The same initrd/kernel pair works fine with ipxe/pxe deployment/cleaning | 11:59 |
| timeu | secure boot is also disabled btw | 12:06 |
| opendevreview | cenne proposed openstack/ironic master: Add api endpoint for setting boot_mode on managed node https://review.opendev.org/c/openstack/ironic/+/800084 | 12:35 |
| opendevreview | Dmitry Tantsur proposed openstack/bifrost master: DNM test the upgrade job https://review.opendev.org/c/openstack/bifrost/+/800673 | 12:44 |
| TheJulia | good morning | 13:02 |
| iurygregory | good morning TheJulia | 13:03 |
| opendevreview | OpenStack Release Bot proposed openstack/ironic bugfix/18.1: Update .gitreview for bugfix/18.1 https://review.opendev.org/c/openstack/ironic/+/801848 | 13:04 |
| iurygregory | I don't see bugfix/18.1 in opendev.org or github.com .-. (maybe is just me...) | 13:06 |
| TheJulia | they sync from review.opendev.org | 13:07 |
| iurygregory | well since the bot pushed the patch I think things are ok :D | 13:08 |
| rpittau | never trust the bot(s) | 13:08 |
| janders__ | see you tomorrow Ironic o/ | 13:08 |
| rpittau | bye janders__ :) | 13:08 |
| dtantsur | iurygregory: I see the branch in git | 13:10 |
| iurygregory | dtantsur, maybe is just my firefox with cache problems... | 13:11 |
| dtantsur | I did it with local git | 13:11 |
| iurygregory | oh =D | 13:11 |
| TheJulia | I need a coffee IV | 13:17 |
| iurygregory | I'm at coffee III atm | 13:18 |
| timeu | TheJulia: Good morning or afternoon (at least here it is ;-)). I got a bit of a success with virtual media. The iso now is properly mounted in a slot and the node boots from it. I can also see that the bootloader tries to load the IPA kernel and ramdisk, however it fails with "can't allocate initrd" and the kernel panics. I checked the contents of the iso and the included | 13:27 |
| timeu | vmlinuz and initrd are the same that work with pxe/ipxe (basically from upstream IPA). iurygregory pointed me to a mailing list thread but I think it's a different issue. | 13:27 |
| timeu | This is the contents of the iso and grub.cfg: https://pastebin.com/EmrLzw91 | 13:29 |
| TheJulia | umm, is it missing the closing bracket or did you just not copy/paste that | 13:31 |
| timeu | copy and paste error | 13:32 |
| timeu | sorry | 13:32 |
| timeu | I only found an old ticket: https://bugzilla.redhat.com/show_bug.cgi?id=1572126 but that's an old grub2 RPM | 13:32 |
| timeu | I will try to enable grub2 debug output to see it provides any information, but the whole bootloader/uefi/ startup is a bit of dark magic for me ;-) | 13:35 |
| TheJulia | yeah, this seems partially familiar | 13:36 |
| TheJulia | but... it should work I think, which is what has me confused | 13:36 |
| timeu | secure boot is disalbed and I copied the /boot/efi/EFI/redhat/grubx64.efi and /boot/efi/EFI/redhat/shimx64.efi (not the shimx64-redhat.efi one) into the ESP iso (as explained in the docs0 | 13:39 |
| opendevreview | Merged openstack/ironic bugfix/18.1: Update .gitreview for bugfix/18.1 https://review.opendev.org/c/openstack/ironic/+/801848 | 13:47 |
| TheJulia | dtantsur: does timeu's issue sound familiar? | 14:06 |
| dtantsur | timeu: maybe it sounds silly, but has worked so many times: have you tried updating firmware? especially UEFI? | 14:36 |
| TheJulia | I do feel like there *was* a case where it was image construction a couple years ago | 14:36 |
| dtantsur | it tends to fail on the early grub stage though | 14:37 |
| TheJulia | Yeah, the initrd not loading is really weird | 14:45 |
| TheJulia | but also familiar | 14:45 |
| timeu | TheJulia: I will check again, but I think the firmware is the latest because of the redfish issues we had | 14:45 |
| timeu | but I will try to play around with the iso and see if I can figure out something more | 14:45 |
| TheJulia | timeu: rhbz 1984968 created for the permission issue | 14:45 |
| timeu | I manually loaded the kernel in the grub prompt and that worked | 14:45 |
| timeu | but when I try it with the initrd I get the cannot allocate | 14:46 |
| TheJulia | wow | 14:46 |
| timeu | TheJulia thanks ! I subscribe to it | 14:46 |
| timeu | I don't know if this is somehow related but we have quite a bit of UEFI entries in that node (ironic, redhat, centos1, centos2, etc) | 14:49 |
| timeu | I can try to clean those also up | 14:49 |
| TheJulia | timeu: unfortunately not related, your post that most likely | 14:53 |
| timeu | for completness I posted the screen grabs of the console here: https://imgur.com/a/5HP8s35 | 14:56 |
| timeu | the first one is the error and the second one is the kernel panic that comes because I guess it can't load initrd and is unable to mount the root filesystem | 14:56 |
| dtantsur | I wonder if your hardware simply has a limit on the allocatable memory | 15:03 |
| dtantsur | how large is your initramfs? | 15:05 |
| timeu | hmm but shouldn't this then also affect pxe/ipxe boot ? 538 MB | 15:05 |
| timeu | it's the IPA one | 15:05 |
| dtantsur | fair enough | 15:05 |
| timeu | maybe the virtual media slot has a limitation ? | 15:06 |
| timeu | I need to check. I would try to mount the generated iso manually via the web interface | 15:06 |
| timeu | of the BMC | 15:06 |
| dtantsur | you can do that | 15:06 |
| timeu | and see if I can figure something out | 15:06 |
| dtantsur | also try upstream images, they're somewhat smaller https://tarballs.opendev.org/openstack/ironic-python-agent/dib/files/ | 15:06 |
| timeu | will do | 15:06 |
| dtantsur | I'm curious if they work differently | 15:06 |
| timeu | the lenovo machnes have different virtual media slots that support different transport protocols | 15:06 |
| dtantsur | which reminds me, we need a plan to switch of CentOS 8 :( | 15:07 |
| timeu | and maybe have different limitations | 15:07 |
| dtantsur | I'm aware of hardware that has 150 MiB limit in virtual media. But you'd probably learn it earlier. | 15:07 |
| timeu | ah maybe I found something. the RDOC | 15:10 |
| timeu | Upload an ISO or IMG image file to the BMC, then mount it to the host as a DVD or USB drive. The BMC storage space is restricted to 50 MB in total. | 15:10 |
| timeu | Note: The client session could be closed without affecting the mounted media.y | 15:10 |
| timeu | has a limit of 50 MB | 15:10 |
| timeu | and the EXT1/2 does not support HTTP I think :-( | 15:10 |
| dtantsur | 50M is a joke. kernel+initramfs on my Fedora would barely fit there. | 15:11 |
| dtantsur | (they're 46M together; add grub and you're probably out) | 15:12 |
| timeu | ah n sorry RODC1 and 2 are limited to 50 MB which does not support http(s)/NFS | 15:13 |
| timeu | I need to check how much the other slots support | 15:13 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/ussuri: Provide a path to set explicit ipxe bootloaders https://review.opendev.org/c/openstack/ironic/+/801867 | 15:18 |
| TheJulia | stevebaker: ^^ backported because it was not in stable/train and without it landing the fix into stable/train would be a stable policy violation | 15:19 |
| TheJulia | also, if there is any stable review braincell carrying core available, I'd appreciate ^^^ getting appropriate buttons mashed | 15:20 |
| TheJulia | dtantsur: timeu: oh wow :( | 15:21 |
| iurygregory | looking now TheJulia =) | 15:30 |
| iurygregory | oh WOW stable/ussuri :D | 15:30 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/ussuri: Refactor RedfishVirtualMediaBoot https://review.opendev.org/c/openstack/ironic/+/801870 | 15:35 |
| opendevreview | Julia Kreger proposed openstack/ironic stable/ussuri: Fix redfish-virtual-media file permission https://review.opendev.org/c/openstack/ironic/+/801871 | 15:35 |
| TheJulia | iurygregory: going to train | 15:35 |
| TheJulia | timeu: fyi, ussuri ^ we'll have to see how it maps out in CI | 15:35 |
| timeu | \o? | 15:36 |
| timeu | \o/ | 15:36 |
| iurygregory | hehehe omg :D | 15:36 |
| iurygregory | I think I will need more coffee and wait food before looking at the backports :D | 15:37 |
| TheJulia | I was going to put it off until next week but since a priority issue came in which also required backport, I was like "eh, I'll just get it started | 15:38 |
| iurygregory | it makes sense =) | 15:39 |
| * TheJulia goes and refills the caffeine transport device | 15:40 | |
| iurygregory | take a look around before trying to refill | 15:40 |
| iurygregory | corgi and cat are looking at you :D | 15:41 |
| dtantsur | folks, do we already have a bifrost release proposed? I cannot remember | 15:41 |
| dtantsur | ah, found it | 15:41 |
| rpittau | yes we... do :D | 15:41 |
| dtantsur | I'm going to try set up integration jobs for the new branches unless somebody is already looking into it | 15:42 |
| * TheJulia wheels the coffee maker into the center of the channel and turns it on | 15:42 | |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-inspector bugfix/10.7: Configure CI for bugfix/10.7 https://review.opendev.org/c/openstack/ironic-inspector/+/801873 | 15:45 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic bugfix/18.1: Configure CI for bugfix/18.1 https://review.opendev.org/c/openstack/ironic/+/801876 | 15:50 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent bugfix/8.1: Configure CI for bugfix/8.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/801898 | 15:55 |
| dtantsur | we're actually supposed to do this ^^^ every time but I guess nobody had time/spoons | 15:55 |
| TheJulia | https://www.amazon.com/The-Tick-Season-1/dp/B0875XX914 <-- clearly we all need to watch that for spooooons | 15:56 |
| dtantsur | looks crazy :) | 15:56 |
| TheJulia | although i don't think that was the battle cry in the new show | 15:56 |
| TheJulia | it was in the animated one like 2 decades ago | 15:57 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic-python-agent bugfix/8.1: Configure CI for bugfix/8.1 https://review.opendev.org/c/openstack/ironic-python-agent/+/801898 | 15:58 |
| opendevreview | Leo McGann proposed openstack/ironic master: Adds Attestation Interface to Node https://review.opendev.org/c/openstack/ironic/+/755836 | 15:59 |
| rpittau | ok, running out of batteries, see ya tomorrow! o/ | 16:01 |
| *** rpittau is now known as rpittau|afk | 16:02 | |
| dtantsur | iurygregory: do we have bugfix releases for IPE? | 16:02 |
| * dtantsur hopes no | 16:02 | |
| * TheJulia wonders if there is a need atm | 16:03 | |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Document making bugfix releases and branches https://review.opendev.org/c/openstack/ironic/+/801903 | 16:09 |
| dtantsur | I should have done this long ago ^^^ | 16:09 |
| TheJulia | ++ | 16:09 |
| dtantsur | ugh, an ugly typo there | 16:10 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic bugfix/18.1: Configure CI for bugfix/18.1 https://review.opendev.org/c/openstack/ironic/+/801876 | 16:10 |
| opendevreview | Leo McGann proposed openstack/ironic master: Adds Attestation Interface to Node https://review.opendev.org/c/openstack/ironic/+/755836 | 16:13 |
| opendevreview | Dmitry Tantsur proposed openstack/ironic master: Document making bugfix releases and branches https://review.opendev.org/c/openstack/ironic/+/801903 | 16:16 |
| dtantsur | TheJulia, iurygregory ^^ | 16:17 |
| TheJulia | dtantsur: thanks | 16:17 |
| timeu | TheJulia: dtantsur: I tried now the upstream IPA kernel and ramdisk ipa-centos7-stable-train and they do work | 16:18 |
| timeu | the iso is now 350 MB instead of 550 MB, not sure if this is due to that | 16:19 |
| timeu | or something else | 16:19 |
| dtantsur | interesting. I wonder if it's because of the size or because of centos 7 | 16:19 |
| dtantsur | are you on Train? | 16:19 |
| timeu | yes OSP 16.1 | 16:19 |
| timeu | I re-used the IPA that is used in the undercloud | 16:19 |
| timeu | to provision the hypervisors/controllers | 16:19 |
| timeu | the nodes are identital (baremetal, hypervisor). | 16:19 |
| dtantsur | if you're up for more experimenting, maybe try centos8-stable-ussuri? I think these should be still compatible. | 16:20 |
| timeu | yeah will give it a try | 16:20 |
| dtantsur | I'm trying to understand if it's more about the size or the contents | 16:20 |
| TheJulia | ditto | 16:20 |
| TheJulia | jungleboyj: anyone at lenovo who may have insight w/r/t virtual media? | 16:20 |
| timeu | I have a backup of both generated ISOs if we need to compare something | 16:22 |
| NobodyCam | Good Morning Ironic Folks | 16:24 |
| TheJulia | NobodyCam: was it you that suggested hash ring entry subsitution on the pain points etherpad? | 16:25 |
| NobodyCam | anyone here ever build IPA on rhel8 aarch64?? | 16:25 |
| NobodyCam | hitting a wall with biosdevname package | 16:25 |
| TheJulia | I have not, I have no aarch64 gear | 16:26 |
| TheJulia | *but* aiui, there is no biosdevname package on aarch64 | 16:26 |
| NobodyCam | LOL I have a moonshot | 16:26 |
| dtantsur | this is correct | 16:26 |
| TheJulia | NobodyCam: deja vu | 16:26 |
| NobodyCam | hehehe | 16:27 |
| dtantsur | biosdevname is supposed to be excluded on the lastest IPA-builder, worth double-checking | 16:27 |
| NobodyCam | ye | 16:27 |
| NobodyCam | yea even | 16:27 |
| TheJulia | bifrost was developed on a moonshot, but with x86 blades | 16:27 |
| NobodyCam | ++ I may have a older clone | 16:27 |
| TheJulia | so I think the arm moonshot was the moonshot second generation, fwiw | 16:28 |
| timeu | hmm as far as I can tell it booted the centos7 IPA but it doesn't seem to clean anything or maybe it did but didn't report back o_0 | 16:28 |
| dtantsur | it can take a while to boot. is there anything on the console | 16:28 |
| NobodyCam | ahh... | 16:28 |
| NobodyCam | `biosdevname:` | 16:29 |
| NobodyCam | `not-arch: ppc64,ppc64el,ppc64le` | 16:29 |
| dtantsur | ugh | 16:29 |
| NobodyCam | I'll add aarch64 there | 16:29 |
| timeu | I just see the login shell | 16:29 |
| dtantsur | NobodyCam: maybe better reverse it | 16:29 |
| timeu | and a warning that lshw uses legacy ethtool | 16:29 |
| dtantsur | to include only intel's arch | 16:29 |
| dtantsur | timeu: okay, so at least something is going on. maybe wait a bit more? | 16:30 |
| NobodyCam | oh I can try that in a few min | 16:30 |
| timeu | yeah will do in the meantime I downloaded the centos8 usuri IPA images | 16:30 |
| timeu | and uploaded it to glance | 16:30 |
| timeu | can test this afterwards | 16:30 |
| TheJulia | dtantsur: that is a good point, bios being specifically an x86 concept amongst the architectures | 16:32 |
| NobodyCam | :) | 16:33 |
| TheJulia | next someone is going to ask for riscv | 16:34 |
| opendevreview | Leo McGann proposed openstack/python-ironicclient master: Adds Attestation Interface to Node and Driver https://review.opendev.org/c/openstack/python-ironicclient/+/755837 | 16:34 |
| dtantsur | me, me, I want riscv! | 16:34 |
| TheJulia | https://static.boredpanda.com/blog/wp-content/uploads/2020/03/corgis-37-5e5d29e880889__700.jpg | 16:36 |
| timeu | ok I aborted the cleaning. I tested on another node the cleaning with ipxe and it was basically super fast. I am now trying the ussuri IPA | 16:40 |
| timeu | ok good news dtantsur: TheJulia: ipa-centos8-stable-ussuri works | 16:43 |
| timeu | node is now available | 16:43 |
| dtantsur | mm, so it's probably about the size | 16:43 |
| timeu | yeah looks like it | 16:44 |
| timeu | why is the redhat IPA so much bigger ? | 16:44 |
| timeu | than the upstream one ? | 16:44 |
| TheJulia | firmware, lots of firmware | 16:44 |
| TheJulia | It has all of the hardware firmware blobs where as upstream we strip out a lot of the excess we know wouldn't be needed in 95% of users | 16:44 |
| timeu | ah I see | 16:45 |
| timeu | btw: I see some FAILED systemd services during startup | 16:45 |
| timeu | but I guess that's ok | 16:45 |
| TheJulia | Yeah, unfortunately to be expected | 16:45 |
| dtantsur | it also has additional introspection modules | 16:45 |
| timeu | NetworkManager-wait-online and System Security Services | 16:45 |
| timeu | so if it's size the big questions is now what's the magical limit ;-) | 16:46 |
| timeu | I will try a deployment to check if the deploymen works with the ussuri IPA but I guess or hope it should | 16:46 |
| timeu | will update the storyboard regarding virtual media about the findings | 16:46 |
| TheJulia | timeu: if you look at ironic-python-agent-builder's steps, you'll be able to see where we start stripping firmware out of the image builds. That may be a good list to start with if you want to unpack and remove a lob or three | 16:46 |
| TheJulia | uncompressed, I know one of the stock blobs is like 200MB | 16:47 |
| timeu | I see ok will look into it | 16:47 |
| TheJulia | k | 16:48 |
| timeu | ok deplopyment worked \o/. I just get following redfish error in the logs: | 16:55 |
| timeu | 2021-07-22 16:54:22.535 7 WARNING ironic.drivers.modules.redfish.management [req-a93e0df4-503c-480c-83d3-ad63121a4ea0 - - - - -] Could not set persistent boot device to hdd for node 01c9e448-b833-4a11-a561-3282b27e1205, using one-time boot device instead: sushy.exceptions.BadRequestError: HTTP PATCH https://10.100.66.71/redfish/v1/Systems/1 returned code 400. | 16:55 |
| timeu | Base.1.8.GeneralError: A general error has occurred. See ExtendedInfo for more information. Extended information: [{'Resolution': 'Choose a value from the enumeration list that the implementation can support and resubmit the request if the operation failed.', 'MessageArgs': ['Continuous', 'BootSourceOverrideEnabled'], '@odata.type': '#Message.v1_1_0.Message', 'MessageId': | 16:55 |
| timeu | 'Base.1.8.PropertyValueNotInList', 'Message': 'The value Continuous for the property BootSourceOverrideEnabled is not in the list of acceptable values.', 'MessageSeverity': 'Warning'}] | 16:55 |
| TheJulia | Now that is a redfish feature we expect to be present | 16:56 |
| timeu | but apart from the the re-boot issue, the node boots fine | 16:56 |
| TheJulia | which lenovo does not have nor intends (aiui) to support | 16:57 |
| TheJulia | *but* I think we patched handling of that.... in ?victoria? | 16:57 |
| TheJulia | dtantsur: do you remember^ | 16:57 |
| TheJulia | err | 16:57 |
| TheJulia | no, it would have been wallaby | 16:57 |
| dtantsur | given that it works, I think the fallback worked? | 16:57 |
| TheJulia | timeu: it just didn't reboot right? | 16:58 |
| timeu | well it did | 16:58 |
| dtantsur | "using one-time boot device instead" | 16:58 |
| TheJulia | it shoudl yeah | 16:58 |
| TheJulia | if that got backported to train which I don't remember right now | 16:59 |
| TheJulia | so peaking of which, I think the key at this point is to ultimately make sure we set boot devices before trying to do bootloader $things | 16:59 |
| timeu | the double reboot is due to UEFI I think | 16:59 |
| timeu | it says it can't find uefi entry, creates a new entry and then does a reboot | 17:00 |
| * TheJulia blinks | 17:00 | |
| timeu | I had this also with ipxe boot | 17:00 |
| timeu | dtantsur: do you remember, I posted about this some time ago. You mentioned that's due to some fallback or so | 17:01 |
| timeu | “System BootOrder not found. Initializing Defaults. Creating boot entry Boot005 with label “CentOS Linux” for file │ coreycb | 17:01 |
| timeu | │ | “\EFI\centos\shimx64.efi” Reset system | 17:01 |
| timeu | this is what's shown in the console | 17:01 |
| TheJulia | omg | 17:02 |
| timeu | but I am quite happy that virtual media works now on the Lenovo nodes ;-) | 17:02 |
| timeu | thanks for all the help! Now i need to organize some dinner ;-) | 17:03 |
| TheJulia | dinner++ | 17:03 |
| TheJulia | timeu: \o/ | 17:03 |
| timeu | :-] | 17:04 |
| TheJulia | timeu: fwiw, there is an issue in rhel8.4 with shim doing that update and things breaking | 17:04 |
| TheJulia | but we've got code in ironic to handle it in advance | 17:04 |
| dtantsur | timeu: it's a normal behavior of RH systems | 17:04 |
| timeu | o_0 good to know | 17:04 |
| dtantsur | https://github.com/rhboot/shim/blob/a1170bb00a116783cc6623b403e785d86b2f97d7/README.fallback#L33-L46 | 17:04 |
| timeu | ah I rememver the link ;-) I read it a couple of times and didn't understand it. Probably need to read it a couple of times more and some UEFI bootloader documentation | 17:05 |
| dtantsur | so if they find you using the default UEFI path, they assume something went wrong | 17:06 |
| dtantsur | because a normal installer would use the OS-specific path | 17:06 |
| dtantsur | so they re-create the OS-specific entry in UEFI | 17:06 |
| dtantsur | timeu: ^^ | 17:06 |
| dtantsur | (re-create in our case just create) | 17:06 |
| opendevreview | Leo McGann proposed openstack/ironic master: Adds Attestation Interface to Node https://review.opendev.org/c/openstack/ironic/+/755836 | 17:41 |
| opendevreview | Leo McGann proposed openstack/python-ironicclient master: Adds Attestation Interface to Node and Driver https://review.opendev.org/c/openstack/python-ironicclient/+/755837 | 17:41 |
| TheJulia | We really ought to add uefi entry clean-up | 17:57 |
| *** dking is now known as Guest1801 | 18:05 | |
| *** sshnaidm is now known as sshnaidm|afk | 20:10 | |
| stevebaker | good morning | 21:18 |
| TheJulia | good morning stevebaker | 23:14 |
| *** pmannidi|brb is now known as pmannidi | 23:17 | |
| stevebaker | TheJulia: regarding that backport, I may have forgotten there is a U between T and V ;) | 23:18 |
| TheJulia | Hehehe | 23:40 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
