)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":7118,"name":"Ian Wienand","email":"iwienand@redhat.com","username":"iwienand"},"change_message_id":"fc34cac2e6484fdb914759ba4da421a1520cab33","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"11fb08f7_14676adf","updated":"2021-12-14 07:59:15.000000000","message":"I would really like to not revert this, as I still haven\u0027t seen anything that suggests that /etc/default/grub is not the correct way to set the kernel command-line variables (indeed, it works on Fedora 35, and I don\u0027t think anything changed in this since Fedora 34/9-stream).   Looking at \n\nhttps://logserver.rdoproject.org/13/36713/6/check/periodic-tripleo-ci-centos-9-ovb-1ctlr_1comp-featureset002-master/b2300e8/logs/undercloud/home/zuul/overcloud-hardened-uefi-full.log.txt.gz\n\nI think that the command-line arguments aren\u0027t getting overwritten, similar to what was happening with Fedora 35.  This was because the machine-id was missing, and I suspect something similar.  I will comment in the bug","commit_id":"936106d5dccb0d5075c7cfbed3ad9879f66bea6e"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"0e736f5384a11259d2315f97ac4943f8cf3712bb","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"de5ec7ff_ca69f921","in_reply_to":"11fb08f7_14676adf","updated":"2021-12-14 14:35:41.000000000","message":"This is happening on Stream-9 initial image, which means the target as far as this change hasn\u0027t moved.\n\nBut perhaps this is just a first time in CI as opposed to a local machine in UEFI mode?\n\nMy impression is we believe grub2-mkconfig should be updating/overwriting the config, but I believe this only happens depending on the mode of the machine, and these jobs are running in bios mode which behaves a little differently from what has been observed.","commit_id":"936106d5dccb0d5075c7cfbed3ad9879f66bea6e"},{"author":{"_account_id":7118,"name":"Ian Wienand","email":"iwienand@redhat.com","username":"iwienand"},"change_message_id":"8bea863e09eeda15c1ffd6fcf2e5be1bf08ce372","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"e76f653d_91154d39","in_reply_to":"33753a09_46f8725c","updated":"2021-12-14 20:23:28.000000000","message":"My feeling is this is the centos image V the centos-minimal image.  \n\nThe situation is confusing between grub-mkconfig and grubby.  [1] sort-of discusses it.\n\nI am still fairly convinced that grub2-mkconfig is the right tool for dib to be using in a generic sense here.  This spans platforms, releases and does not require bifurcating the bootloader path -- leaving the grub2-mkconfig path broken and not matching correctly led to the Fedora 35 failures which were equally hard to unravel (and we can\u0027t just revert this as-is, without fixing that too).\n\n9-stream grub2-mkconfig has code intended to update the BLS kernel parameters [2] -- and we know it works because it updates our centos-minimal 9-stream kernel correctly.\n\nThis is why I would like to root-cause *why* it\u0027s not working here.  machine-id mismatch?  something set by the upstream image in /etc/default/grub?\n\nThis is complicated by the fact that our boot tests happen on Ubuntu Focal, which doesn\u0027t have kernel support to be able to open the xfs on centos-9 .qcow2 images -- i.e. we can not do an end-to-end test on the centos (not -minimal) element.\n\nI\u0027ll keep looking, but this is my current thinking. \n\n[1] https://docs.fedoraproject.org/en-US/fedora/f34/system-administrators-guide/kernel-module-driver-configuration/Working_with_the_GRUB_2_Boot_Loader/\n\n[2] https://gitlab.com/redhat/centos-stream/rpms/grub2/-/blob/c9s/0074-Add-BLS-support-to-grub-mkconfig.patch#L230","commit_id":"936106d5dccb0d5075c7cfbed3ad9879f66bea6e"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"d9c106920e418227e7b896f93e487074bdad8bb2","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"33753a09_46f8725c","in_reply_to":"de5ec7ff_ca69f921","updated":"2021-12-14 18:56:14.000000000","message":"It is likely worth noting, Steve Baker did run into some inconsistencies which drove the explicit grubby command to force the update across all kernels. That is really and truly what is needed it seems.","commit_id":"936106d5dccb0d5075c7cfbed3ad9879f66bea6e"}]}