Problem
Yesterday, one of my CoreOS systems refuses to boot after an update.
Deployment 0 won’t boot, while Deployment 1 works as normal
Details
Here’s a screenshot of the behavior from a virtual machine created following the general steps listed in the Replication section further down this post.
Here’s the output of rpm-ostree status and bootc status of the affected physical system
#rpm-ostree status
State: idle
AutomaticUpdatesDriver: Zincati
DriverState: active; periodically polling for updates (last checked Sat 2026-07-04 16:07:34 UTC)
Deployments:
ostree-image-signed:docker://quay.io/fedora/fedora-coreos:stable
Digest: sha256:40f8190851a49d9a02c89f91171ccc9ec59b9090c96d551b8aa36a7641744f79
Version: 44.20260607.3.1 (2026-06-22T00:27:36Z)
LayeredPackages: avahi wsdd
● ostree-image-signed:docker://quay.io/fedora/fedora-coreos:stable
Digest: sha256:075bf47207b42f01b0bdf9c9d6705ba6dc6a6e7e87bc0b474c87f0d444357752
Version: 44.20260523.3.1 (2026-06-08T22:00:22Z)
LayeredPackages: avahi wsdd
#bootc status
● Booted ostree
Commit: cef08b5add4e66dc5c9ffbb51d7136e638f3d9092f6a1342354bf5856d82e231
Rollback ostree
Commit: 97e25d137182b587b683da7550f1f605a0e97695306a893fd07a7de3827e3316
Replication
- Create the ignition file with the following content
-
#test.yml variant: fcos version: 1.6.0 passwd: users: - name: fedora # Password is same as username groups: - wheel # "wheel" adds this user to the admin group password_hash: $6$aK.HN1rvo86y1s.8$OZTkf3NVtIycyRsicLu1mOuxaeqIimMQLhuV2fq/6Cx9sROPxYe2tbyJ4fp7HcgkjojN.LpC1FsQaTUpp1zVf1 boot_device: luks: tpm2: true discard: true storage: disks: - device: /dev/disk/by-id/coreos-boot-disk partitions: - label: root number: 4 size_mib: 12000 resize: true - label: var size_mib: 0 luks: - device: /dev/disk/by-partlabel/root name: root wipe_volume: true options: ["--hash","sha512"] - device: /dev/disk/by-partlabel/var name: var wipe_volume: true clevis: tpm2: true discard: true options: ["--hash","sha512"] filesystems: - device: /dev/mapper/var path: /var format: btrfs wipe_filesystem: true with_mount_unit: true
-
- Run the code block below can be used to embed the ignition above into a CoreOS installer disk image to be used on a bare metal system or virtual machine.
-
sudo dnf install coreos-installer butane rm -f ./test.ign butane -o ./test.ign ./test.yml image=$(coreos-installer download -s stable -f iso 2> /dev/null) # Automatic boot device selection ## `--dest-device` is set to /dev/vda for the first virtio drive connected to UEFI Secure Boot + TPM2 enabled QEMU virtual machine ### Warning: the drive specified by --dest-device will be wiped coreos-installer iso customize --dest-device /dev/vda --dest-ignition ./test.ign -o ./fcos-test-headless.iso ${image}
-
- Install from the ISO made in the previous step and reboot.
Notes
I’ve tested both the stable and next streams and both get booting forever as shown in the image in the Problem section
Used the latest iso images provided by coreos-installer from CoreOS Stable and Next streams as listed below.
# Stable stream
fedora-coreos-44.20260607.3.1-live-iso.x86_64.iso
# Next stream
fedora-coreos-44.20260621.1.1-live-iso.x86_64.iso