During an OpenShift upgrade from 4.11.41 to 4.12.18 all cluster operators upgraded except the machine-config because reached the degraded status. All the workers node and one master were upgraded successfully. So the upgrade is stuck on the second master to be updated, MCD error is:

Marking Degraded due to: unexpected on-disk state validating against rendered-master-df8668e3a25e8ec50b1a9cc4cfbb16e7: expected target osImageURL "quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:127670aaa6073e97bd6d919e1b57bd16b60435410b096aeefce7cd64f6c83d24"

Solution https://access.redhat.com/solutions/5598401 was tried different times with no success.

The expected image is pulled successfully.

But the rpm-otree fails with a checksum error:

Jun 10 16:29:13 master-1 systemd[1]: Started /usr/bin/podman run --env-file /etc/mco/proxy.env --privileged --pid=host --net=host --rm -v /:/run/host quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:127670aaa6073e97bd6d919e1b57bd16b60435410b096aeefce7cd64f6c83d24 rpm-ostree ex deploy-from-self /run/host.
Jun 10 16:29:13 master-1 podman[43758]: 
Jun 10 16:29:13 master-1 podman[43758]: NOTICE: Experimental commands are subject to change.
Jun 10 16:29:16 master-1 podman[43758]: error: Invalid checksum of length 0 expected 32

upgrade from 4.11.41 to 4.12.18

Only at customer environment

1.
2.
3.

Upgrade stuck.

Upgrade to complete.

ostree fsck doesn't report any error.