After upgrading systemd-udev from 241-4.gitcbf14c9.fc31.x86_64 to 242~rc3-1.fc31.x86_64 eth0 network interface stops being renamed to ens3. This results into nonconfigured network and making the machine unreachable over a network. The network interface is a virtio device handled by virtio_net kernel module. There is another strange thing that the kernel module is not loaded automatically. Loading it manually creates eth0 interface, but udev seems not renaming it. After a reboot, "systemctl list-units --failed" reports these failed services: network.service (because the renamed ens3 interface does not exist) systemd-udevd.service (exit code=exited, status=226/NAMESPACE; follwed by "start request repeated to quickly") systemd-udevd-control.socket (service-start-limit-hit) systemd-udevd-kernel.socket (service-start-limit-hit) There seems to be some issue with namespaces in udevd daemon. Please note that I observe this error message from DNF on the upgrade: Running scriptlet: systemd-udev-242~rc3-1.fc31.x86_64 14/14 Failed to send reload request: No such file or directory warning: %triggerin(systemd-udev-242~rc3-1.fc31.x86_64) scriptlet failed, exit status 1 Error in <unknown> scriptlet in rpm package systemd-udev That maybe depends on this scriptlet error: Running scriptlet: systemd-libs-242~rc3-1.fc31.x86_64 1/1 warning: Unable to get systemd shutdown inhibition lock: Failed to connect to socket /run/dbus/system_bus_socket: No such file or directory That's caused probably by a broken DBus unit (bug #1623781).
*** This bug has been marked as a duplicate of bug 1697370 ***
After disabling SELinux enforcing I was able to upgrade, reboot and udevd started succesfully. Then I enabled enforcing and tried to downgrade, reboot, upgrade, and reboot to reproduce this issue. But I was unable. Now any upgrade to systemd-udev-242~rc3-1.fc31.x86_64 passes for me. I have selinux-policy-3.14.4-9.fc31.noarch.