Blank screen at boot after update

Hi there,

Let me give you full context on the issue and how it happened. I have a DELL 7440 with dual Boot (Windows 11), with the linux partition encrypted. The login is managed by sssd to give access to an Active Directory, but there are local users as well. The laptop doesn’t have a dedicated GPU. It has sophos (an antivirus) and pakiti (a monitoring tool) installed.

Yesterday morning I updated rockylinux from 9.4 to 9.5, and after that I rebooted it at least twice and all was fine. Then I logged in with a local user and set up a cronjob for pakiti. After this rockylinux hangs after decrypting the disk, with a blank screen (well, there is a blinking cursor on the top left corner). Changing tty’s won’t help at all, although the 6th tty allows me to type stuff (to no effect whatsoever).

After trying different things I managed to get these error messages:

    [FAILED]  Failed to listen on D-Bus System Message Bus Socket.
    [FAILED]  Failed to listen on D-Bus System Message Bus Socket.
    [DEPEND]  Dependency failed for Power Profiles deamon.
    [FAILED]  Failed to start User Login Management.

My best guess is that the login with a local user messed up with the User Login Management and that’s why the greeting screen won’t show up. But in this case the other ttys should be working…

From what I have read, “Failed to listen on D-Bus System…” could mean anything, and the dependency failure could not be the cause for this.

For what it’s worth, I accessed the disk with a live usb and dug around in the log files of sssd. I found these messages in the logs dated after the failure:

    Dynamic DNS update failed
    SRV query failed[11]. Could not contact DNS servers
    Unable to retrieve primary servers: SRV lookup error
    No available servers for service AD_GC
    No available servers for service AD

I tried different things with chroot and sssd seemed to be working fine… but I know very little to nothing about sssd (or ldap for that matter). I removed the sssd cache to no avail.

I am completely out of ideas. Is there any way to fix this?

Many thanks.

If you have root account, you could try starting in single user emergency mode.

Something like, edit the grub menu and tell it to set systemd unit to ‘emergency’ target. let it boot and you should be able to log in as root.

Thanks, the root user is disabled but I’ll try to enable it and boot the single user emergency mode from grub.

Really good tip, @gerry666uk . I was able to get into a bash terminal by adding init=/bin/bash to the linux /boot... grub command.

Now that I am in, I’d like to check all the boot steps to see which one fails. Could anyone point out where to look?

I suspect that the culprit has to do with this error message that keeps appearing:

[FAILED]  Failed to listen on D-Bus System Message Bus Socket.

I fixed the issue: there was a problem with SELinux. I think the labelling system had been corrupted. Booting with the flags selinux=1 enforcing=0 fixed it. In this mode selinux doesn’t enforce policies but it still labels all the files correctly, so when I applied these flags it overwrote the corrupted labelling system. This and this were my main sources of documentation.

Now it boots normally with all the systems working.

It’s great that it’s working again.

It doesn’t really explain what went wrong.

Sounds like the update went ok, first and second reboot both working.

How can creating a cron job break selinux?