CentOS 6 – Lost Console Login

Home » CentOS » CentOS 6 – Lost Console Login
CentOS 7 Comments

Just had a problem with CentOS 6.5 x86_64 after the last kernel upgrade. It looks like the boot sector got hosed somehow and the system was unbootable. I was able to fix it with a boot to an install image USB and running grub-install. System is now bootable.

However, I’ve lost the system console. No login prompt is provided at the console.

The system boots and is remotely accessible. Services seem just fine. I
don’t see any error messages after booting, either dmesg or
/var/log/messages.

Can anyone tell me what packages provide console login so I can try a reinstall.

Thanks.

7 thoughts on - CentOS 6 – Lost Console Login

  • Did I respond, init 3 – text only?

    I also recalled last night that I may have interrupted an selinux relabel process. So this morning I did “touch /.autorelabel” and then rebooted. Alas still no text login.

  • Okay, it looks like upstart isn’t running /etc/init/start-ttys.conf

    After running “initctl list” and seeing that tty was in stop/waiting, I
    manually ran:

    initctl start tty TTY=/dev/tty1

    six times to get the six usual ttys. After looking at start-ttys.conf I
    guess I could have just run that once instead. Now “initctl list” looks like:

    $initctl list rc start/running, process 1507
    tty (/dev/tty3) start/running, process 6169
    tty (/dev/tty2) start/running, process 6167
    tty (/dev/tty1) start/running, process 6165
    tty (/dev/tty6) start/running, process 6176
    tty (/dev/tty5) start/running, process 6174
    tty (/dev/tty4) start/running, process 6172
    plymouth-shutdown stop/waiting control-alt-delete stop/waiting rcS-emergency stop/waiting readahead-collector stop/waiting kexec-disable stop/waiting quit-plymouth stop/waiting rcS stop/waiting prefdm stop/waiting init-system-dbus stop/waiting readahead stop/waiting splash-manager stop/waiting start-ttys stop/waiting readahead-disable-services stop/waiting rcS-sulogin stop/waiting serial stop/waiting

    Which looks right to me. At the moment I’m at work and this host is at home so I can’t check the console.

    I looked at start-ttys.conf:

    start on stopped rc RUNLEVEL=[2345]

    That looks right. Can anyone tell me why it’s not running?

  • Starting the ttys that way didn’t work. I was unable to login. So I
    stopped them and ran:

    initctl start start-ttys

    Which *seemed* to work better. It loaded all six ttys. I was able to login at the console. I see an entry in /var/log/secure for my login.

    *But* the terminal session is all messed up. I can run any commands. They are accepted but do not execute. The first character of the command is echoed back but that’s all. In addition, during login, the password is fully echoed back on the screen instead of being hidden the way it should be.

    As I said, I can login through SSH and work just fine. So the user environment and all user tools are intact.

    But something else isn’t running at boot time interfering with proper login at the console and execution of the ttys.

    Does anyone have any thoughts?

    Thanks.

  • I had hoped someone here might be able to shed a bit more light on the startup processes for the console.

    After educating myself a bit, I reinstalled a few packages that seemed relevant:

    yum reinstall pam
    yum reinstall initscripts
    yum reinstall bash

    But the console ttys are *still* not loading at boot time. I have to run
    /etc/init/start-ttys manually.

    *And* tty1 is still messed up. The password is displayed in clear text and not blanked out. Sometimes the session is fine. Most times the session refuses to run any commands at all.

    Now weirdly tty2 through tty6 are all just fine. The password is blanked as expected. The session executes just fine and dandy.

    Once again I’ll ask if anyone has any experience with Upstart. Can anyone tell me why /etc/init/start-ttys isn’t running at boot time?

    Thanks again for your help.