"Unable to open initial console" due to full disk?

"Unable to open initial console" due to full disk?

Post by Leslie Mikese » Tue, 08 Dec 1998 04:00:00




>Hmmph. I've got a machine running RedHat 5.1 --- it's been running happily for
>months as a web app server. So happily, in fact, that I haven't even logged
>onto the machine in weeks except using my browser.

>Unfortunately, today I come in and the machine isn't responding. I look over
>at the console and cron is *ing about not being able to stat stuff. I
>reboot, and get this ugly "unable to open initial console" message and ...
>kaput.

>My theory is that I ate up all my disk with Apache/custom logs. So I boot to
>a rescue disk and mount my /dev/hda1 and /dev/hda5. Both mount fine, but
>appear to have no files on them at all! Hmm. I run e2fsck on them; many
>errors are found and fixed, but no luck. I cat /dev/hda1 and see that,
>indeed, all my *is there (and LOTS of it looks like web logs).

The disks should be unmounted when you run e2fsck, but otherwise this
sounds like it should have fixed up most problems.  

Quote:>I can't figure out for the life of me how to mount these disks in such a way
>that I can clean them up. I'm pretty new to Linux so I'm hoping this is just a
>lack of knowledge; the alternative is that my machine is just screwed.

If you are sure that you have the partition names right and you still
don't see any files after remounting it probably isn't worth the
effort to try to salvage things.   However, simply letting the disk
fill and rebooting should not cause this kind of problem and it is
extremely unlikely that you would see the same thing on 2 different
partitions at once.  This sounds more like something accidentally or
intentionally removed all of your files (like 'rm -rf /') and the
'unable to open' message is because the /dev entries are gone.

  Les Mikesell

 
 
 

"Unable to open initial console" due to full disk?

Post by John Lewi » Wed, 09 Dec 1998 04:00:00


I just ran into the exact same problem.  Our RedHat 5.1 machine ran
happily for weeks and then at some point hung itself.  I am getting the
same "Unable to open initial console" message on reboot.  When I booted
with rescue boot/root disks, I also fixed a bunch of errors with
e2fsck.  However, upon mounting each of the filesystems at /mnt they are
all empty.  The one exception to this (and the only noticable difference
from  your case) is that my root filesystem contains only the
directories where all the other filesystems would be mounted--empty of
course.  Obviously we can simply reinstall the operating system, but it
would be nice to know what could have caused this and how to prevent it
in the future!
--
John A. Lewis


> Hmmph. I've got a machine running RedHat 5.1 --- it's been running happily for
> months as a web app server. So happily, in fact, that I haven't even logged
> onto the machine in weeks except using my browser.

> Unfortunately, today I come in and the machine isn't responding. I look over
> at the console and cron is *ing about not being able to stat stuff. I
> reboot, and get this ugly "unable to open initial console" message and ...
> kaput.

> My theory is that I ate up all my disk with Apache/custom logs. So I boot to
> a rescue disk and mount my /dev/hda1 and /dev/hda5. Both mount fine, but
> appear to have no files on them at all! Hmm. I run e2fsck on them; many
> errors are found and fixed, but no luck. I cat /dev/hda1 and see that,
> indeed, all my *is there (and LOTS of it looks like web logs).

> I can't figure out for the life of me how to mount these disks in such a way
> that I can clean them up. I'm pretty new to Linux so I'm hoping this is just a
> lack of knowledge; the alternative is that my machine is just screwed.

> Thanks in advance for any help...
> ---S

> Sean Nolan



 
 
 

1. Serial console and "unable to open an initial console"

Running RedHad 6.1, I tried to follow the instrucitons in
/usr/src/linux/Documentation/serial-console.txt to create a serial console.

I get a bunch of "boot-type" messages (hardware addresses, IRQs, etc.) on
ttyS0, then it gets to "Freeing unused kernel memory" and complains "Warning:
unable to open an initial console".  initd then proceeds to write all its
messages to the VGA display.

ls -l /dev/console shows crw--w--w- 1 root root 5,0 and ls -l /dev/ttyS0
shows crw--w--w- 1 root tty 4,64

I think I rebuilt the kernel with CONFIG_CONSOLE_SERIAL=y (and that appears
in /usr/src/linux/configs/kernel-2.2.12-i686.config) and did the other things
the doc called for, and clearly it's partly working, but I must have
overlooked something.

Any suggestions greatly appreciated.

Jonathan Edwards

2. CRC source code

3. "Unable to open a initial console" message

4. I'd like to get network respone time?

5. How to correct "Unable to open an initial console"?

6. Slackware 3.0 installation problems

7. Kernel 2.5.62 "unable to open initial console"

8. Modem card on FreeBSD

9. NFS root : get "Unable to open initial console"

10. Error message "unable to open an initial console"

11. "Unable to open an initial console" on boot

12. Boot problems -- Suddenly "unable to open initial console"

13. "Unable to open initial console"