Linux Mounting problem

Hello,
I am using Redhat Linux 7.1
During reboot, i get the message " Mounting / as readonly" And, it enters into maintenance mode...( & all other steps fail -
/proc not mounted, swap not mounted, fsck fails)
I did the following :
mount -t proc proc /proc
fsck /dev/hda1
The following error messages : ERROR : Couldn't open /dev/null
(Read-only file system)

It goes into a never-ending loop, and never i am able to recover from
this problem.

Has anyone come across such a problem? How to tackle it? Do we need to use a bootdisk, to get into the read-write mode of root
filesystem ?
How to change root filesystem from read-only to read-write?

This will be of very great help to me and my team.

Thanks in advance,
Venkatesh.

First, verify that /dev/null is a character device, major=1, minor=3. file /dev/null.
It may have gotten changed to some real file because of some errors.

In maintenance mode do:

/sbin/mount -n -o remount /dev/hda1 # Mount r/w, no write to /etc/mtab

/bin/rm /dev/null             # Delete it
/bin/mknod /dev/null c 1 3    # Make a new one
/bin/chmod 777 /dev/null      # Accessible by everyone in all modes

/bin/chown root.sys /dev/null # Standard ownership
/sbin/umount /dev/hda1 # Now unmount it
/sbin/mount /proc /proc -t proc # You can mount it if you want
/sbin/e2fsck -f /dev/hda1 # Fix the disk
... After you fix the errors ...
exec /sbin/init auto (or exit if from a startup script)

Cheers,
Dick Johnson

Penguin : Linux version 2.4.1 on an i686 machine (799.53 BogoMips).

I was going to compile a list of innovations that could be attributed to Microsoft. Once I realized that Ctrl-Alt-Del was handled in the BIOS, I found that there aren't any.

He is getting these errors because his /etc/rc.d/rc.sysinit is shuttling all output from the commands in the script to be written to /dev/null and his FS is mounted as read-only. It never gets mounted as read-write because of fsck failing to approve the check with an OK status. If 1-2 boots in rescue mode (or as "linux single") don't fix the errors, the last step (e2fsck) should be run and should fix the error, unless there is serious drive corruption.

Regards,
Jim Roland, RHCE

[SNIPPED...]

But you can write to a device that exists on a R/O disk! If the write is not allowed, then the device-file has become corrupt and "seems to be" a file, instead of a device..

Script started on Fri Aug 31 08:04:09 2001

# mount -o ro /dev/sdc1 /alt
# cd /alt
# ls
System.old  boot   dos	   exports  initrd  lost+found	proc  tmp  vmlinuz
alt	    cdrom  etc	   guru     laptop  mnt		root  usr  vmlinuz.old
bin	    dev    export  home     lib     opt		sbin  var
# >foo
bash: foo: Read-only file system
# cd dev
# >null
# ls -a >null
# file null
null: character special (1/3)
# pwd
/alt/dev
# cd ~
# umount /alt
# exit
exit
Script done on Fri Aug 31 08:05:25 2001

Cheers,
Dick Johnson

Penguin : Linux version 2.4.1 on an i686 machine (799.53 BogoMips).

I was going to compile a list of innovations that could be attributed to Microsoft. Once I realized that Ctrl-Alt-Del was handled in the BIOS, I found that there aren't any.

It will enter maintenance mode if fsck was unable to fix errors as a previous boot-up. Boot with your RH71 cd, enter "linux rescue" at the "boot:" prompt, and your system will be mounted under /mnt/sysimage. It may delay for a little while while coming up if it's re-running fsck. After you have a prompt (navigate throught the text dialog boxes first), run fsck and you can fix the partition or inode errors.

Regards,
Jim Roland, RHCE

Sounds like devfs is compiled in, but not mounted. Can you verify?