I am having problems getting my server to boot up because as soon as I start to mount filesystems, it says that it is having problems because it can't write to /dev/null. I'm not sure why exactly this would be necessary, but oh well. It doesn't work. i can't even get my system to get to a workable point. It freezes in boot when it fails to mount all of the filesystems and then puts me into shell so that I can fix the problem. The thing that mystifies me is that it is still able to read the filesystem because I can see everything on the hard drive when i get into shell. Any suggestions on how to fix this? I'm completely stumped. This is the first time that I have ever had anything like this happen to me. Thanks for any help that anyone can provide.

I'm running Fedora Core 1 with a few updates
Athlon 1.4GHz
512 MB Ram
40 GB hard drive (100 mb /boot, 1 gb swap, the rest for /)
60 gb hard drive (/home)
dual NICs
ATI graphics card (although I feel that this information might be irrelevant)

Thanks

-Jalsk

Try this from your shell:
This is from the man page

as root exec 'chmod 666 /dev/null' and try again

That didn't work. Here is exactly what I'm getting:

/etc/rc.d/rc.sysinit: line 88: /dev/null: Read-only file system
/etc/rc.d/rc.sysinit: line 91: /dev/null: Read-only file system
Configuring Kernel Parameters: Dup2: Bad File Descriptor
[ FAILED ]
Setting clock (localtime): Mon May 10 07:33:20 MDT 2004 dup2 bad file descriptor
[ FAILED ]
Loading Default keymap (us): /etc/rc.d/rc.sysinit: line 159: /dev/null: Read-only file system
[ FAILED ]
Setting hostname mercury: dup2: Bad file descriptor
[ FAILED ]
/etc/rc.d/rc.sysinit: line 183: /dev/null: Read-only file system
Initializing USB controller (usb-uhci): dup2: Bad file descriptor
[ FAILED ]
/etc/rc.d/rc.sysinit: line 194: /dev/null: Read-only file system
Checking root file-system
dup2: Bad file descriptor
[ FAILED ]

*** An error occured during the file system check.
*** Dropping you to a shell: the system will reboot
*** when you leave the shell.
Give root password for maintenance
(or type Control-D to continue):

please post output of:

ls -la /dev/null

try the following:

mount -o remount, rw /
rm /dev/null
mknod -m 666 /dev/null c 1 3

Last note help me to fix this problem. I don't know how It happend, but I was "playing" with users accounts in /etc/passwd and changed UID.

Thanks for help me

That last reply fixed the same problem for me.

My boss was ready to toss me AND the Linux server out.

Have the same problem as above.

tried your suggestion but OS tells me "rm cannot remove directory '/dev/null': Is a directory"

tried "chmod 666 /dev/null" and it says 'read-only file system'.

Any more suggestions?

The funny thing about the situation I am in is that I was only editting the samba.conf file and setting up the OS to be a domain controller. In doing this I was following directions to setup a computer account with home directory of /dev/null.

I then tried to restart smb and it failed. I then restarted and it has been rooted since???????

No user should have /dev/null as a home directory because /dev/null is never a directory, it is a device file. When you made the account, you probably removed /dev/null and recreated it as a directory. You need to rmdir /dev/null and recreate it using the directions given upthread. Any directions telling you to make a user with a home directory of /dev/null are flawed. Use /fake or some other nonexistent directory for a user that is not allowed to log in and will own no files.

cannot delete directory as it says it is not empty.

try ls /dev/null and it returns nothing (indicating that there is no contents in the directory)

How do I deal with these conflicting reports?

cd /dev
rm -rvf null/

And for god's sakes... be careful with that command. More sysadmins have killed their systems with than than {cr,h}ackers ever will.


The reason it says the directory isn't empty is because it probably has some hidden files.


Now... follow the mknod and chmod commands listed above... then reboot.

EDIT: While it is not wise to have a usr directory of /dev/null it can be done. As long as that directory stays as a device. Any files saved will be thrown away and any attempts to read files will find none. But it is wiser to create a folder like, /tmp/emptydir and make that the home directory.

I had the exact problem described here at the weekend. Being a bit of a Samba newbie, I bought the official "Samba Howto & Reference Guide". This, along with several online tutorials I found, will advise you to create machine accounts (in my case W2K domain clients) with the home directory of /dev/null. Now my system dies on bootup (RH9) with the same errors described above, doh! I will try the solutions suggested here tonight, thanks for the help!

This conflicting info online and in print really does not help newbies...

Hello,
I was wondering if anyone had found a permanent solution for this problem. I am running RH9 and about 4 months ago, starting experiencing the same problem on boot (dup2:bad file descriptor). Using the solution of:
mount -o remount, rw /
rm /dev/null
mknod -m 666 /dev/null c 1 3
works great, but on the next reboot - its the same thing all over again. I don't shut this machine down very often, but it would be nice to have it just reboot.

Thanks,
Andy

I actually didn't fix the error per se, I just reformatted the machine. I did, however, after a few minutes of searching on google find a site that claims to work, although i haven't tested it yet. Hope this helps!

http://www.jroller.com/comments/tdak...ile_descriptor

-Jalsk