HELP!!!

I am not well versed in linux at all. I managed to get my system configured, and it has been running very nicely as a server for the past 3 weeks.

Yesterday, while transfering files, it crashed. When I try to reboot I get - no such file or directory. I fried doing a fsck but it too found nothing.

I am running RH 7.1.

As it is going through the boot cycle, it fails and gives me the following

mounting proc: file system dup2:bad file descriptor
Configuring kernel parameters: dup2: bad file descriptor
Setting Clock: dup2: bad file descriptor

then goes into
/etc/rc.sysinit : dev/null; : Input/output error - a few times and then tells me

An error Occurred during the file system check. Dropping you to shell; the system will reboot when you leave the shell. Give root password for maintenance.

If I go through the installation screens, and get to the drive partitioning are it looks like when I try to view the partitions that the names ie /home /boot etc are not there. Does that make sence.

HELP!!!!

Chris

bad news.

dup2 "Bad file descriptor" in essence means "cant check, cant mount, cant read". Processes spawned tru rc.sysinit (the boot script) can't get file descriptors off of /proc, cuz the partition /proc is on isnt fsck'ed, so /proc can't be mounted in read-write mode.

try booting off the install cd with "linux rescue" or use a bootfloppy, or use a bootdisk like tomsrtbt.
run fsck with the -f (force) on /dev/(all partitions).
if any errors come tru, plz tell us exactly what.

*just my curiousity, did u have a good security scheme on ure box?

HTH

OK,
I tried booting off the CD -and did the rescue mode

I am going through the hundreds of errors in my hde6 (which is my /home)

it gets to a point and then spits out a sequence of increasing numbers - ie. -5646026 -5646027 -5646027, etc.

What do I do - I was able to stop it with a <CTRL> C and then when i did the fsck again - same thing - should I let it continue??

Chris

NEW INFO

Got through all of the fsck ... here are the results

HDE1 /boot clean

HDE5 swap - The super block could not be read or does not describe a correct ext2 file system. Bad Magic Number in Super-Block while trying to open /dev/hde5

HDE6 /home clean
HDE7 /usr clean
HDE8 / clean
HDE9 /var clean

Now when I re-boot I get the dup2 error, but now I get
/etc/rc.sysinit : dev/null; : Read Only file system instead on inut/output error

Chris

Ok, kewl.
U didnt know, but Ext2fs has more than one superblock.

Now since this is /swap we wont bother (no data on it). Boot off some disk and delete & create a new spaw in its place.

BIG PROBLEM,

I killed the /swap partition and it re-assigned hde numbers to the remaining drives

Now i cannot get anything from my /home directory - I get the error that I had from the swap directory.

At this point I only want to save the data - I don't care if I have to re-install Linux.

Chris

I'm also getting this on about 10 of my machines. Is it hardware? I have a auto update script running which compares RPM's on the local system to updates.redhat.com and if updates.redhat.com is newer it does a rpm -U --nodeps package name. Could this be a problem?

Here's my quick fix for this but it sucks because it keeps happeneing again and again.

boot off toms boot disk

mount /dev/hda7 /mnt (hda7 = / partition on my machine)
cd /mnt/dev
rm null
mknod null c 1 3
chmod 0666 null
cd /
umount /dev/hda7
reboot


if you have the redhat cd you can do this 2 by just gettign into the install and then pressing alt+f2 and it should bring you to a sh term.