Cannot recover data, raid problem

[enid]

Hello to all,
I have a server system which is configured with raid5 software (mdadm) . It is composed by 4 HDD SCSI each 37GB.
I have installed on it RedHat Linux 9, in the default configuration (so /boot a separate partition and the other ones with LVM)

My problem is that one of the partitions configured with raid 5 (Linux raid autodetect) i believe that it is on degraded or faulty state,

my system does boot up but it shows on the screen "Kernel panic , cannot find initrd"

I have very sensible data in there, and the problem is that I didn't make a backup. I want to just retrieve some of the data files there.

I managed to boot into the rescue mode with CD1 of RH9 and the commands which I tried to do:
# mdadm --examine --scan /dev/sda2 /dev/sdb1 /dev/sdc1 /dev/sdd1 > /etc/mdadm.conf
#mdadm -A -s
#cat /proc/mdstat --->>> it showed the raid array but only with three partitions exactly /dev/sda2 /dev/sdb1 /dev/sdc1

I have lvm so I entered:
#vgchange -a y vol0 (because volume group wasn't active)
#mount /dev/vol0/lv0 /mnt/lv0 -->> which I think is the / partition, and I couldn't copy my files (not all of them but the important ones),

it gave some I/O errors in copying (transfering with scp in another server).

so I gave:
#fsck /dev/vol0/lv0 and responded with Y (yes) to all of the questions (a huge number of inode problems)

and now again I can't copy my files....

My question is how can I recover my files, is there another way?

Your help would be very appreciated!

Thanks in advance.

[Rubberman]

I suspect that other than using a sector editor to find specific file data you are SOL. Given the relatively small size of this RAID, not keeping a full backup on an external drive was, in hind sight, not prudent...

FWIW, a lot of people seem to think that RAID-5 and such are "magic bullets" that will never lose data. Not so, though when they work correctly, a single disc failure is recoverable. However, some sorts of failures, such as corruption of the inodes, directory structures, etc are not recoverable. Also, RAIDs have a tendency for more than one drive to fail at about the same time, often because the drives tend to be from the same manufacturing batch and thus have about the same life expectancy. Rule 1 for RAIDs: if one drive goes bad, back the volume up and replace them all. Personally, I prefer mirrored drives, and I make sure they are from different manufacturing batches so that I have a better hope that if one goes bad, the other won't before the replacement is re-mirrored.