Disk Problems

matrixcubed · 09-24-2005, 10:48 PM

Hello all,

I thought I'd do a bit of checking around before I go doing anything dumb...

This evening the primary slave disk on my little Redhat 9 server started making disk-access noises when I tried accessing some files on it via FTP, then the web.

I logged in via SSH, and checked the 'du' table, which looked fine. When I cd'd to the offending mount-point of the sec.slave, and did an ls, I got "0 total" response, instead of the expected directory listing.

I tried unmounting the disk, but got a 'device is busy' error.

I haven't attached a monitor to the server to see any error messages that have come up, and I don't know which logfiles might tell me something bad is going on.

I'm assuming that something software-wise messed up, and I'd like to restart the system and hope to hell the file-system fixes it automagically, but have not encountered this behavior before, and was wondering whether there is a behavior pattern and resulting data loss or something similar which I can expect? This is a fairly new disk (maybe 3 months old), so shouldn't be experiencing data loss 'yet'. Additionally, the server was not having problems earlier today.

Any suggestions are appreciated.

addy86 · 09-25-2005, 02:12 AM

Quote:

Originally posted by matrixcubed
I tried unmounting the disk, but got a 'device is busy' error.

This usually means that at least one program still has its current working directory in the mounted partition (I think this also applies to shells).

Quote:

I haven't attached a monitor to the server to see any error messages that have come up, and I don't know which logfiles might tell me something bad is going on.

/var/log/messages, /var/log/syslog

matrixcubed · 09-25-2005, 06:01 AM

Killed the processes which would have access to the drive. Unmounted and remounted. Now I get...

Code:

mount: wrong fs type, bad option, bad superblock on /dev/hdb1,
       or too many mounted file systems

After digging, came across dmesg, which tells me...

Code:

end_request: I/O error, dev 03:41 (hdb), sector 104
hdb: drive not ready for command
hdb: status error: status=0x00 { }

hdb: drive not ready for command
hdb: status error: status=0x00 { }

end_request: I/O error, dev 03:41 (hdb), sector 4200
hdb: drive not ready for command
EXT3-fs error (device ide0(3,65)): ext3_readdir: directory #2 contains a hole at offset 0
hdb: status error: status=0x00 { }

end_request: I/O error, dev 03:41 (hdb), sector 4240
hdb: drive not ready for command
hdb: status error: status=0x00 { }

end_request: I/O error, dev 03:41 (hdb), sector 312576512
hdb: drive not ready for command
hdb: status error: status=0x00 { }

end_request: I/O error, dev 03:41 (hdb), sector 2
hdb: drive not ready for command
EXT3-fs: unable to read superblock

Sigh... Currently googl'ing for a fix.

matrixcubed · 09-25-2005, 06:13 AM

Thank the Gods of Rebooting ... it is working again!