fsck reports bad blocks

[radi_1]

Hi,
some file systems are not mounted at boot time,
when trying manual mounts,an error message saying the /dev/vg00/lvolxx needs fsck,
when running fsck(with -p and without any options ,an error is reported:
CANNOT READ:BLK xxxxxx (xxxxxx is block number)

CONTINUE?
when entering y,many block no.'s are displayed.
so it seems there are few bad blocks in the file system.
Is there a way to fsck and mount these file systems?and does the above fsck error means i have some corrupted data?
Am thinkig of using the copy utility provided by the support plus media.
any other suggestions please?
thanks.

[Stephen Keane]

You can try adding the -y flag (means you answer yes to all prompts) to see if the filesystem can be recovered.

Which fileystem(s) is/are affected?

[Bill Hassell · Accepted solution]

Unfortunately, fsck cannot fix any bad blocks. fsck's purpose is to untangle corrupted numbers in the directory and free space table. But it cannot relocate bad blocks and cannot reconsturct missing data due to failed hardware. The disk has gone bad and must be replaced. You should see this documented in /var/adm/syslog/syslog.log

[RAC_1]

What file system type is this??

Did you try fsck -y -o full
Have backups ready??

[HGN]

Hi

If you are seeing bad blocks it might because of a bad disk. It is better to take a full backup of the filesystem and get the disk replaced.

Rgds

HGN

[Kent Ostby]

CANNOT READ: BLK ######

is pretty much alway bad hardware.

if it's a secondary disk, you could try mediainit (depending on your OS version), but then you will have to reload all data to the disk.

mediainit MAY be able to clean things up but in the long run, you're going to want to replace that disk.