[lug] recovering data

[Nick Golder]

On 2009-09-22 16:25 -0400, Gordon Golding wrote:
> 
>  I have a data server with 2 drives on it.
> 
> I was keeping a copy of data on Drive1(system drive) and generation
> copies on Drive2
> 
> Drive1 almost filled up, so I did a last big tar of Drive1data to
> Drive2, in case somebody had changed stuff since the last official
> backup.
> 
> Tar took forever, the tar file on drive2 had problems and couldn't be
> copied to a usb drive.
> 
> Then: segmentation fault and WON'T reboot - has disk errors.
> 
> I took Drive2 to another machine, it reports disk errors.
> 
> I don't want to just be experimenting - hope somebody has more
> experience than I - I am working first on Drive2.
> 
> fsck says error.? When I say: Ignore, it finds many more.
> 
> Is the only thing to do a "fsck -a" and hope for the best?
> 
> 
> [root at icare ~]# fsck /dev/sdb1
> fsck 1.41.4 (27-Jan-2009)
> e2fsck 1.41.4 (27-Jan-2009)
> /dev/sdb1 contains a file system with errors, check forced.
> Pass 1: Checking inodes, blocks, and sizes
> Error reading block 601 (Attempt to read block from filesystem
> resulted in short read).? Ignore error<y>? yes
> 
> Force rewrite<y>? no
> 
> Error reading block 602 (Attempt to read block from filesystem
> resulted in short read).? Ignore error<y>? 
> 
> [root at icare ~]# fsck /dev/sdb1
> fsck 1.41.4 (27-Jan-2009)
> e2fsck 1.41.4 (27-Jan-2009)
> /dev/sdb1 contains a file system with errors, check forced.
> Pass 1: Checking inodes, blocks, and sizes
> Error reading block 601 (Attempt to read block from filesystem
> resulted in short read).? Ignore error<y>? no
> 
> Error while iterating over blocks in inode 7: Attempt to read block
> from filesystem resulted in short read
> e2fsck: aborted
> 

It might be worth using ddrescue to 'clone' Drive 2  off to a new
drive and do any invasive work on that new drive.  At that point
your better off either running fsck on the new drive with the -a or 
-y switch and redirect the output so you can review later.


-- 
Nick Golder