hacking SCO....

[Doug Russell]

Oh, I love replying to my own posts....  :)

On Mon, 27 Sep 2004, Doug Russell wrote:

> Try adding    conv=sync,noerror    to your dd line.  If most of the data
> after the defect(s) can be read, you'll end up with an almost complete
> partition which will likely run.  You can then fsck and restore from tape.
>
> for example,
>
> dd if=/dev/daX of=/dev/daY conv=sync,noerror bs=128k

Actually, remove the bs=128k from above (force of habit).  When you're
trying to recover a disk like this, you want the block size to be single
sectors (bs=512, the default) so you get every sector that is readable.

It's slower, but it'll get you a more complete copy if it only skips 1
sector on an error instead of 256.  :)

If you know the defects are only in a certain range, you can get creative
with the   skip   directives to dd and copy most of the disk in larger
blocks, and go back and do the bad part one sector at a time (very handy
when recovering today's large IDE disks).

See the dd(1) manpage for more info.

Later......						<Doug>