So now that I have explained how ddrescue 1.19 operates, let’s look at
some practical commands to make it work. Understand that these
commands are what I feel would work best, which means that they are
somewhat of an opinion. But my “opinion” is based on simulated
benchmark testing that I have done, so I base my commands on the
knowledge I have learned from that testing.
The first thing we need to figure out is what do we know about the
drive? The fact is at this stage we really don’t know anything! We
probably know (or suspect) that the drive has errors. But until we try
reading it, we don’t know the extent or nature of the errors.
So let’s start with a single command that would attempt to do the best
it can without the user making any further decisions. Obviously you
must change your commands for the proper input and output device.
[code]
ddrescue /dev/sda image.dd image.log --skip-size=5Mi,50Mi
[/code]
This command will allow for a larger than default starting skip size
so it will skip out of bad areas faster, but limit the maximum to a
reasonable number so it doesn’t go crazy and skip too far ahead which
could miss good data. Both the initial and maximum skip size are up
for discussion, as they are just what I think may be best, but it is
very difficult to know what is really best.