[Discuss] ddrescue, anyone?

[Greg Rundlett (freephile)]

Hi Dan,

The two-pass examples do a "shallow" then "deep" rescues as you mention.
 The second pass uses the contents of the logfile generated by the first
pass to be more efficient.  And, since rescuing a large partition can take
so long, re-trying failed sectors repeatedly on a large drive can take
"forever".  So that's why they recommend the first pass with -n -- it gives
you the most bang for your buck, and then you can dive in with a full
rescue.

~ Greg

Greg Rundlett


On Mon, Jul 15, 2013 at 1:42 PM, Daniel Barrett <dbarrett at blazemonger.com>wrote:

>
> I used ddrescue (http://www.gnu.org/software/ddrescue/ddrescue.html) for
> the first time yesterday, as my boot disk developed errors and I needed to
> copy the contents to a fresh drive.
>
> Can anyone explain why the examples in the manual all run ddrescue twice?
>
>
> http://www.gnu.org/software/ddrescue/manual/ddrescue_manual.html#Examples
>
> The pattern seems to be:
>
>   $ ddrescue -f -n /dev/one /dev/two logfile
>   $ ddrescue -d -f -r3 /dev/one /dev/two logfile
>
> but the reasoning is never explained, and plenty of other sources on the
> say simply to run it once as "ddrescue /dev/one dev/two logfile".
>
> I believe that the first invocation avoids problematic areas of the disk
> (-n), whereas the second invocation retries errors up to 3 times. But I
> don't understand if these two invocations are independent, i.e., they will
> both take the full 10 hours for a 1 gig drive, or does the second
> invocation
> somehow make use of the result from the first run and go more quickly?
>
> I ran only the first invocation, e2fsck'ed all partitions, and booted, and
> everything worked fine.
>
> Thanks for any advice/understanding.
>
> --
> Dan Barrett
> dbarrett at blazemonger.com
>
> _______________________________________________
> Discuss mailing list
> Discuss at blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>