I am just copying a disk with a number of bad blocks near the end.
Initially, the Remaining time estimate was 9 hours, but once it hit
the bad block, the estimate is all over the place, from 1 hour to 15
days. I guess it is based on an average of how much has been
transferred in the last x seconds. If they are good blocks, this will
be a high number. If they are problematical, it may be zero.
So my suggestion is that maybe basing it on the average amount
transferred over the last x seconds may not be helpful when there are
a lot of bad blocks in an area of the disk. (I would guess that bad
blocks are more likely clustered together than scattered randomly
over the disk.)
Perhaps some calculation that incorporates how many failed reads
there have been in the last y seconds as well as how much has been
transferred may be more useful?
ofc, none of this really means anything, since the copy will take as
long as it needs. It just seems that it might be possible to give a
better estimate knowing that there are a series of bad blocks being
processed.
Cheers
David
_______________________________________________
Bug-ddrescue mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/bug-ddrescue
