From: Enrique Perez-Terron (enrio_at_online.no)
Date: Sat, 22 Oct 2005 21:48:39 +0200
On Sat, 22 Oct 2005 15:32:34 +0200, Nico Kadel-Garcia <firstname.lastname@example.org> wrote:
> <email@example.com> wrote in message
>> No need to zero a partition to discover bad blocks either...
>> Well, not if your filesystem fsck utility has the ability.
>> fsck -c /dev/hda will perform a scan for bad blocks, mark them as
>> and then continue with a full fsck scan for damaged files resulting from
>> those blocks (which then get dumped into lost+found)
>> This works with ext2/3, but I don't think reiserfsck used to have the
> Historically, for me, I've found zero-ing the partition to be slightly more
> reliable. It also doesn't require a filesystem or any management of it.
Curious, how do you measure (or "measure" :) ) that reliability?
Also, how do bad blocks manifest themselves when they occur, using dd?
CRC errors reported from OS to dd and from dd to the stderr? But does
the drive read back the contents of the block after the write to check