I'm in the process of recovering a dead disk and this has brought the following questions.
Background: I tend to run spinrite on the dead drives first. The dynastat process seem to recover data that has bit rot. I then run ddrescue -n /dev/sda /dev/sdb to fast clone the drives. I then run ddrescue -r 1 /dev/sda /dev/sdb to see if there is anything else. Q1:Hard Disk - Are drive access commands queued? Q2: Hard Disk - On modern hard disks is a read enough for a drive controller to detect a sector is going bad and replace it with a spare? Q3: ddrescue - Can dd or ddrescue be used on a single byte of the raw disk to refresh it? ex: ddrescue -n /dev/sda /dev/sda --block-size=1 Q4: ddrescue - Can the disk be mounted and active? Are there any risks? Q5: Os - Is there a way to refresh the disks at the os level (linux/bsd/unix)? Yes, I know I can use raid to swap disks. It is my goal to prevent or detect the error before the drive fails. Nick _______________________________________________ Bug-ddrescue mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-ddrescue
