I've read a bit about that mdadm (and the kernel) now have bad block support but I am having a hard time finding any useful documentation on this subject.
- Is the bad block list for disk X duplicated on all disks in the array?
- Will a disk with bad blocks be indicated differently in /proc/mdstat ?!
- If more than two disks in a raid 6 array has bad blocks, is it possible to set a threshold for how many bad blocks is tolerable?!