I have a Ubuntu 11.10 Server 64-bit. It was setup with a Raid1 /dev/md0. /dev/sda1 failed completely. /dev/sdb1 is still there and booted as a degraded array. I shutdown the server to remove the faulty drive and replaced with a 160gb drive. (The two original drives were 80gb).
I used sfdisk to copy the partitioning scheme from /dev/sdb to /dev/sda - went ok
used mdadm to add /dev/sda1 back to the raid /dev/md0
at 46.5% of rebuilding the raid, it fails stating an I/O read error on /dev/sdb1
running bad blocks on /dev/sdb1 I found 65 bad blocks.
How can I mark those bad blocks on /dev/sdb1 since it is a linux_raid_member, rebuild the array onto /dev/sda1, That way I can fail /dev/sdb and replace it, and rebuild again?