[arch-general] Raid 1 array with mismatched blocks

[Mauro Santos]

I have 2 usb disks that are part of a raid 1 array and after checking
the array for consistency the mismatch count is not zero.

>From all I could find by searching the internet this could be due to
disk failure (not the case), or because of user mistake (most probable).

In this case one can either blindly repair/resync the array and hope for
the best or try to get the correct data before the resync.

My doubt here is what would be the best way to access each disk
individually (it's raid 1 so it should not be hard) and allow me to copy
the correct data, sync the array and write back the correct data.

I was thinking about something along the lines of connecting only one
disk at a time, issue 'mdadm -R -o /dev/md0' to avoid any more changes
between disks, mount the filesystem and check if the files are ok or
not. If it helps, I do have a md5sum of all files and I know which files
I have written that may not be consistent across disks.

-- 
Mauro Santos