indeed /usr/share/mdadm/checkarray exits with an error message when it encounters a redundant MD device in auto-read-only state. this state is normal for assembled MD device which haven't yet been written to, such as the swap area mentioned by Philipp Kern.
any write operation to the respective device would take it out of auto-read-only (in my case it was a swapoff, mkswap, swapon which did the trick), but one can't expect checkarray to figure out what sort of write is safe on the device. i think that checkarray should look for and skip any md device it finds in the auto-read-only state. hope this helps, adc -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
