raid 1 problems fedora core 3 - fsck fails
Going a little nuts here.
mdadm --create --level 1 -n2 /dev/md1 /dev/hda2 /dev/hdb2
(both hda2 & hdb2 have ext3 fs on them)
This works fine, eventually copies over everything and /proc/mdstat finds the array. This is going to be mounted as root.
Then, as I did for the /boot partition (which seems to mount fine), I did:
fsck -f /dev/md1
and it eventually finished. Then according to the HOWTO, I ran:
to get the superblock to say the right thing for raid. Everything fine so far, I can mount /dev/md1 no problem (running rescue).
I change lilo.conf to have:
and chroot/lilo, and it tells me it's updating two disks (correct).
But when I reboot, it fails the fsck, complaining about wrong superblock info! This didn't happen with /boot (/dev/md0). I can't figure it out.
Yes, both hda2 and hdb2 (and hda1/hdb1, the /boot partitions) are set to 'fd' in fdisk. They get detected at boot, but the fsck just fails.