It sounds like you havn't get raid-0 compiled into the kernel or the raid
modules setup.
There are two ways of getting raid into the kernel:
1. use a initrd image that the kernel loads. This image has the raid
modules in it.
2. just do a make menuconfig and select RAID - 0 as '*' not 'M' so that
the raid will be part of the kernel. You will need to rebuild the kernel
image after this.
Have a look at your messages before the kernel mounts the raid partition. you
should see some raid error messages.
Since raid will always be used there is no real overhead in just compiling it
into the kernel.
Julienna Chu wrote:
> Guys,
>
> For my etc/raidtab I have
>
> raiddev /dev/md0
> raid-level 0
> nr-raid-disks 2
> persistent-superblock 1
> chunk-size 4
> device /dev/hdc1
> raid-disk 0
> device /dev/hdd1
> raid-disk 1
>
> /etc/fstab has
>
> /mnt/md0 /backup ext2 defaults 1 1
>
> I know it worked before it rebooted. I was ablr to mount it and write
> files to it just fine.
> When the system rebooted I started to get that messgae about reading short
>
> fsck.ext2:Attempt to read block from filesystem resulted in short read
> while trying to open /dev/md0
> Could this be a zero-length partition? (linux actually asked me this)
>
> I even went into cfdisk to create a primary partition with partition id
> of fd
> for both /dev/hdc1 and /dev/hdd1 before I rebooted the second time and I
> still
> got these fsck.ext2:Attempt to read block from filesystem resulted in
> short read while trying to open /dev/md0
> mesgs upon boot up.
>
> I then go to maintenannce mdoe and I do a raidstart -a, mount -a, and
> my raid is back and alive.
>
> When I reboot, I then get these messages about
> fsck.ext2:Attempt to read block from filesystem resulted in short read
> while trying to open /dev/md0
>
> What gives?
