I don't know the internals of RAID but I have just done a little
experiment - it probably won't help you sleep at night but I thought you
would like to know and it may help you.
1. I made a RAID-0 array (raidtab & mkraid /dev/md0) under Mandrake
7.0-2 using a fresh 2.2.14 kernel with IDE & RAID patches.
Copied data onto the RAID then shutdown.
2. Moved the drives so they were in different positions and booted.
3. During the startup it came up as an error loading RAID and
dumped me into a shell.
4. From the shell I edited raidtab and did a mkraid -f /dev/md0
which it then replied with the fact that superblocks were different or
something (sorry didn't think to copy it down) I then did another
mkraid -f /dev/md0 and the superblocks were rewritten and the event
counter set back to 0.
5. Without doing a mkfs or even a fsck I mounted /dev/md0 and was
able to access all the data again.
>From this it seems that mkraid just modifies the RAID superblock (if
there is such a beast) and not the superblock of the filesystem. This
means to me at least that the problem is not with the mkraid -f /dev/md0
but what else RedHat decided to do.
I realise that this does not immediately solve your problem but I hope
it points you in the right direction.
If you have any questions feel free to e-mail me directly.
Adrian Head
> -----Original Message-----
> From: Volker Wysk [SMTP:[EMAIL PROTECTED]]
> Sent: Tuesday, May 16, 2000 9:09 PM
> To: [EMAIL PROTECTED]
> Subject: Still trying to save my data
>
> Hello.
>
> I've tried Gelnn's tip (thanks!), but still all superblocks seem to be
> corrupted. This seems quite strange to me, since the volume has not
> been formatted.
>
> Is there anyone familiar with the internals of the RAID system, who
> could
> tell me what actually happens when mkraid is called? *Should* "e2fsck
> -n
> -b ... /dev/md0" work?
>
> bye.
