Re: Mirror of an existing filesystem with 'md'

[Chris Adams <cmadams hiwaay net>]

Once upon a time, Sean Millichamp <sean compu-aid net> said:
> It was my understanding that when a raid device was created a raid
> superblock was written to the very end of the block device.  If this is
> indeed the case, then you would be overwriting whatever it located at
> the end of your filesystem on that block device.

Yep.  I created a single-disk RAID 1 device today (with a "missing" disk
for some testing), and here is what /proc/partitions says the sizes are:

major minor  #blocks  name
   8     8     811251 sda8
   9     0     811136 md0

MD took 115 blocks (1kB sized blocks).  I don't know what determines how
much is used, but some amount is definately used.

If you can figure out how much, you can resize your ext2/3 filesystem
down a little to allow for the MD superblock (or resize the FS down a
good bit, make the RAID, then resize the FS back up to the new available
size).

-- 
Chris Adams <cmadams hiwaay net>
Systems and Network Administrator - HiWAAY Internet Services
I don't speak for anybody but myself - that's enough trouble.