Subject: Re: RAID disk failure
To: Jukka Marin <jmarin@pyy.jmp.fi>
From: Robert Elz <kre@munnari.OZ.AU>
List: netbsd-users
Date: 12/26/2001 22:20:12
    Date:        Wed, 26 Dec 2001 14:11:09 +0200
    From:        Jukka Marin <jmarin@pyy.jmp.fi>
    Message-ID:  <20011226141109.A6955@pyy.jmp.fi>

  | What is the easiest way of moving the RAID1 system to two bigger disks?

Just a possibility - remove the old raid drives, plug in two new ones
(assuming that's what you mean), build the (empty) raid filesystems,
get parity rebuilt, ...   Then shutdown, remove one of the new drives,
replace it with the half of the old raid that still works - now you have
two degraded raid filesystems, both usable.   Copy data.  Remove old
raid drive, replace with the new one temporarily removed, and allow
parity to rebuild.

  | I guess it would be easiest just to drop a 60 GB disk
  | in the machine and 'raidctl -R' it .. but then I would lose the 15 gigs
  | and still be limited to 45 GB of storage (which doesn't seem to be too
  | much these days).

That would be easier, but no need to lose the 15GB (though you will have
no way to provide redundancy for it, so use it for /usr/obj or something).

Stick in the new drive, label it - one partition 45G to match the half
drive that is left (say wd1e) - set as type RAID, and the other 15G as one,
or two (or whatever you like) partitions for normal filesystems (ffs, lfs,
whatever takes your fancy).   Add wd1e to the raid, and allow the parity
to rebuild, and do what you like with the filesystem(s) that are left over.

kre