atacontrol Raid, cannot re-add member to array

[Paul Mather]

On Sun, 2004-07-04 at 18:10, Simon L. Nielsen wrote:
> On 2004.07.04 23:33:28 +0200, Harald Schmalzbauer wrote:
> 
> > I've never tried ataraid with "non-raid" controllers but I doubt that 
> > detach/attach would work.
> 
> It does work, you "just" have to hotswap the disk [1].  I tried it
> some time ago, and I successfully did a rebuild, though I did kill one
> of the disks shortly after since I tried to plug in the power cable
> the wrong way during another test (yes that's stupid, I know :-) ).

So does this mean ATA RAID doesn't work on "non-raid" controllers that
have non-hot swappable drives attached?  (E.g., a drive get hard errors,
is marked as failed and the RAID as DEGRADED, and you have to shut down
the machine to remove and replace it---but ATA RAID won't recognise
it/rebuild onto it when you reboot with the replacement drive.)

If that is the case, the man page really should note that serious
limitation.

I have tried the detach/attach on a non-raid controller to simulate
failure and have *never* managed to get rebuild to work.  I've also
tried the shutdown/remove/replace/reboot method but, again, *never*
managed to get rebuild to work on a "non-raid" controller. :-(

I've never had any hotswap-capable drives to test the hot-swap
replacement/rebuild method. :-(

> Well, I'm using it just fine on my main mailserver (running 5.2.1) but
> that's a RAID1 with standard a ATA controller.

I wish it would work for me. :-)  I have a non-raid controller 5.2.1
system that I tried it on (with non-hot swap drives) that I eventually
had to bail and use Vinum on.  Alas, Vinum seems to be rotting at a fast
pace in CURRENT, so I'm worried about the prospect of being able to
upgrade that system when 5.3 rolls around... :-(

Cheers,

Paul.
-- 
e-mail: paul at gromit.dlib.vt.edu

"Without music to decorate it, time is just a bunch of boring production
 deadlines or dates by which bills must be paid."
        --- Frank Vincent Zappa