[KLUG Members] RAID problems.

[Adam Williams]

>>4ft for 9 drives including an external cabinet?  This is ***VERY*** hard
>>to believe.
>I don't see what's so hard to believe. Is it too long, or too short?

You have the cable between the host and the external cabinet.  You have
the cable inside the cabinet.  You have the cable inside the host.

For the four drives in my server I have two feet of cable.  My 8-bay
cabinet has 2.5 feet inside.  My 4 bay cabinet has 2 feet of cable inside.
The cable in between each cabinet and the host is four feet.  No matter
what I end up with => 6ft.  And some cabinets have ALOT more cable inside
that these.  But I doubt it is a length thing anyway.

>Anyway, I'm finding out it's probably a parity error due to excessive heat,

I can believe this,  but I'd still double and triple check your
termination.

>and that a verify of the drive should fix it, although it won't save the
>data.  I was instructed by adaptec to reformat the drives and make a new
>RAID.  I just might, but losing 30+ GB of data (although it's already
>"lost") is pretty hard.  Hurts, too.

Is it too far gone to try and regenerate the array when the drives are
cold?  That both the parity and data are equally corrupt seems not so
likely,  although altogether possible.  I know some very smart people who
say that RAID-5 is an entirely false sense of security.

-- 
-----------------------------------------------------------
Ximian GNOME, Evolution, LTSP, and RedHat Linux + LVM & XFS
-----------------------------------------------------------