Am 21. Jul, 2007 schwätzte Randy Melder so:

> This is what gets me:
> "If up to N-1 disks are removed (or crashes), all data are still
> intact. If there are spare disks available, and if the system (eg.
> SCSI drivers or IDE chipset etc.) survived the crash, reconstruction
> of the mirror will immediately begin on one of the spare disks, after
> detection of the drive fault."
> ( http://www.tldp.org/HOWTO/Software-RAID-HOWTO-1.html#ss1.4 )
>
> Does that mean the spare disk is empty until a fault or removal of one
> of the drives?

The spare disk is unavailable for any use other than being a spare.

Essentially you have the space available of the smallest disk in the RAID
1 ( and spares ).

The spare theoretically isn't in use, but maybe some RAIDs do some sort of
spare updates during slow times. You shouldn't count on that.

> I guess I'm just looking for something that says "Spare disks in the
> array maintain the same mirror at all times even after failure of N-1
> disks."

Spare disks in the array should be counted as disks with no information on
them, but unavailable for other use.

Unless, of course, you're doing a security wipe after use, then you should
presume they have the most sensitive data that was in the RAID :).

ciao,

der.hans
-- 
#  https://www.LuftHans.com/        http://www.CiscoLearning.org/
#  "I interviewed Jonathan Winters once in the late '80s, it was as if he were
#  narrating a hallucination that I didn't see." -- Terry Gross, 03Aug2006