On Wed, 25 Jul 2001, Nadav Har'El wrote:
> Am I missing something, or isn't RAID (*Redundant* Array of Inexpensive Disks)
> supposed to fix exactly that problem, so that when only one disk dies, it can
> be taken out (and later replaced) without a single byte lost?
the folks that chose the acronym RAID have missed something. RAID levels 1
(mirroring) 4 and 5 (extra parity data) support such redundancy. RAID 0
(striping) or linear mode (as is used on linux.org.il) don't support any
redundancy. when one disk goes - they all go.
> Or was the raid set up with zero redundancy? If so, what was its purpose
> exactly? To allow for larger disk throughput? To allow for a giant virtual
> partition?
the last one - a large (certainly not giant, in today's terms) virtual
partition, to avoid having to jaggle data between partitions, and allow
maximal usage of disk space. the space we have no isn't sufficient as it
is, so we'll not go with raid 5 for now.
--
guy
"For world domination - press 1,
or dial 0, and please hold, for the creator." -- nob o. dy
----------------------------------------------------------------------------
To unsubscribe, send a message to [EMAIL PROTECTED]
Archives available at http://www.mail-archive.com/[email protected]/
