IBM ServeRAID-4 Ultra160 Series, User Reference

Page: 40 / 208

If a physical drive fails in the array,
the data from the failed drive is
compressed into the distributed spare
drive. The logical drive remains RAID
level-5E.
When you replace the failed drive, the
data for the logical drive
decompresses and returns to the
original striping scheme.
9
3
5
7
1
*
*
*
*
2
6
8
4
*
If you use a RAID level-5E logical drive in a failover or cluster configuration, the
RAID level-5E logical drive will not failover while undergoing compression or
decompression.
Note: The ServeRAID Manager program Express configuration does not default to
RAID level-5E. If you have four physical drives, Express configuration
defaults to RAID level-5 with a hot-spare drive.
Understanding RAID level-x0
RAID level-x0 refers to RAID levels-00, 10, 1E0, and 50. RAID level-x0 includes
more physical drives in an array. The benefits of doing so are larger logical drives,
increased performance, and increased reliability. RAID levels-0, 1, 1E, 5, and 5E
cannot use more than 16 physical drives in an array. However, RAID levels-00, 10,
1E0, and 50 include more physical drives by managing an array of arrays, or a
spanned array. The operating system uses the spanned array logical drive the
same as a regular array logical drive.
Note: RAID level-x0 is only available on the IBM ServeRAID-4 Ultra160 SCSI
controller.
RAID level-x0 requires a minimum of two drives and supports a maximum of 60
drives.
The following illustration is an example of a RAID level-10 logical drive.
Start with six physical drives.
Create three arrays (labeled
A, B, and C), each array
using two physical drives.
A A B C A
22
IBM
®
User’s Reference: ServeRAID
™
-4 Ultra160 SCSI Controller