Chapter 3: Creating Disk Arrays

Initialization

RAID 5/50 arrays must have consistent parity before they

 

can be used to protect data. Initialization writes a known

 

pattern to all drives in the array. If you choose not to

 

initialize an array, this is known as a “trusted array” and

 

any drive failure will result in data corruption. It is possible

 

to later perform a parity rewrite, which recalculates the

 

parity based on the current data, thus ensuring the data

 

and parity are consistent.

 

 

Reserved Capacity

In order to allow drives from a different family or

 

manufacturer to be used as a replacement for a drive in

 

an array, we recommend that a small percentage of the

 

drive’s capacity be reserved when creating the array. This

 

is user selectable, from 0 to 10 percent.

 

 

RAID Level 0

RAID 0 is defined as disk striping where data is striped or

 

spread across one or more drives in parallel. RAID 0 is

 

ideal for environments in which performance (read and

 

write) is more important than fault tolerance or you need

 

the maximum amount of available drive capacity in one

 

volume. Drive parallelism increases throughput because

 

all drives in the stripe set work together on every I/O

 

operation. For greatest efficiency, all drives in the stripe set

 

must be the same capacity. Because all drives are used

 

in every operation, RAID 0 allows for single-threaded I/O

 

only (i.e., one I/O operation at a time). Environments with

 

many small simultaneous transactions (e.g., order entry

 

systems) will not get the best possible throughput.

 

 

RAID Level 1

RAID 1 is defined as disk mirroring where one drive is an

 

exact copy of the other. RAID 1 is useful for building a

 

fault-tolerant system or data volume, providing excellent

 

availability without sacrificing performance.

 

 

RAID Level 5

RAID 5 is defined as disk striping with parity where the

 

parity data is distributed across with parity all drives in the

 

volume. Normal data and parity data are written to drives

 

in the stripe set in a round-robin algorithm. RAID 5 is multi

 

threaded for both reads and writes because both normal

 

data and parity data are distributed round-robin. This is

 

one reason why RAID 5 offers better overall performance

 

in server applications. Random I/O benefits more from

 

RAID 5 than does sequential I/O, and writes take a

 

performance hit because of the parity calculations. RAID

 

5 is ideal for database applications.

 

 

20

www.gateway.com