Appendix; Introduction To Raid; Raid 0 - VIA Technologies P4PB ULTRA-URL User Manual

Via mainboard

Table of Contents

Appendix

The basic idea of RAID (Redundant Array of Independent Disks, formally

known as Redundant Array of Inexpensive Disks) is to store the same data in

different places on multiple hard disks. It is primarily a performance option,

but it too can be used for data protection against the loss of data due to human

carelessness, mechanical failures or natural disasters. It has parity check func-

tion such that when one of the hard disk in the array fails, after replacing the

disk with a new one, all data is kept intact as nothing has happened.

RAID 0 is also know as "striping". The minimum of disks require to imple-

ment RAID 0 is two. All the disks connected to this array are seen as a single

large disk. For example, if you have two identical hard disks which all come

with 20GB of capacity, the operating system will see these two hard disks as

one 40GB hard disk.

Data blocks are split equally among the disks. For example, if you have 8

blocks of data to be stored in two hard disks, blocks 1, 3, 5 and 7 will be stored

in disk 1 while blocks 2, 4, 6 and 8 will be stored in disk 2. Performance is

increased as read and write can be carried out in parallel, but with no data

redundancy. The disadvantage of RAID 0 array is that when one of the disk

fails, the whole arrays fails since the data is spread across the array. RAID 0 is

best used where cost/performance is critical, but data integrity is not.

RAID 1

RAID 1 is also known as "mirroring". Two identical hard disks are seen as

only one hard disk, but with only with capacity of just one. For example, if

you have two 20GB hard disks connected to this array, instead of showing

40GB, the operating system will only see a single 20GB hard disks.

Data is mirrored between the disks. This means that you have two disks both

containing exactly the same data. The advantage of RAID 1 is when one disk

fails, the array will keep working as nothing has happened. Read perfor-

mance is improved since the disks can be read just once, while write perfor-

mance is more or less identical as writing to a single disk.

A-

Table of Contents