| MODE |
RAID 0 / Striping |
RAID 1 / Mirroring |
RAID 5 / Striping with Parity |
RAID 10 /Striping & Mirroring |
| Short Description |
Performance |
Data Security |
Performance and Data Security |
Performance and Data Security |
| No. of needed HDD |
2 or more, same size |
2 - same size |
3 or more |
4 - same size |
| Formula |
Volume capacity = (# of HDDs) * (capacity of smallest HDD) |
Volume capacity = capacity of smallest HDD |
Volume capacity = ((# of HDDs) – 1) * (capacity of smallest HDD) |
Volume capacity = 2 * (capacity of smallest HDD) |
| Example |
4 x 500GB = 2.000GB gross storage capacity |
2 x 500GB = 500GB gross storage capacity |
4 x 500GB = (4-1)*500GB = 1.500GB gross storage capacity |
4 x 500GB = 2.000GB / 2 = 1.000GB gross storage capacity |
| Description |
RAID 0: This is called "striping". This is a performance-oriented data mapping technique since
there is a near-linear performance scaling with the addition of connected hard drive(s). Data
written to the array is divided into stripes and written across the disks in the array. This procedure
enables high I/O performance at a low cost. |
RAID 1: This is called "mirroring". It is a data redundancy technique. All data in the system is simultaneously written to two hard disks instead of one. The redundancy provides full protection
against data loss in the event of the failure of either of the disks. When a hard drive failure
occurs, all data is immediately available on the second hard drive. |
RAID 5: RAID 5 stripes both data and parity information across three or more drives. RAID 5
has a distributed parity algorithm, where data and parity blocks are written across all drives in the
array. Fault tolerance is maintained by ensuring that the parity information for any given block is
placed on a drive separate from those used to store the data itself. Read operations outperform
write operations.
RAID 5 requires a minimum of three hard drives. |
RAID 10: This is also known as RAID 1+0 (striped mirrors) and is similar to RAID 0+1
(mirrored stripes). It uses a combination of RAID 0 and RAID 1. This combines the best features
of striping and mirroring to yield large arrays with high performance and superior fault tolerance. |
| Advantages |
The fact that individual disks can be accessed in parallel improves performance.
In a two disk RAID 0 volume, the performance is roughly twice that of each component disk.
Three and four disk RAID 0 volumes also have near linear scaling in performance. |
Complete redundancy of data
The performance is higher for read-intensive applications because only one read request needs to
be issued, but the data can be read in parallel from either disk. |
Fault tolerant, data security. One harddrive (any one in the array) can break down and all data is still available. More HDD boost the performance |
RAID 10 provides excellent overall performance by combining the speed of RAID 0 with the
redundancy of RAID 1. |
| Disadvantages |
RAID 0 offers
no data redundancy. If we have a four disk RAID 0 volume, the chance of a hardware disk drive
breakdown is four times greater than that of a single disk.
Separate Backup Solution is recommended to ensure data redundancy. |
The performance is lower for write-intensive applications since two write requests need to be
issued, one to each disk. |
Reduced write performance. Read requests can be read from any of the disks, but in the case of write
operations, the data must be written, parity must be calculated, and then parity must be written. |
Four drives are required so it is more expensive in terms of redundancy
information. |
