1. What is RAID?

The term RAID stands for Redundant Array of Inexpensive Disks (or Redundant Arrays of Independent Disks), which first appeared in the late 1980s. RAID is a virtualization technique that allows multiple physical hard drives to be grouped into one. Logical drive system aims to increase data access speed or reduce the risk of data loss or corruption caused by hardware disk failure, or sometimes a combination of both.

Watching: What is a Hard Drive Raid

The way RAID works is to copy data onto two or more physical hard drives linked together by a RAID Controller. RAID Controllers can be based on either hardware or software.

Most of the different RAID types use an error-limiting technique called “parity” data that allows for fault tolerance when data is duplicated This can reduce the impact of data loss. data in the event of a hardware failure.

RAID can be used for SATA, SAS and SSD drives.

There are different ways to install RAID. Each type of RAID serves a different goal based on specific needs to address certain requirements such as:

Data Reliability/Data Reliability – ensures data is error free.

– Data Availability – ensures data is available even in the event of hardware failure.

Data Performance/Data Performance – ensures fast data access for both read and write operations.

– Data capacity / Data Capacity – ensures the ability to store large amounts of data.

2. Common RAID Types

RAID is developed according to many different standards, each with differences in performance, storage capacity, and reliability. RAID can be grouped into two directions as follows:

– RAID according to common standards

– RAID improved and developed according to its own standards

RAID Standards

RAID 0:

Raid 0 requires at least two hard disks, RAID 0 allows computers to write data in a special method known as Striping. For example, you have 8 data segments numbered from 1 to 8, the odd numbered segments (1,3,5,7) will be written to the hard disk first and the even numbered segments (2,4,6). ,8) will be burned to the second disc, which cuts the theoretical working time in half

READ MORE  Peer Review In Vietnamese

In general, with n hard disks, each disk only needs to read/write 1/n of the requested data. Theoretically, the speed will increase n times..

However, RAID 0 has a high risk of data loss, the main reason lies in the way the information is written separately because so the data is not completely located on a single hard disk and whenever it is necessary to access information (for example, that), the computer will have to synthesize from the hard disk.

As long as a hard disk fails, that information (file) is considered unreadable and completely damaged.

The final capacity of a RAID 0 system is equal to the sum of the capacities of the single drives (two 80GB drives running RAID 0 will see only one 160GB RAID drive).

It can be seen that RAID 0 is really suitable for users who need to quickly access large volumes of data, for example gamers or those who specialize in digital graphics and video.

RAID 1

This is the most basic form of RAID capable of ensuring data security.

Just like RAID 0, RAID 1 requires at least two hard disks to work. Data is written to 2 identical drives (Mirroring). In the event that one drive fails, the other will continue to function normally. You can replace damaged drives without worrying about damaged information.

See also: What is Email Marketing – A New Perspective In Marketing 4

As for RAID 1, performance isn’t a top factor, so it’s not surprising that it’s not the number one choice for speed enthusiasts. However, for network administrators or those who have to manage a lot of important information, the RAID 1 system is an indispensable choice.

The final capacity of a RAID 1 system is equal to the capacity of a single drive (two 80GB drives running RAID 1 will show the system only one 80GB RAID drive).

RAID 2

The error checking mechanism of this type of raid at the Bit level uses the Hamming code mechanism, but almost by default all current hard drives have integrated Bit level checking with Hamming code, so this type of Raid is no longer available. use anymore.

READ MORE  What is the age of 96 years old?

RAID 3

This is the first raid to use data redundancy with Parity.

Parity is the result produced by calculating the XOR between the bits in the stored data block. For example, a data file is divided into 1 block stored in 6 parts, the storage part from Block A1 – A3 is stored on 3 hard drives, then Raid 3 will calculate the XOR of the bits in this Block 1 to 3 word into a block. parity stored on the 4th hard drive.

When a failure occurs, the system will recalculate the data from the operation:

A-Parity(1-3) = Block A1 Block A2 Block A3

However, in fact, this type of Raid is no longer used because the performance is too low, especially with the Database because the bit-by-bit processing is very time consuming when storing and retrieving.

RAID 4

RAID 4 is similar to RAID 3 but at a larger level of data blocks rather than individual bytes. They also require a minimum of 3 hard disks (at least two disks for data storage and at least one disk for parity data storage)

However, this type of raid is not used much anymore.

RAID 5

This is probably the most powerful and popular form of RAID. Data and backups are shared across all hard drives. This rule is quite complicated

We go back to the example of 8 data segments (1-8) and now 3 hard drives. Fragments 1 and 2 will be written to drives 1 and 2 separately, their backup segments are written to hard drive 3. Segments 3 and 4 are written to drives 1 and 3 with backup segments respectively writes to drive 2. Fragments 5, 6 are written to drives 2 and 3, and backups are written to drive 1 and then this sequence repeats, segments 7,8 are written to drives 1, 2 and the backup is written to drive 3 like the original. Thus, RAID 5 both ensures improved speed and high safety. The final hard disk space is equal to the total disk space used minus one drive. That is, if you use 3 80GB drives, the final capacity will be 160GB.

RAID 6

RAID 6 is somewhat like RAID 5 but is used more repeatedly to separate data to write to different hard disks. For example, in RAID 5, each data is separated into two storage locations on two different hard disks, but in RAID 6, each data is stored in at least three (or more) locations. This helps to increase the security of data compared to RAID 5.

READ MORE  Sketch Drawing Practice At Home For Beginners

In RAID 6, we see that the tolerance to hard failure is greatly increased. If with 4 hard drives, they allow simultaneous failure of 2 hard drives and the system still works normally, this creates a very large safety probability. That is why RAID 6 is usually only used in servers containing extremely important data

3. Non-standard RAIDs

RAID 10 & RAID 01

This is a type of Raid that combines Raid 1 and Raid 0 but are two completely different mechanisms.

Raid 10: Split data into two parts by Raid 0 mechanism, each part is mirrored by Raid 1 . mechanism

Raid 01: The data is mirrored the same on the first write under raid 1 but split into two parts on the second write under raid 0

In terms of performance, it is clear that Raid 10 has a clear advantage in both read and write access speed compared to Raid 01. For systems that require high performance while ensuring data security, Raid 10 is the choice. choose optimal.

Raid 50

As a perfect combination of Raid 5 and Raid 0, the data is first written according to Raid 0 mechanism but divided according to Raid 5 mechanism on the second write.

As a result, this type of RAID still ensures very good query speed similar to RAID 10 but makes better use of hard drive capacity than Raid 10.

See also: What is a Young Master – Young Master or Young Master

Hot spare disks

This is a hard drive whose primary function is for backup use only. When a hard drive in the Raid group fails, the system will automatically take the HotSpare hard drive to replace the damaged hard drive to ensure the best data protection. Hot spare disk can be used for 1 raid cluster or can be shared for many different raid clusters.