RAID Description

RAID stands for Redundant Arrays of Independent Disks. The basic idea behind RAID is to combine multiple physical disks into one virtual disk unit with a greater capacity. This provides higher storage performance, higher I/O performance, and higher reliability than a single disk, providing an effective solution for data security and storage performance.

The main purpose of RAID technology is to improve the performance and reliability of storage systems through data distribution and redundant storage. It combines multiple independent hard disks into a single logical unit. In RAID, data is divided into multiple blocks, and these blocks are stored on different physical disks. In this way, data can be read and written in parallel to improve the throughput and response speed of the storage system.
In addition to improved performance, RAID also provides high data reliability through redundant storage. In some RAID levels, data is copied to multiple disks so that even if one disk fails, the system can still work and the lost data can be reconstructed with redundant data.
There are multiple RAID levels, such as RAID 0, RAID 1, and RAID 5, and each level has different characteristics and application scenarios. The performance, reliability, and cost vary with RAID levels. You can select an appropriate RAID level and configure the storage system based on your business needs.

RAID Features and Concept

Features/Concept	Explanation
Disk Pack and Virtual Disk	A disk pack combines multiple physical disks into a logical unit and a virtual disk is a continuous data storage unit created from a disk pack. A disk pack provides scalability and redundancy, while a virtual disk can be a complete disk pack, a combination of multiple disk packs, or a part of a disk pack.
Fault Tolerance	Fault tolerance refers to the ability to maintain data integrity and system availability when a disk error occurs or a disk fails in RAID. Fault tolerance is available in RAID levels such as RAID 1, RAID 5, RAID 6, and RAID 10 by using redundant storage and data verification algorithms to protect data against the effect of the failure of a single disk.
Consistency Check	A consistency check is to check data consistency for RAID levels with redundancy, such as RAID 1, RAID 5, RAID 6, and RAID 10. The RAID controller card verifies and calculates data in a disk pack and compares it with redundant data to ensure data integrity. If data inconsistency is detected, the system attempts to automatically fix the error and record the error information.
Disk Striping	Disk striping is a technology that divides data into multiple blocks and stores them on different physical disks. In this way, data can be read and written in parallel to improve the throughput and response speed of the storage system. With striping, multiple processes can access different parts of the data at a time, avoiding disk conflicts while maximizing the I/O parallel processing capability.
Disk Image	Disk image applies to RAID 1 and RAID 10 and achieves 100% data redundancy by writing the same data into two disks at a time. When one disk fails, the data on the other disk is still available and the system can still work normally. Disk image also enables fast data recovery. After a faulty disk is replaced, the system automatically copies data to the new disk to restore the redundancy state.
Just a Bunch Of Disks (JBOD)	Just a Bunch Of Disks (JBOD) is a mode of data transmission that allows RAID controllers to transmit commands transparently to hard disks without the need for processing in the transmission device. In this way, upper-layer service software or management software can directly access and control hard disks. For example, the hard disk mounted to the RAID controller card can be used as the installation drive when you install the server OS, without limitation by the virtual disk.

Commonly Used RAID Levels

RAID stands for Redundant Arrays Of Independent Disks. Commonly used RAID levels include RAID 0, RAID 1, RAID 5, RAID 6, RAID 10, RAID 50, and RAID 60. Each level has different characteristics and application scenarios.

RAID 0

RAID 0 distributes data across multiple hard disks in striping mode to improve storage performance. Data parallel processing can make full use of bus bandwidth, but this level does not have data redundancy and fault tolerance features, and is suitable for scenarios that require high I/O but low data security.

RAID 1

RAID 1 uses disk mirroring. Each working disk has a mirror disk for full data redundancy. Data is also written to the mirror disk, and can be read from the working disk or mirroring disk. RAID 1 provides high data reliability but reduces the effective capacity by half and is suitable for scenarios that require high fault tolerance, such as financing and banking.

RAID 5

RAID 5 achieves fault tolerance and high performance with cyclic redundancy check (CRC) and decentralized data storage. Data and parity data are distributed on each member disk in RAID. If a disk fails, the data on the failed disk can be reconstructed using the data on another disk. RAID 5 is suitable for massive data operations and transaction processing, providing fast, large capacity, and reasonable fault-tolerant disk arrays.

RAID 6

RAID 6 adds a second independent parity check information block on the basis of RAID 5 to provide higher data reliability. Data is still available even when two disks fail. However, compared with RAID 5, RAID 6 requires more disk space for parity check and has poor write performance.

RAID 10

RAID 10 is a RAID configuration that combines RAID 0 and RAID 1. It divides multiple hard disks into multiple groups and mirrors them, and then the mirroring groups are striped. RAID 10 provides the same data security as RAID 1 and the basically same storage performance as RAID 0. RAID 10 is suitable for scenarios that balance performance and security.

RAID 50

RAID 50 is a RAID configuration that combines RAID 5 and RAID 0. Data is partitioned into stripes, the parity bits are used to ensure data security, and stripes with parity are evenly distributed on disks. RAID 50 combines the features of RAID 5 and RAID 0 and is suitable for scenarios that require high storage performance and fault tolerance.

RAID 60

RAID 60 is a RAID configuration that combines RAID 6 and RAID 0. Data is partitioned into stripes and two independent parity check information blocks are used to ensure data security. RAID 60 provides high data reliability and fault tolerance, and is suitable for scenarios that require high data security.

Performance Comparison Among RAID Levels

RAID Level	Number of Disks Required	Fault Tolerance	I/O Performance	Storage Capacity	Storage Capacity Utilization
RAID 0	1 or more	None	High	Minimum capacity of member disk × number of member disks	100%
RAID 1	2	100%	Low	Minimum capacity of member disk	50%
RAID 5	3 or more	One disk fails	High	Minimum capacity of member disk × (number of member disks - 1)	(N-1)/N
RAID 6	3 or more	Two disks fail	Medium	Minimum capacity of member disk × (number of member disks - 2)	(N-2)/N
RAID 10	4 or more	One disk fails	High	Minimum capacity of member disk × number of member disks/2	50%
RAID 50	6 or more	One disk fails	High	Capacity of subgroup × number of subgroups	(N-M)/N
RAID 60	6 or more	Two disks fail	Medium	Capacity of subgroup × number of subgroups	(N-M*2)/N

Note: In the preceding table, N is the total number of disks and M is the number of member disks in each RAID subgroup.

As can be seen from the preceding table, fault tolerance, I/O performance, storage capacity, and storage capacity utilization differ depending on RAID levels:

RAID 0: Provides no fault tolerance and features high I/O performance and storage capacity, but low data security.
RAID 1: Provides 100% data redundancy, but reduces the effective storage capacity by half. It is suitable for scenarios that require high fault tolerance.
RAID 5: Provides fault tolerance for one disk, and features high I/O performance and storage capacity utilization.
RAID 6: Provides fault tolerance for two disks and is suitable for scenarios that require high reliability and I/O performance.
RAID 10: Provides fault tolerance for one disk, combines the advantages of RAID 0 and RAID 1, and applies to scenarios that require high performance and fault tolerance.
RAID 50: Provides fault tolerance for one disk and is suitable for scenarios that require high performance and fault tolerance.
RAID 60: Provides fault tolerance for two disks and is suitable for scenarios that require high data security and performance.

You can select a RAID level based on the performance, fault tolerance, and storage capacity requirements of your application.

Creating a RAID

You can create a RAID using EVS. For more information, see Create a RAID Using EVS.

Physical Server

Configuring RAID