RAID 3 (redundant array of independent disks 3)
What is RAID 3 (redundant array of independent disks 3)?
RAID 3 (redundant array of independent disks 3) is a RAID configuration that uses a parity disk to store the information generated by a RAID controller and uses striping for storing the data.
Because the parity information is on a separate disk, RAID 3 does not perform well when tasked with numerous small data requests. RAID 3 is a better choice for applications that have long sequential data transfers, such as streaming media, graphics and video editing.
RAID 3 uses block-level striping, which distributes block-level data across storage devices in the array, with parity handled by a separate drive. RAID 3 requires a minimum of three drives, one of which is the parity disk.
The rationale for RAID storage is to spread data across multiple devices, with the ability to rapidly retrieve the data when needed and also to increase data redundancy and survivability in the event of a disruptive event.
As RAID technology evolved, different versions emerged, some with high redundancy and recoverability, while other versions had different capabilities based on the design. RAID 3, for example, functions best with large data files, such as video files.
How is RAID 3 used?
Using a RAID controller as the administrative component, RAID 3, like other RAID iterations, stores large amounts of data. As noted, the design of RAID 3 makes it better suited for large files. The associated computer system's operating system (OS), such as Windows, views the RAID array as a single logical storage resource, despite the number of disk drives in the array.
When responding to an input/output (I/O) operation request from the OS, RAID 3 accesses all drives at the same time. This prevents overlapping I/O operations, ultimately making RAID 3 ideal for single-user systems with lengthy records used by the system's applications.
Other RAID versions provide greater redundancy and fault tolerance, faster data retrieval and the ability to accommodate smaller data requests, such as database inquiries. RAID 3 is often superseded or replaced by more powerful, redundant and fault-tolerant RAID arrays, such as RAID 0, 1, 5, 6, 10 and 50.
RAID 3 advantages and disadvantages
As noted, RAID 3 provides high throughput when storing and retrieving data, making it a good choice for large data files. RAID 3-based systems are generally fault-tolerant, and large data files can be retrieved quickly.
On the minus side, RAID 3 requires a separate parity disk, adding to the overhead and introducing a possible single point of failure. Despite the throughput, it is not well suited to rapid-fire data requests, such as database queries.
Alternatives to RAID 3
Other RAID levels can be selected in lieu of RAID 3, based on business requirements. Separate hard disk drives, solid-state drives and cloud storage services are among the most important alternatives.
Another possibility is network-attached storage (NAS), which supports a variety of requirements and can use a RAID level and replication to protect data. NAS is often used with a cloud storage service in a hybrid arrangement, using the NAS controller to store high-demand data within the local storage array and moving less critical data to the cloud service. The hybrid arrangement also provides another way to prevent data loss.
Strategies for selecting the proper RAID level
To select the most appropriate RAID level to implement, you must first determine the business requirements, followed by storage capacity requirements, system redundancy and fault tolerance. RAID levels address a broad spectrum of user storage and performance requirements, so a careful analysis of the requirements is important to avoid over- or underconfiguring a RAID array.
RAID 0 is considered the fastest implementation, RAID 1 has the best reliability and fault tolerance, and RAID 5 provides a good balance of the two. Disk failure and the associated data loss are always important considerations, especially in large arrays.
Aside from storage capacity, key considerations include data redundancy, disaster recovery, speed of data retrieval, data retention policies, data protection and privacy requirements, data security, physical space for disk arrays, and power and environmental requirements. Selection becomes a careful balance among business requirements, performance, reliability, data access and device survivability.
RAID protects data and improves storage performance and availability. Learn about the pros and cons of different RAID levels and where they work best.
