Redundant Array of Independent Disks (RAID)

1. Overview

RAID(Redundant Array of Inexpensive Disks, or Drives, or Redundant Array of Independent Disks) is a data storage virtualization technology that combines multiple physical disk drive components into one or more logical units for the purposes of data redundancy, performance improvement, or both. This was in contrast to the previous concept of highly reliable mainframe disk drives referred to as Single Large Expensive Disk(SLED). 

Data is distributed across the drives in one of several ways, referred to as RAID levels, depending on the required level of redundancy and performance. The different schemes, or data distribution layouts, are named by the word "RAID" followed by a number, for example, RAID 0 or RAID 1. Each scheme, or RAID level, provides a different balance among the key goals: reliability, availability, performance, and capacity. RAID levels greater than RAID 0 provide protection against unrecoverable sector read errors, as well as against failures of whole physical drives.

2. RAID 0

2.1 Feature

• Split data across any number of disk allowing higher data throughput

2.2 Advantages

• No overhead caused by parity controls
• All storage capacity is used
• Easy to implement

2.3 Disadvantages

• No parity
• No mirroring
• No redundancy
• No-fault-tolerant

2.4 Usages

• Non-critical storage of data with high-speed read/write
• image-retouching or video editing station

2.5 Concatenation

• Each disk stores a whole contiguous section of data
• Easy to grow by simply adding another disk
• Offers some performance benefits with random read and writes
• Failures tend to be localized

 2.6 Striping

• Process of dividing data into small blocks and spreading these data blocks across several physical disks
• Each neighboring sequential block residing n another disk
• A single disk failure would result in most of the data is useless because they would have holes in them
• Cannot be grown easily since to add additional disks all the data on the existing disks would have to relocate.

3. RAID 1

3.1 Features

• Minimum number of disk is 2

3.2 Mirroring

• Writing all data to two different drives at the same time
• To get data redundancy
• Prevent disk failure from being a single point of failure
• Fault-tolerant

3.3 Advantages

• Offer excellent read speed and a write-speed that is comparable to that of a single drive
• If a drive fails, they just have to be copied to the replacement drive
• Very simple technology

3.4 Disadvantages

• Effective storage capacity is only half of the total drive capacity
• Do not always allow a hot swap of a failed drive
  • Need hardware controllers to support hot-swapping

3.5 Usages

• Mission-critical storage such as accounting systems
• Suitable for small servers only two data drives will be used

3.6 RAID 1E

• Use the striping technique increase in reading speed even for degraded configurations
• Allows more than two disks in the set

4. RAID 2

 

• Consists of bit-level striping with dedicated Hamming-code parity
• Data is striped such that sequential bit is on a different drive
• Hamming-code parity is calculated across corresponding bits and stored on at least one parity drive
• Not used by any commercially available system
• Parity
  • A technique of checking if data has been lost or written over when it's  moved from one place in storage to another
  • Stored separately against extra data
  • Recreating missing block from the remaining blocks and the parity block

 

5. RAID 3/4

5.1 RAID 3

• Consist of byte-level striping with dedicated parity
• All disk spindle rotation is synchronized
• Each sequential byte is on a different drive
• Parity is calculated across corresponding bytes and stored on a dedicated parity drive
• Not commonly used in practice

 

5.2 RAID 4

• Consist of Block-level striping
• RAID-DP, two parity disks, used
• I/O parallelism
  • Not require reading across all data drives
  • Improving the performance of small transfers

6. RAID 5

6.1 Features

• Minimum number of disks 3
• Stripe data with parity to recover the data in case of disk failure
  • Offer both speed and redundancy as parity data is stored across all of the disks

6.2 Advantages

• Very Fast Read data transactions 
• Somewhat slower Write data transactions
• If a drive fails, Still have access to all data, even while the failed drive is being replaced and the storage controller rebuilds the data on the new data

6.3 Disadvantages

• Drive failures have an effect on throughput(degraded mode), although this is still acceptable
• Complex technology
  • Consume quite Rebuild time 
  • If another failure during rebuilding, losing data

6.4 Usages

• Good all-round system combines efficient storage with excellent security and decent performance
• File and application servers have a limited number of data drives

7. RAID 6

7.1 Features

• Parity data are written to two drives
• Requires at least 4 drives 
• Withstand 2 drives dying simultaneously

7.2 Advantages

• Read data transaction very fast
• More secure than RAID 5
• Tolerant to two drive fail even while the failed drives are being replaced

7.3 Disadvantages

• Write data transactions slower than RAID 5 cause of additional parity data
  • About 20% slower
• Drive failures have an effect on throughput
• Complex technology which rebuilds drives takes a long time
  • If another failure during rebuilding, losing data

8. RAID 10

8.1 Features

• Minimum number of disk is 4
• Combines striping of RAID 0 and mirroring of RAID 1
• Redundancy of RAID 1 with 

8.2 Advantages

• Rebuild time is very fast  because all needed is copying all the data from the surviving mirror to a new drive
  • 30 minutes for drives of 1 TB

8.3 Disadvantages

• Expensive redundancy
  • Half of the storage capacity goes to mirroring

8.4 Usages

• Highly utilized database servers
• servers performing a lot of writing operations

9. Comparison of RAID levels

Features	RAID 0	RAID 1	RAID 1E	RAID 5	RAID 5EE	RAID 6	RAID 10
Minimum # Drives	2	2	3	3	4	4	4
Data Protection	No Protection	Single-drive failure	Single-drive failure	Single-drive failure	Single-drive failure	Two-drive failure	Up to one disk failure in each sub-array
Read Performance	High	High	High	High	High	High	High
Write Performance	High	Medium	Medium	High	High	High	Medium
	100%	50%	50%	67% ~ 94%	50% ~ 88%	50% ~ 88%	50%
Typical Applications	High-end workstation, data logging, real-time rendering, very transitory data	OS, a transaction DB	OS, transaction DB	Data warehousing, web serving, archiving	Data warehousing, web serving, archiving	Data archive, backup to disk, high availability solutions, servers with large capacity requirements	Fast databases, application servers

 

10. References

https://en.wikipedia.org/wiki/RAID

https://en.wikipedia.org/wiki/Data_redundancy

https://en.wikipedia.org/wiki/Standard_RAID_levels

http://www.acnc.com/advanced_raid_technology

https://slideplayer.com/slide/9404744/

http://www.electricmonk.org.uk/2013/01/03/raid-levels/

https://www.ivshub.com/understanding-different-raid-levels/

https://quickcse.wordpress.com/2018/08/12/raid/

http://www.datacenterprofessionals.net/profiles/blogs/revisiting-raid-storage-remains-relevant-and-resources

https://www.mowasay.com/2016/04/raid-level-comparison-types/

http://vceit.com/p/security-raid.htm

https://slideplayer.com/slide/6389171/

https://www.golinuxhub.com/2014/04/raid-levels-0-1-2-3-4-5-6-01-10.html

https://www.unifore.net/ip-video-surveillance/enterprise-grade-nvr-and-raid-level-knowledge.html

https://jwprogramming.tistory.com/24

https://www.booleanworld.com/raid-levels-explained/

https://www.prepressure.com/library/technology/raid

http://web.cs.ucla.edu/classes/spring09/cs111/scribe/16/index.html

https://en.wikipedia.org/wiki/Degraded_mode