Table of contents
When your computer requests data from a single storage drive, it can only retrieve it as fast as the single drive can provide it. What if you could access multiple drives simultaneously, allowing data to be read/written significantly faster?
Think of your file as a group project in school. You can either do the entire project yourself and take a week or divide it amongst seven other students and complete it in a day. That is how data striping works.
With the increasing demand for higher storage capabilities, improved performance, and data loss situations, RAID has emerged as a viable storage solution.
Definition
RAID stands for Redundant Array of Inexpensive Disks or Redundant Array of Independent Drives. This technology is developed to increase the performance and reliability of data storage. A RAID system consists of multiple drives that work in parallel. This setup can comprise hard disk drives or SSDs (solid state drive).
All these drives are linked together to prevent data loss and improve performance. This becomes possible by mirroring data on one or more drives in the same array. Drive mirroring helps in preserving data if one or more drives fail. To improve performance, a stripping technique is used that spreads data over multiple drives.
RAID levels
RAID storage devices can be configured in multiple ways, known as RAID levels. The different RAID levels determine how data is dispersed among multiple drives. Initially, there were six levels of RAID – 0 to 5. However, as the number of levels expanded, the RAID level was divided into three categories: standard, nested, and nonstandard RAID levels. The Standard RAID level consists of RAID 0- 6. The Nested RAID (a combination of RAID levels) includes RAID 10 (RAID 1+0), RAID 01 (RAID 0+1), and RAID 50 (RAID 5+0). The non-standard RAID levels are RAID 7, RAID 1E, RAID S, and more.
All these different RAID levels offer different levels of performance and reliability. However, based on the techniques used, the most common types of RAID are level 0, level 1, level 5, level 6, level 10, level 50, and level 60.
Comparison
There are multiple RAID levels and each level offers a unique combination of performance and redundancy. The RAID level you choose will determine the speed and fault tolerance you can expect from the RAID setup. When choosing a RAID level, you need to consider many factors, such as cost, reliability, capacity, and performance. So, let’s compare the most popular RAID levels to help you select the best level for your storage system. Below we’ve discussed the common RAID levels and the pros and cons of each type.
RAID 0 (stripe)
RAID 0, also called striping, works by dividing data evenly across two or more drives (HDD or SSD). RAID o divides data into blocks and is distributed across the drives in the array. This helps in speeding up performance but RAID 0 fails to provide redundancy. This means this level offers superior I/O performance but cannot be used for mission-critical situations. The performance of RAID 0 can be further improved by using multiple controllers- one controller per drive. Due to superior overall performance,
RAID 0 is best suited for applications where improved performance is the primary driver. It is ideal for non-critical data storage where high read/write speed is more important. Drive striping is most suited for large applications like image retouching or video editing. However, due to a lack of data redundancy, any raid data failure could result in complete data loss.
Advantages
- Superior performance- both in reading and writing operations
- RAID 0 is the most affordable and easy to set up
- Full capacity use, no overhead
- The technology is easy to implement
Disadvantages
- RAID 0 is not fault-tolerant (no redundancy). Single-drive failure could lead to complete data loss.
RAID 1 (mirror)
RAID level 1 stores data twice by writing them to the data drives and the mirror drives. With mirroring and duplexing, data is duplicated on a second drive. This means that if one drive fails, data can be read or written from the other drive. RAID 1 offers fault tolerance because it duplicates data by simultaneously writing on two storage devices. In RAID 1, each drive has an exact copy on another drive to ensure protection against data loss. But on the downside, RAID 1 doesn’t provide any performance improvements.
This technique does not include parity or striping. This RAID level can be used for mission-critical storage and small servers.
Advantages
- Increased read speed
- Fault-tolerant
- Simple technology, easy to set up
- Doesn’t require data rebuilding if a driver fails
Disadvantages
- Uses only half of the storage capacity because all data get written twice
- More expensive
- If a drive fails, you need to power down your computer to replace a failed drive.
RAID 5 (parity)
RAID level 5 uses drive striping and parity to strip data across three or more drives. This is the most popular RAID level for organizing independent drives. As one of the most common secure RAID levels, RAID 5 can comprise 3 to 16 drives. In this technique, data is striped but not duplicated. Here the parity data are spread across all drives to rebuild data if needed. This means that the RAID 5 array can withstand a single drive failure without losing data.
This RAID level ensures protection against drive failure and provides high performance and reliability. RAID 5 offers a great combination of data security and improved read/write performance. With good storage, superior security, and reliable performance, RAID 5 can be used for applications that have a limited number of data drives.
Advantages
- Better performance because data can be read or written simultaneously across multiple drives
- Fault-tolerant because even if a drive fails, data can be accessed from other drives
- Consistent access to all data
Disadvantages
- Uses only half of the storage capacity
- It is a complex technology, and RAID rebuild may take a day or longer
- Parity overhead that causes lower performance rates
RAID 6 (double Parity)
RAID 6 is one of the most commonly used RAID levels for business use. It is similar to RAID 5; however, this level writes the parity data to two drives. RAID 6 needs at least four drives and can withstand two drives failures at a time. This technique combines block-level striping with double-distributed parity. RAID 6 stripes the data and parity information across all drives, duplicating redundancy information.
This level offers improved performance and can withstand the complete failure of two drives. RAID 6 is considered safer than RAID 5 but data writing can be slower than RAID 5. This happens because RAID 6 involves calculating the additional parity data. RAID 6 is the right choice for application servers that use many large drives for data storage. Even though this type of RAID is expensive to build, it gives you much confidence when it comes to avoiding RAID data loss scenarios.
Advantages
- Data reading transactions are super fast
- Data can be accessed even when two drives fail
Disadvantages
- Write data transactions are slower than RAID 5
- Complicated to implement, and rebuilding an array can take a long time
RAID 10 (mirror + stripe)
RAID 10 combines RAID 0 and RAID 1, where data is mirrored and striped simultaneously. By combining RAID levels 1 and 0, this level achieves the benefits of both levels. This means that RAID 10 gets the high performance of RAID 0 and fault-tolerance of RAID 1. RAID 10 has a minimum of four (4) drives, where two (2) drives are striped and mirrored onto two (2) other drives to create a single array. It is noteworthy here that RAID 10 and 01 are not the same.
Advantages
- Excellent performance
- Fault-tolerant
- Quick to rebuild
Disadvantages
- More expensive
- Limited scalability
RAID 50 (parity + stripe)
RAID 5+0 is a combination of RAID 0’s block-level striping with the distributed parity of RAID 5. This configuration combines several RAID 5 arrays and strips them together. With a minimum of 6 drives, RAID 50 can sustain one disk per sub-array failure without data loss. It means that a RAID 50 array with three RAID 5 sets can withstand three drive failures simultaneously. This configuration is expensive and more complex but performance and fault tolerance is much superior, with usable capacity starting at 67%.
Advantages
- Better Performance than RAID 5
- High Fault Tolerance
Disadvantages
- Costly
- Limited Scalability
RAID 60 (double parity + stripe)
RAID 6+0 is a combination of RAID 0 array striped across RAID 6 elements. With a minimum of 8 disks (2 sets of 4 disks), RAID 60 consists of several RAID 6 sub-arrays striped together. It means that RAID 60 can withstand two disks failure per sub-array without data loss. It has extreme fault-tolerance capacities but read performance is less than RAID 50.
Advantages
- Higher Fault Tolerance than RAID 6
- Superior Performance
- Higher Efficiency
Disadvantages
- Cost
- Complex Implementation
Comparison table
RAID level | Min. drives | Fault tolerance | Read Speed | Write Speed | Efficiency |
---|---|---|---|---|---|
RAID 0 | 2 | None | High | High | 100% |
RAID 1 | 2 | 1 drive | High | Medium | 50% |
RAID 5 | 3 | 1 drive | High | Low | 67% – 94% |
RAID 6 | 4 | 2 drives | High | Low | 50% – 88% |
RAID 10 | 4 | 1 drive in each sub-array | High | Medium | 50% |
RAID 50 | 6 | 1 drive in each sub-array | High | Medium | 67% – 94% |
RAID 60 | 8 | 2 drive in each sub-array | High | Medium | 50% – 88% |
Cost, speed and space calculator
Now you know what RAID is and its different levels. If you would like to purchase or upgrade your RAID system and opt out for cloud backup because of slow upload speed and monthly charges to store your files, then you can use check out a free RAID calculator tool. It can calculate the cost, speed, and usable space.