Key Points
- RAID 10 is a combination of RAID 0 and RAID 1.
- RAID 10 has higher fault tolerance than RAID 5 as it maintains an exact duplicate striped volume of the main striped volume.
- RAID 5 uses parity to create fault tolerance while RAID 10 uses disk mirroring.
- RAID 10 is a better choice for databases than RAID 5.
- RAID 5 takes a decent amount of time to rebuild applications and cannot be used while rebuilding.
- RAID 5 can experience an Uncorrectable Read Error (URE) which does not occur with RAID 10.
- RAID 10 can be fixed quicker and used during repair.
RAID (Redundant Array of Inexpensive Discs) is virtualization technology which can be used to combine several inexpensive disc drives into a single hard disc drive. RAID 5 requires a minimum of three drives and is an improved version of RAID 0, while RAID 10 is a combination of RAID 0 and RAID 1. Read on for a full explanation of the two, similarities and differences.
What is RAID?
© g0d4ather/Shutterstock.com
RAID is a virtualization technology used to combine multiple independent or inexpensive disk drives into one or more logical units for data security, performance improvements, or both.
It uses three different techniques combined with a physical dock for the drives to save data across multiple drives, clone the main drive in real-time, and check for data integrity.
The three techniques are called:
- Data striping
- Disk mirroring
- Parity
Data Striping
Data striping is the process of writing file contents split up on multiple drives within the RAID array. Striping can help with read/write speeds and overall performance.
Disk Mirroring
Disk mirroring creates an exact copy of a drive-in real-time, meaning every change is made to both the original drive and the clone. Disk mirroring provides data security by creating an exact clone of the main drive as long as each drive in the array has the same or more storage capacity than the cloned drive.
Parity
Parity is a simple error detection code used to check data integrity.
RAID arrays use either one or more of these techniques to increase performance or provide data redundancy. The different RAID arrays are called RAID levels.
Here is a list of RAID levels:
- RAID 0: Data striping with no disk mirroring or parity.
- RAID 1: Disk mirroring with no parity or striping.
- RAID 2: Bit-level data striping that uses Hamming-code parity.
- RAID 3: Byte-level data striping with dedicated parity.
- RAID 4: Block-level striping with dedicated parity.
- RAID 5: Block-level striping with distributed parity.
- RAID 6: Block-level striping with double distributed parity.
- RAID 10: Combination of data striping (RAID 0) and disk mirroring(RAID 1).
This is not an all-inclusive list of RAID levels, but it should give you a general understanding of the differences between RAID levels. Some RAID levels require specific hardware and are made for specific purposes.
What is RAID 5?
RAID 5 is a more secure version of RAID 0 that requires a minimum of three hard drives. It uses block-level striping with distributed parity. That means it saves data across multiple hard drives in larger sections than bit-level or byte-level striping. The parity code is distributed across the drives along with the file contents.
RAID 5 is used to increase operation speed by using multiple drives’ throughput to increase read and write speeds. Overall performance can be increased by adding more disks, but the price will increase with it.
What is RAID 10?
©Pixza Studio/Shutterstock.com
RAID 10 is a combination of RAID 1 and RAID 0. For this reason, it is sometimes referred to as RAID 1+0 or RAID 1&0. That means it performs disk striping and disk mirroring.
To set up a RAID 10 array, you’ll need at least four hard drives. First, you’ll need to create two RAID 1 mirrored volumes. Then, you’ll use these mirrored volumes as to the RAID 0 striped volumes. The result is that data is broken into stripes and written across two different drives which increases the drive performance against a single independent disk.
Then, those striped volumes are cloned in real-time to the mirrored volumes. This allows the user to get the overall performance advantages of a RAID 0 array along with the security of storage redundancy.
RAID 5 vs. RAID 10: Side-by-Side Comparison
| RAID 5 | RAID 10 | |
|---|---|---|
| Primary use: | Instruct RAID array on how to handle storage devices in the array | Instruct RAID array on how to handle storage devices in the array |
| Conceived: | 1986 | The 1980s |
| Initial release: | 1986 | 1988 |
| Technical committee: | IBM | IBM |
| Influential developers: | Norman Ken Ouchi, David Patterson, Garth A. Gibson, Randy Katz, IBM | Norman Ken Ouchi, David Patterson, Garth A. Gibson, Randy Katz, IBM |
| Open format: | Yes | Yes |
| Technologies influenced: | Storage drives | Databases, storage drives |
Pros and Cons of RAID 5
| Pros | Cons |
|---|---|
| Uses parity for fault tolerance | Slower read/write speeds than RAID 10 |
| Can tolerate a single disk failure | Difficult to repair and can be unrepairable |
| Increases speed compared to a single drive. | Fewer data integrity as it requires a parity calculator to recover lost data and can fail. |
| Lower price than RAID 10 array | |
| Cheaper to create higher data storage capacity |
Pros and Cons of RAID 10
| Pros | Cons |
|---|---|
| 100% data storage integrity as main striped volumes are mirrored. | Uses 50% of available data storage due to disk mirroring |
| Faster overall performance when compared to RAID 5. | Requires a minimum of four disk drives in the array |
| Suitable for high-level applications like database management. | |
| Uses disk mirroring and disk striping. | |
| Easy to repair if data loss or a disk failure occurs |
Similarities and Differences
| Similarities | Differences |
|---|---|
| Both RAID 10 and 5 use disk striping. | RAID 10 will only use half of the total storage capacity in the array due to mirroring. RAID 5 starts with a higher storage capacity and can be increased quickly and easily. |
| Both RAID levels offer fault tolerance. | RAID 5 relies on parity code and can tolerate a single disk failure. However, RAID 10 has 100% redundancy which provides far better data integrity. |
| Both are RAID levels used to instruct a computer to handle storage devices in the array. | RAID 10 has faster write and read speeds than RAID 5 as RAID 5 requires that both the data and parity are read before writing additional parity or data to the volume. |
| Both combine multiple independent disks into one or more virtualized data volumes. | RAID 5 offers less data security and is not recommended for critical applications, while RAID 10 can be used for critical applications and databases. |
| RAID 10 can not tolerate a single drive failure, RAID 5 can. | |
| RAID 5 requires a minimum of three disks to set up, while RAID 10 requires a minimum of four disks. |
