Multi-level RAID systems are very complicated. Reconstruction and data recovery of multi-level RAID systems is also very complicated. Gillware is very proud of its technical staff and their RAID analysis and reconstruction skills. If you’ve ever tried to reassemble a complicated multi-level RAID system, you know that the tools available to do so are very limited. At Gillware, we analyze your system in detail to determine its parameters and if necessary, write custom reconstruction software specific to your system. Gillware looks forward to helping you with your RAID recovery needs.
Multi-Level RAID Naming Convention
Multi-level RAID naming conventions are confusing. The standard multi-level RAID naming convention that is most commonly used is to list the Sub-Array first and the Super-Array second. For example, if you have three RAID 1 Sub-Arrays made up of two disks each and put them together to make a RAID 0 Super-Array, you are implementing a RAID 1+0 system of six disks. Alternatively, if you have two RAID 0 Sub-Arrays of three disks each and put them together to make a RAID 1 Super-Array, you are implementing RAID 0+1. See our
RAID 1+0 and
RAID 0+1 pages for more details.
RAID 5+0
RAID 5+0 is a RAID 0 Super-Array consisting of RAID 5 Sub-Arrays. Data is striped into the RAID 0 array and then striped several more times with a parity calculation into several RAID 5 arrays.
| |
RAID 5 Sub-Arrays (stripes w/parity) |
|
| RAID 0 Super-Array (stripes) |
Disk 1 |
Disk 2 |
Disk 3 |
Sub-Array A |
| Disk 4 |
Disk 5 |
Disk 6 |
Sub-Array B |
RAID 0+5
RAID 0+5 is a RAID 5 Super-Array consisting of RAID 0 Sub-Arrays. Data is striped with parity into the RAID 5 array and then the data and parity are striped several more times into several RAID 0 arrays.
| |
RAID 0 Sub-Arrays (stripes) |
|
| RAID 5 Super-Array (stripes w/parity) |
Disk 1 |
Disk 2 |
Sub-Array A |
| Disk 3 |
Disk 4 |
Sub-Array B |
| Disk 5 |
Disk 6 |
Sub-Array C |
RAID 3+0
RAID 3+0 is a RAID 0 Super-Array consisting of RAID 3 Sub-Arrays. Data is striped into the RAID 0 array and then striped several more times with a parity into several RAID 3 arrays each with their own dedicated parity disk.
| |
RAID 3 Sub-Arrays (stripes w/ dedicated parity) |
|
| RAID 0 Super-Array (stripes) |
Disk 1 |
Disk 2 |
Disk 3 |
Sub-Array A |
| Disk 4 |
Disk 5 |
Disk 6 |
Sub-Array B |
RAID 0+3
RAID 0+3 is a RAID 3 Super-Array consisting of RAID 0 Sub-Arrays. Data is striped with dedicated parity into the RAID 3 array and then the data and parity are striped several more times into several RAID 0 arrays. With RAID 0+3, one of the RAID 0 Sub-Array stores all of the parity calculations.
| |
RAID 3 Sub-Arrays (stripes) |
|
| RAID 0 Super-Array (stripes w/ dedicated parity) |
Disk 1 |
Disk 2 |
Sub-Array A |
| Disk 3 |
Disk 4 |
Sub-Array B |
| Disk 5 |
Disk 6 |
Sub-Array C |
RAID 5+1
RAID 5+1 is a RAID 1 Super-Array consisting of RAID 5 Sub-Arrays. Data is striped w/parity into one RAID 5 Sub-Array and then the entire Sub-Array is mirrored. RAID 5+1 is called a "belt + suspenders" because of its extremely high fault tolerance. In the six-disk RAID 5+1 example shown below, it is possible to recover all data even after 4 out of 6 disks fail simultaneously! All that is needed to recover the data is two stripes. This means that as long as the four failed disks are NOT sets of matching stripes (disks 1, 2, 4, 5) for example, data can be recovered. As a result of this high fault-tolerance, storage efficiency is very low and cost is very high.
| |
RAID 5 Sub-Arrays (stripes w/parity) |
|
| RAID 1 Super-Array (mirrors) |
Disk 1 |
Disk 2 |
Disk 3 |
Sub-Array A |
| Disk 4 |
Disk 5 |
Disk 6 |
Sub-Array B |
RAID 1+5
RAID 1+5 is a RAID 5 Super-Array consisting of RAID 1 Sub-Arrays. Data is striped w/parity into one RAID 5 Sub-Array and then each stripe in the RAID 5 array is mirrored. RAID 1+5 is called a "belt + suspenders" because of its extremely high fault tolerance. In the six-disk RAID 1+5 example shown below, it is possible to recover all data even after 4 out of 6 disks fail simultaneously! All that is needed to recover the data is two stripes. This means that as long as the four failed disks are NOT sets of matching stripes (disks 1, 2, 3, 4) for example, data can be recovered. As a result of this high fault-tolerance, storage efficiency is very low and cost is very high.
| |
RAID 1 Sub-Arrays (mirrors) |
|
| RAID 5 Super-Array (stripes w/ parity) |
Disk 1 |
Disk 2 |
Sub-Array A |
| Disk 3 |
Disk 4 |
Sub-Array B |
| Disk 5 |
Disk 6 |
Sub-Array C |
Gillware looks forward to helping you with your multi-level RAID reconstruction needs.
To place an order please call our 800 number.
