RAID 6 Recovery
RAID 6 parameters
RAID 6 layout utilizes two different parity functions which helps a RAID 6 array survive up to double disk failure. Despite of the fact that RAID6 is perceived to be more reliable than a RAID 5, the probability of a controller failure is the same. So if your RAID6 fails, this will most likely be due to a controller failure. To recover a RAID6 from a controller failure, you need to know or recover the following parameters:
- number of member disks
- disk order and which disk was the first in the array
- block size
- start offset on the member disks
- positions of each parity function
- rotation of parity functions
- algorithm used to calculate the second parity function
- pattern of parity functions
You should know all these parameters to reconstuct RAID 6 successfully. Our software ReclaiMe Free RAID Recovery (download) can recover RAID6 configuration automatically. See the step-by-step RAID Recovery instructions.
RAID 6 variations
RAID 6 is significantly more complex than RAID 5.
- The first parity function (XOR, denoted as P on the diagrams) naturally comes from RAID5. However, there is no widely accepted standard for the second parity function (Q). Typically, a Reed-Solomon code or some variation thereof is used.
- Reed-Solomon code depends on the order in which input is provided. This produces four more variations: left-to-right or right-to-left and if P or Q is calculated first.
- There are several variations of layout accomodating two parity functions.
Missing disk limitations
Maximum of one missing disk is allowed. Although RAID 6 should be recoverable with two disks missing, the implementation of ReclaiMe Free RAID Recovery cannot work with two disks missing.
With Promise hardware RAID 6, no missing disks are allowed at all (as of build 381).
Post a Comment