RAID6 disk array introduction
1, RAID 6 arrays of N (N> = 4) block array of discs, the storage capacity of disk capacity is the sum of the N blocks (N-2) / N.
2, RAID 6 RAID 5 technology is based on a RAID system in order to further strengthen data protection and design, in fact, is an extension of RAID 5 level. Differs from RAID 5 and in addition to have the same level of XOR parity data area on each disk, there is a region XOR checksum for each data block. Of course, the current parity data disk data blocks can not exist in the current disc but interleaving memory, the specific form of the figure.
. 3, RAID 6 each data block has two check protective barrier (a layered check, a check is general), and therefore the data redundancy RAID 6 is quite good. Even if two disks while loss does not affect the use of the data, but the addition of a check, the efficiency of writing worse than RAID 5, and the design of the control system is also more complex, the second calibration region also block reducing the effective storage space.
RAID 6 disk array configurations:
1, add six to the virtual machine disk, and then try to update the system.
2, the newly added sdb, sdc, sdd, sde, sdf, sdg, be configured to set the partition type raid type. (Used here as an example the rest of the sdb disk operations are the same) 
3, then began to create a RAID5 volume.
Entry
mdadm -C -v /dev/md6 -l6 -n4 /dev/sd{b..e}1 -x2 /dev/sd{f..g}1Wherein
mdadm: command word
-C: create a RAID volume
-v: show details of the
/ dev / md6: Create a directory file
-L6: Level6
--n4: Number of disks 4
-x2: the number of spare disk of 2
/ dev / sd {f..g} 1: disk directory names and 
4, the use of cat / proc / dmstat, view disk case.
cat /proc/mdstat5, the RAID5 volume is formatted xfs format. 
6, create a file in mnt md6 as a mount point, raid6 will be mounted. 
7, it can also be mounted automatically.
Verify RAID 6 arrays
1, create some empty folders in md6 years. 
2. Check with mdadm -D / dev / md6 information about the disk, you can see sdb1, sdc1, sdd1, sde1 is active, sdf1, sdg1 standby. 
3, enter mdadm -f / dev / sdb1 and sdb2 deleted, see sdf1, whether sdg1 will replace up. 
4, input mdadm -D / dev / md6 View active state disk, you can see sdf1 and sdg1 in the active state and replace sdb1 and sdb2. 
5, this time to see if we mnt / md6 file is still there.