SO, HOW DOES THIS WORK? WHAT’S REALLY GOING ON?
Assuming a 3-disk mirror, when one writes 6 blocks of data to the
mirror LUN, the 6 blocks (and likewise, subsequent blocks) will be
arranged on the 3 disks as shown below. The "D" refers to data, and
the "M" refers to mirror.
With the data arranged in this fashion, it is easy to see that if a
disk fails, the data will still be available.
disk-1 disk-2 disk-3
Data -> | blk1 blk2 blk3
Mirror -> | blk3 blk1 blk2
Data -> | blk4 blk5 blk6
Mirror -> | blk6 blk4 blk5
Note that a 3-disk mirror of 18GB (actually 17.9GB) disks yields 26.8GB
of usable space. That’s 17.9GB times 3, then divided by 2 because the
data is mirrored.
17.9 x 3
——– = 26.8
CREATING THE VOLUME:
Here is how one can create, via the CLI, a 3-disk RAID-1 (mirror).
Each of the disks used here are 18GB.
t3a:/:<10> vol add v1 data u1d1-3 raid 1
t3a:/:<11> vol init v1 data
WARNING – Volume data will be initialized to zero.
WARNING – Volume initialization can take a significant amount of time.
Continue ? [N]: Y
t3a:/:<12> vol list
volume capacity raid data standby
v1 26.8 GB 1 u1d1-3 none