1.5.3 STRIPING (FAST)
The STRIPING storage policy distributes access across all hard disks, also called
FAST
(equivalent to RAID 0).
STRIPING presents the best data speed but no data redundancy. STRIPING storage policy accelerates hard disk
operating speed by using many disks in parallel. Hard drive data segments are written to different disks
simultaneously which increases performance while sacrificing data redundancy.
To implement the STRIPING storage policy, the AES-3502UE Storage creates a single virtual volume that is
striped
across both hard drives, with a storage capacity that is equal to the sum of both hard disk drives.
It is possible to set any node within a cascaded configuration to STRIPING, although there is no performance
benefit when using multiple layers of striping.
1.5.4 MIRRORING (SAFE)
The MIRRORING storage policy stores all data in duplicate on separate drives to protect against data loss due to
drive failure. One drive
mirrors
the other at all times, equivalent to RAID 1. Every write operation goes to both
drives. MIRRORING provides the highest level of data protection for critical data that you cannot afford to lose if a
hard drive fails, but halves the amount of storage capacity because all data must be stored twice. The resulting
storage capacity of the virtual MIRRORING volume will be equivalent to the size of one hard drive (if both drives
are the same) or the smaller of the two drives (if they are different).
If one drive fails, the MIRRORING volume is still usable, but it is in a vulnerable state because its mirrored hard
drive is inaccessible. When the offline drive comes back online, the appliance begins a rebuild process
immediately to restore data redundancy. A message appears in the LED indicator to notify you that a rebuild is in
progress.
Although the volume remains available during the rebuild process, the volume is susceptible to data loss through
damage to the remaining drive until redundancy is restored at the end of the rebuild and verification process. Host
access takes precedence over the rebuild process. If you continue to use the MIRRORING volume during the rebuild,
the rebuild process will take a longer time to complete, and the host data transfer performance will also
be affected.
Any node within a cascaded configuration can be set to MIRRORING, but it is more efficient to use the
MIRRORING policy at the lowest possible level within the hierarchy.
It is also possible to create a MIRRORING volume using one hard disk drive connected to Port 0 of the TEAR
Ver. 071004
AES-3502UE User’s Manual
8
