HP StorageWorks EVA Controller HSG V8.8 Software - Manual (Page 133)

Glossary
133 Command Console V2.5 User Guide
maximum block
transfer
For HS-series controllers, the virtual disk parameter that
specifies the maximum number of data blocks to be cached
before writing data to virtual disk members. A data transfer
greater than this size is not cached and is written immediately to
disk. Valid values are 1-1024 (blocks).
member
Any device used in a virtual disk based on multiple devices.
Any device in the device group on which the virtual disk is
based.
metadata
Special data written to a device for the purpose of controller
administration. Metadata improves error detection and media
defect management for a device. It is also used to define device
and virtual disk configuration.
mirrored cache
Some controllers offer high-performance cache hardware that
can be set to operate in mirrored mode. In this mode, the cache
data is duplicated and stored in cache memory in physically
separate locations. If one copy of the data becomes corrupted or
unavailable for some reason, the other copy is available for use.
A mirrored cache configuration provides complete protection
for cached data. However, the cache memory effectively
contains twice as much data for each host request, so the
effective size of the cache is cut in half.
mirrored virtual disk
A group of storage devices organized as duplicate copies of
each other. Mirrored virtual disks provide the highest level of
data availability at the highest cost. Another name for RAID 1.
mirroring
The simplest form of data redundancy. Two or more devices
form a mirrored device group. Each device is an exact copy of
the other.
For read requests, data can be read from either device therefore
increasing the throughput. Both devices can handle different
requests simultaneously.
For write requests, data is written to both devices. If one device
fails, all reads/writes are executed on the mirrored device
group’s other device. Therefore, mirroring provides excellent
availability (unless the second device fails before the first
(failing) device can be repaired or replaced).
Sometimes called shadowing.
The mirroring technique can also be applied to cache memory.