HP NetRAID-4M Configuration and Upgrade Guide (Release 5)

Chapter 10 Glossary

RAID containers (NetRAID-4M)

Containers are logical disks (also known as arrays) created from free space and made

up of one or more partitions on one or more physical disks. A container that spans

multiple physical disks can be larger than any one of the physical disks. Containers

differ from most other RAID arrays in that their underlying partitions can be smaller

than a physical disk. Consequently, several containers' partitions can reside on a

single physical disk. A container's file system appears in Windows Explorer as a disk

drive with its own drive letter. A single level container is made up of one or more

partitions. A multi-level container is made up of one or more single level containers.

RAID critical

If one of the members of the container fails, that container is marked as “critical.” If

this is RAID5, the container can be forced back to its original configuration without

rebuilding.

RAID degraded

Although the term is sometimes used to mean “critical”, degraded is more

commonly used to describe a container that is rebuilding while it is being accessed,

in effect, resulting in degraded performance.

RAID disk verify

Equivalent to the "background consistency check" option that would be enabled

through the BIOS or the FAST GUI. The Verify command checks only the

underlying data blocks, not the file system or parity. Note that verifying a disk is not

the same as running chkdsk under Windows NT. RAID disk verify is recommended

in the following cases:

• Verify a disk after adding it to your system to determine if the disk has any

obvious defects

• Occasionally check a disk (without the repair option) to be sure it is reliable

• Verify a disk before removing any dead partitions

RAID fault tolerance illustration

Error detection requires single dimension checking (parity). Error correction

requires two-dimensional checking. If a single bit error can be located in two

dimensions, then the error can be corrected. In this illustration, data is written

horizontally bit by bit across the drives in sequential stripes. Parity is written into a

separate drive.