Array
A group of several hard disks on which one or more logical
drives are contained is called an array.
Array Initialization
See Initialization.
ATA
AT Bus Attachment. Standard parallel interface to IDE hard disk drives which is
commonly used in desktop computers and some entry-level servers. A successor of
ATA is SATA (Serial ATA interface).
Automatic Rebuild
Automatic procedure which restores faulty segments on preconfigured volumes. If
a drive in a SAF-TE enclosure which has no hot spare assigned should fail, a
rebuild starts only when the failed drive has been replaced by a new one.
Background Initialization
In the case of background initialization of a drive the redundant logical drive
can be accessed while the initialization is running.
BGI
Background Initialization is a type of initialization
which is started automatically in the background at low priority by an Avago
controller. See also Initialization and
Background Initialization.
Bus
See Channel.
Cache
Quickly accessible memory on a controller which serves as a buffer for data
which is read from or written to devices.
Capacity
Total storage capacity available on a drive; often specified in megabytes or
gigabytes. A distinction is made between physical and logical capacity.
CAS
CAS (Central Authentication Service) is an enterprise single sign-on solution for web services.
Single sign-on (SSO) means a better user experience when
running a multitude of web services, each with its own means of authentication.
With a SSO solution, different web services may authenticate to one authoritative
source of trust, that the user needs to log in to, instead of requiring the
end-user to log in into each separate service.
Channel
Path between a controller and storage devices used for data transfer and
control of information. Each channel of a controller is identified by a number
between 0 and the maximum number of channels minus 1. A channel is also
referred to as a port or bus.
Concatenation
Connection in sequential order.
Consistency Check
An action in which the controller checks all segments of the logical dive. Depending
on the logical drive type various things can be checked and the checks can therefore
take different lengths of time. RAID-5 checks the data for consistency and parity.
RAID-1 checks whether both drives are consistent. See also
MDC and Verify.
Create
See Automatic Rebuild.
Dedicated Hot Spare
A physical drive which, when required, takes over the place of a failed
physical drive in a specially assigned, fault-tolerant logical drive.
Degraded
A redundant logical drive in which one or more members have failed. The data is
still intact, but redundancy has been impaired and is in a worse status. The
logical drive and all the data are still available, but a further drive failure
leads to the failure of the logical drive and loss of data. A check plus
correction can return an impaired logical drive to its optimal status.
Dirty Data
Data which has been written into a cache and has not yet been updated on the
actual target volume.
Drive
See Physical and Logical Drive.
Enclosure
An enclosure for physical drives which generally contains several power
supplies, fans, and temperature sensors. Enclosures are normally outside the
computer to which they are connected. Some computers also contain internal enclosures.
Event
Notification or warning message when changes occur in the system.
Event Log
Information on controller activities or other events are stored in a file.
Expand
Assignment of more storage space to a logical drive.
Failed
Status of a non-redundant logical drive with a single drive failure or a
redundant logical drive with multiple drive failures. The status generally
results in a loss of data as access to the logical drive is no longer possible.
Failed Segment
A segment which is no longer used by a logical drive because it is either
logically or physically damaged.
Fast Initialization
The logical drive is available immediately, but has a special internal status.
In RAID-5 and RAID-50 the write performance is affected until a check plus
correction has been performed on this logical drive.
Fault-tolerant Logical Drives
Logical drives with redundant components (RAID-1, RAID-5, RAID-10, RAID-50).
Format
A process performed by the firmware in which all data on the hard disk is
totally deleted.
Global Hot Spare
Volume which can replace a failed component in the logical drives on the same
controller. The available storage capacity must be at least as large as that of
the failed component. See also Hot Spare.
Hot Spare
A physical volume available as a replacement if a drive fails. In a redundant
logical drive this permits automatic data recovery. The storage capacity
available must be at least as large as that of the failed component. See also
Automatic Rebuild, Global Hot Spare and
Dedicated Hot Spare.
Hot Swap
Replacement of system components while the system is running.
Initialization
A fault-tolerant logical drive must be initialized before it is used. This
operation deletes all blocks on the logical drive. In RAID-5 a parity is then
generated on the basis of the current content of the member segments, whereas
RAID-1 copies the content of the first drive (master) to a second drive
(slave). Depending on the RAID type initialization runs in the background at
differing speeds. In addition, Fast Initialization
is also provided for immediate access to a RAID-5 drive.
Initialized Logical Drive
A logical drive which is ready for read and write operations.
Java
Java is an object-oriented, platform-independent programming language developed
by Sun Microsystems. All Java programs run on a wide range of platforms without
any adjustments. However, to execute, JAVA programs generally need a special
runtime environment, the virtual machine, and only this environment need be
adjusted to the various operating systems.
JBOD
JBOD stands for Just a Bunch Of Disks. According to the latest definition of
the Storage Networking Industry Association (SNIA) a single hard disk should be
understood here today, whereas earlier it referred to multiple hard disks (also
referred to as concatenation).
KByte
Is a unit of information or computer storage, commonly abbreviated as "KiB" and
stands for 210 bytes = 1,024 bytes.
LBA
Logical Block Addressing is an addressing method for hard disks. In contrast to
other methods the sectors on the hard disk are counted singly, beginning with 0.
LED
A light-emitting diode is an electronic semiconductor which is used, for
example, to indicate read or write operations on hard disks. Often an LED is
also used for determining the location of the drives.
Locate
An aid in uniquely identifying hard disk which consists of the controller
number, the channel or port number, the LUN, and the SCSI-ID.
Logical Drive
A drive which consists of one or more physical drives, mostly hard disks. For
the operating system, all the storage volume available constitutes a single drive.
Logical Drive Order
The order in which, at system startup, the server's operating system recognizes
individual hard disks and other devices connected to the controller.
LUN
Each SCSI device can contain up to eight subdevices. A LUN is the number of the
logical unit - 0 through 7 - which has been assigned to this device. However,
generally only one subdevice (LUN 0) exists.
MDC
Make Data Consistent is a consistency check with additional correction.
Depending on the drive type various things can be checked and the checks can
therefore take different lengths of time. RAID-5 checks the data for
consistency and parity and optionally corrects parity errors. RAID-1 checks
whether both drives are consistent. Either if inconsistencies found it will be try to
correct if is possible to fix it. See also
Consistency Check and Verify.
Migration
Process of transforming a logical drive from one RAID type to another,
changing the strip size, or expanding the size of a logical drive by
adding new disks.
Monitoring
The process for ascertaining, displaying, and logging system events.
Notification
Aid used by the system for communication relating to events that have occurred.
NVRAM
A non-volatile memory which can retain information even if the power supply is
not maintained. This memory is often used on RAID controllers as a memory for
the configuration and for error logging.
Offline
The status of a logical or physical drive which can no longer be accessed.
Offset
The distance from the start of a drive to the start of a segment.
Optimal
A logical drive in its normal operating status in which all components are
present and fully operational.
Parity
A form of redundancy which is used for checking the user data for errors.
Additional data is generated from the user data which is also stored and can be
used to reconstruct the original data.
Patrol Read
Patrol Read examines the drive's interface for errors. If an error is detected
which cannot be rectified automatically, this error is logged. The faulty part
of the medium is assigned to the defective blocks. With some controllers the
disk is automatically taken out of service if multiple errors are found.
Physical Drive
Generally a physical hard disk drive, also known as hard disk for short. This
is a randomly accessed, rewritable data storage device.
Port
See Channel.
RAID Signature
From the RAID signature on a hard disk the RAID controller can recognize, among
other things, whether the hard disk has already been initialized. In future it
is to contain a complete, non-proprietary RAID configuration to make it easier
to exchange drives or controllers.
RAID Volume
Two or more logical drives of the same type which
need not necessarily have the same capacity are connected in a RAID volume.
RAID-0
A logical drive with one level, consisting of two equal-sized segments on
different hard disk drives. RAID-0 uses the striping
process to distribute the data evenly in equal-sized sections over the
drives concerned.
RAID-0/1
See RAID-10.
RAID-1
A logical drive with one level, consisting of two equal-sized segments on different
hard disk drives. Offers redundancy by storing the same data on both hard disks.
RAID-5
A logical drive with one level, consisting of three equal-sized segments on
different hard disk drives. The capacity of one segment is used for parity data
which is distributed over all the drives in equal-sized sections.
RAID-10
A logical drive. Two equal-sized RAID-1s are used to create a RAID-10.
RAID-10 thus needs four physical drives.
RAID-50
A logical drive. Two equal-sized RAID-5s are used to create a RAID-50. RAID-50
thus needs six physical drives.
Reconfiguration
See Migration.
Redundancy
Redundancy is the general term to describe the additional existence of resources
which are functionally identical or comparable if these are not normally required
when operation is error-free. Here redundancy is used to manage data in a system
with the aim of automatically replacing failed components by operational ones.
For example, logical drives of the type RAID-5 are redundant because the surviving
members can combine to replace the data of a failed component.
Replacement
See Hot Spare.
Role Based Access Control
Role based access control (RBAC) manages access control by defining a set of
user roles (security roles). One or more roles are assigned to each user, and
one or more user privileges are assigned to each role.
SAF-TE
SCSI Accessed Fault-Tolerant Enclosure.
SAS
Serial Attached SCSI is to replace the current parallel SCSI interface because
the Terminators which are typical for SCSI are not
required for SAS. SAS takes over the SATA connections. SATA devices can be used
on SAS but not vice versa.
SATA
Serial ATA is a successor to ATA which transfers data
serially instead of in parallel.
SCSI
Small Computer System Interface is a parallel high-speed communication method
which permits data transfer rates of up to 320 Mbytes/s. The current
specification supports up to 15 devices per channel.
SCSI ID
A unique number (0-15) which is assigned to each device on a SCSI bus.
Second Level Array
A logical drive can consists of more than one level. The logical device of the
second level (lower level array in an array with two levels) is never visible
for the operating system and can only be used by other logical devices. For
example, a RAID-10 array's member in the top level is a RAID-0 array and in the
second level two or more RAID-1 arrays.
Segment
Reserved area on a physical hard disk. A segment is always part of a logical
drive and cannot be used by more than one logical drive at the same time.
Single Sign-On (SSO)
Single sign-on allows participating applications to share a single sign-on session.
Users complete a centrally managed authentication experience once per
browser session and while authenticated to CAS can log into multiple applications
without again being prompted for credentials and without those applications
ever seeing the user's password.
SMART
Self-Monitoring Analysis and Reporting Technology (SMART). This drive
function is designed to determine the reliability status of a hard disk drive.
If SMART detects a potential problem that could be serious, the user is
notified and receives assistance on how to rectify this problem.
Spare Drive
See Hot Spare.
Striping Process
Small, contiguous data areas which are distributed over all hard disks in the
logical drive are called stripes. For example, with the striping process RAID-0
distributes the data in equal-sized sections over the drives concerned to
expedite access.
Stripe Size
In logical drives using the striping process (RAID Levels 0, 5, 10 and 50) the
data is distributed over the member hard disks in equal-sized sections. The
amount of data in each section is the stripe size.
Temperature Warning Threshold
The temperature threshold defined by the user above which a warning message is issued.
Terminator
In a SCSI bus system both ends of the cable route must be provided with a
terminator. There are two types of termination, the passive one using resistors
and the active one using an internal voltage source.
Verify
An action in which the controller checks all segments of the hard disks. This determines
whether the hard disks are sending back data from the blocks. See also
Consistency Check and MDC.
Virtual Drive Order
See Logical Drive Order.