Glossary

attach logging device

To add a logging device to an existing trans metadevice. If the trans metadevice is mounted, DiskSuite attaches the log when the file system is unmounted or the system is rebooted.

attach submirror

To add a submirror to an existing mirror. DiskSuite automatically resyncs the submirror with other submirrors.

block

A unit of data that can be transferred by a device, usually 512 bytes long.

boot

To start a computer program that clears memory, loads the operating system, and otherwise prepares the computer.

browser

In DiskSuite Tool, a window for browsing through DiskSuite objects in list form. There is a separate browser for slices, metadevices, and hot spare pools.

byte

A group of adjacent binary digits (bits) operated on by the computer as a unit. The most common size byte contains eight binary digits.

canvas

In DiskSuite Tool, the main region where DiskSuite objects are displayed and manipulated.

collapse

A DiskSuite Tool command that decreases (minimizes) the size of DiskSuite objects, as shown on the canvas.

commit

A DiskSuite Tool command that commits changes that have been made to DiskSuite objects. The changes are stored in the md.cf file.

concatenated metadevice

See concatenation.

concatenated stripe

A metadevice made of concatenated groups of striped slices.

concatenation

In its simplest meaning, concatenation refers to the combining of two or more data sequences to form a single data sequence. In DiskSuite:

(1) Another word for concatenated metadevice.

(2) Creating a single logical device (metadevice) by sequentially distributing disk addresses across disk slices.

The sequential (serial) distribution of disk addresses distinguishes a concatenated metadevice from a striped metadevice.

configuration

The complete set of hardware and software that makes up a storage system. Typically, a configuration will contain disk controller hardware, disks (divided into slices), and the software to manage the flow of data to and from the disks.

configuration log

A history (log) kept by DiskSuite Tool of all top-level operations and input-validation errors during a session.

controller

Electronic circuitry that acts as a mediator between the CPU and the disk drive, interpreting the CPU's requests and controlling the disk drive.

cylinder

In a disk drive, the set of tracks with the same nominal distance from the axis about which the disk rotates. See also sector.

detach logging device

To remove a logging device from a trans metadevice.

detach submirror

To remove a submirror's logical association from a mirror.

Disk View window

In DiskSuite Tool, a graphical view of the physical devices attached to the system. It can be used to show the relationship between the logical and physical devices.

diskset

A set of disk drives containing logical devices (metadevices) and hot spares that can be shared exclusively (but not concurrently) by two hosts. Used in host fail-over solutions.

DiskSuite objects

In DiskSuite Tool, a graphical representation for the state database, metadevice or part of a metadevice, or hot spare pool.

driver

Software that translates commands between the CPU and the disk hardware.

drop site

In DiskSuite Tool, the region of the Disk View window where any metadevice, group of metadevices, or physical device can be dragged and dropped. The physical layout of the device mappings is displayed after the metadevice is dropped on a specific color in the drop site.

encapsulate

To put an existing file system into a one-way concatenation. A one-way concatenation consists of a single slice.

Evaluate

A DiskSuite Tool command that displays errors and warning messages in the configuration log for the selected metadevice.

Expand

A DiskSuite Tool command that increases (magnifies) the view of DiskSuite objects.

fault tolerance

A computer system's ability to handle hardware failures without interrupting system performance or data availability.

formatting

Preparing a disk to receive data. Formatting software organizes a disk into logical units, like blocks, sectors, and tracks.

full mirror resync

See resyncing.

Gbyte

(Gigabyte), 1024 Mbytes (or 1,073,741,824 bytes).

head

In a magnetic disk drive, an electromagnet that stores and reads data to and from the platter. Controlled by a disk controller.

high-availability

A term describing systems that can suffer one or more hardware failures and rapidly make data access available.

hot spare

A slice reserved to substitute automatically for a failed slice in a submirror or RAID5 metadevice. A hot spare must be a physical slice, not a metadevice.

hot spare pool

A group of hot spares. A hot spare pool is associated with submirrors or RAID5 metadevices.

icon well

In DiskSuite Tool, the region containing icons that are the source for new DiskSuite objects. Icons are used as templates to create metadevices and hot spare pools. See also templates.

interlace

(1) To distribute data in non-contiguous logical data units across disk slices.

(2) A value: the size of the logical data segments in a striped metadevice or RAID5 metadevice.

interleave

See interlace.

Kbyte

(Kilobyte), 1024 bytes.

latency

The time it takes for a disk drive's platter to come around to a specific location for the read/write head. Usually measured in milliseconds. Latency does not include the time it takes for the read/write head to position itself (head seek time).

local diskset

A diskset that is not in a shared diskset and that belongs to a specific host. The local diskset contains the metadevice state database for that specific host's configuration. Each host in a diskset must have a local diskset to store its own local metadevice configuration.

logical

An abstraction of something real. A logical disk, for example, can be an abstraction of a large disk that is really made of several small disks.

logging

Recording UFS updates in a log (the logging device) before the updates are applied to the UNIX file system (the master device).

logging device

The slice or metadevice that contains the log for a trans metadevice.

master device

The slice or metadevice that contains an existing or newly created UFS file system for a trans metadevice.

Mbyte

(Megabyte), 1024 Kbytes.

md.cf

A backup file of the DiskSuite configuration which can be used for disaster recovery. This file should not be edited or removed. It should be backed up on a regular basis.

md.conf

A configuration file used by DiskSuite while loading. It can be edited to increase the number of metadevices and disksets supported by the metadisk driver.

mddb.cf

A file to track the locations of state database replicas. This file should not be edited or removed.

md.tab

An input file that you can use with the command line interface utilities metainit(1M), metadb(1M), and metahs(1M) to administer metadevices and hot spare pools.

MetaDB object

The graphical object in DiskSuite Tool that represents the metadevice state database. The MetaDB object administers the metadevice state database and its copies (the state database replicas).

metadevice

A group of physical slices accessed as a single logical device by concatenation, striping, mirroring, setting up RAID5 metadevices, or logging physical devices. After they are created, metadevices are used like slices.

The metadevice maps logical block addresses to the correct location on one of the physical devices. The type of mapping depends on the configuration of the particular metadevice.

Also known as pseudo, or virtual device in standard UNIX terms.

Metadevice Editor window

The main window for DiskSuite Tool. It provides a view of metadevices and hot spare pools in which you can graphically create, display, or edit your configuration.

metadevice state database

A database, stored on disk, that records configuration and state of all metadevices and error conditions. This information is important to the correct operation of DiskSuite and it is replicated. See also state database replica.

metadisk driver

A UNIX pseudo device driver that controls access to metadevices, enabling them to be used like physical disk slices. The metadisk driver operates between the file system and application interfaces and the device driver interface. It interprets information from both the UFS or applications and the physical device drivers.

mirror

A metadevice made of one or more other metadevices called submirrors. It replicates data by maintaining multiple copies.

mirroring

Writing data to two or more disk drives at the same time. In DiskSuite, mirrors are logical storage objects that copy their data to other logical storage objects called submirrors.

multi-way mirror

A mirror that has at least two submirrors.

Objects list

In DiskSuite Tool, a pseudo-browser in the Metadevice Editor window that displays metadevices, hot spares, and configuration problems.

one-way mirror

A mirror that consists of only one submirror. You create a one-way submirror, for example, when mirroring slices that contain existing data. A second submirror is then attached.

online backup

A backup taken from a mirror without unmounting the entire mirror or halting the system. Only one of the mirror's submirrors is taken offline to complete the backup.

optimized mirror resync

A resync of only the submirror regions that are out of sync at a system reboot. The metadisk driver tracks submirror regions and can determine which submirror regions are out of sync after a failure. See resyncing.

Panner

In DiskSuite Tool, the region where a miniature view of the canvas shows small representations of the DiskSuite objects currently displayed on the canvas.

parity

A way for RAID5 configurations to provide data redundancy. Typically, a RAID5 configuration stores data blocks and parity blocks. In the case of a missing data block, the missing data can be regenerated using the other data blocks and the parity block.

partial mirror resync

A resync of only a replacement part of a submirror or RAID5 metadevice, rather than the entire submirror or RAID5 metadevice. See full mirror resync and optimized mirror resync.

partition

See slice.

On a SPARC system, a slice and partition are the same.

On an x86 system, a slice and partition are distinct. A partition is a part of a disk set aside for use by a particular operating system using the fdisk program. Thus partitioning the disk enables it to be shared by several different operating systems. Within a Solaris partition, you can create normal Solaris slices.

platter

The spinning disk that stores data inside a disk drive.

Put Away

A DiskSuite Tool command that returns DiskSuite objects on the Metadevice Editor window canvas to the Objects list.

RAID

Redundant Array of Inexpensive Disks. A classification of different ways to back up and store data on multiple disk drives. There are seven levels of RAID:

Level 0: Nonredundant disk array (striping)

Level 1: Mirrored disk array

Level 2: Memory-style Error Code Correction (ECC)

Level 3: Bit-Interleaved Parity

Level 4: Block-Interleaved Parity

Level 5: Block-Interleaved Distributed-Parity

Level 6: P + Q Redundancy

DiskSuite implements RAID levels 0, 1, and 5.

resync region

A division of a mirror that enables tracking changes by submirror regions rather than over the entire mirror. Dividing the mirror into resync regions can reduce resync time.

resyncing

The process of preserving identical data on mirrors or RAID5 metadevices.

Mirrors are resynced by copying data from one submirror to another after submirror failures, system crashes, or after adding a new submirror.

RAID5 metadevices are resynced during reboot if any operations that may have been halted from a system panic, a system reboot, or a failure to complete are restarted.

SCSI

Small Computer Systems Interface. An interface standard for peripheral devices and computers to communicate with each other.

sector

The smallest divisions of a disk platter's tracks. Usually 512 bytes. See block.

seek time

The time it takes for a disk drive's read/write head to find a specific track on the disk platter. Seek time does not include latency nor the time it takes for the controller to send signals to the read/write head.

shared diskset

See diskset.

simple metadevice

A term usually reserved for a concatenated metadevice, striped metadevice, or concatenated stripe metadevice.

slice

A part of each physical disk that is treated as a separate area for storage of files in a single file system, or for an application such as a database. Before you can create a file system on disk, you must partition it into slices.

Slice Filter window

In DiskSuite Tool, a menu available from the Disk View window and the Slice Browser that filters the slices to view those available to be parts of metadevices, hot spares, state database replicas, and trans metadevice logs.

state database replica

A copy of the metadevice state database. Keeping copies of the metadevice state database protects against the loss of state and configuration information critical to metadevice operations.

stripe

(1) A metadevice created by striping (also called a striped metadevice).

(2) An interlaced slice that is part of a striped metadevice.

(3) To create striped metadevices by interlacing data across slices.

striping

Creating a single logical device (metadevice) by transparently distributing logical data segments across slices. The logical data segments are called stripes.

Striping is sometimes called interlacing because the logical data segments are distributed by interleaving them across slices.

Striping is generally used to gain performance, enabling multiple controllers to access data at the same time.

Compare striping with concatenation, where data is mapped sequentially on slices.

submirror

A metadevice that is part of a mirror. See also mirror.

system (/etc/system)

A file used to set system specifications. DiskSuite uses this file, for example, when mirroring the root (/) file system.

templates

In DiskSuite Tool, the template icons create new, empty metadevices. The new metadevices cannot be used until they are "populated" with their necessary parts. Templates can also be combined to build additional metadevices.

Tbyte

(Terabyte), 1,024 Gbytes, or 1 trillion bytes (1,099,511,627,776 bytes).

three-way mirror

A mirror made of three submirrors. This configuration enables a system to tolerate a double-submirror failure. You can also do online backups with the third submirror.

trans metadevice

A metadevice for UFS logging. A trans metadevice includes one or more other metadevices or slices: a master device, containing a UFS file system, and a logging device. After they are created, trans metadevices are used like slices.

two-way mirror

A mirror made of two submirrors. This configuration enables a system to tolerate a single-submirror failure.

UFS

UNIX file system.

UFS logging

The process of recording UFS updates in a log (the logging device) before the updates are applied to the UNIX file system (the master device).