3DM 2 Online Help

Table of Contents Previous Next Index



Glossary
*
3DM 2. 3ware Disk Manager. The 3ware disk manager is a web-based graphical user interface that can be used to view, maintain, and manage 3ware controllers, disks, and units. It is available for download from
http://www.3ware.com/support/download.asp.
*
3ware. Named after the 3 computer wares: hardware, software and firmware. A leading brand of high-performance, high-capacity Serial ATA (SATA) RAID storage solutions.
*
A-Chip. AccerATA chip. Automated data port to handle asynchronous ATA disk drive interface.
*
AMCC. Applied Micro Circuits Corporation provides the essential building blocks for the processing, moving and storing of information worldwide.
*
Array. One or more disk drives that appear to the operating system as a single unit. Within 3ware software, arrays are typically referred to as units.
*
Array Roaming. The process of swapping out or adding in a configured unit without having to shut down the system. This is useful if you need to move the unit to another controller.
*
Background rebuild rate. The rate at which a particular controller initializes, rebuilds, and verifies redundant units (RAID 1, RAID 5, RAID 10, RAID 50).
*
Boot volume size. The size to be assigned to volume 0 when creating a unit through 3BM on a PC machine. Note that the resulting volume does not have to be used as a boot volume. However, if the operating system is installed on the unit, it is installed in volume 0.
*
Carve size. The size over which a unit will be divided into volumes, if auto-carving is enabled.
*
Chassis Control Unit (CCU). A device within a chassis (or enclosure) used to identify a drive or display status of a RAID unit by flashing the appropriate LEDs.
*
CLI. Command Line Interface. The 3ware CLI is a text program, rather than a GUI (graphical user interface). It has the same functionality as 3DM, and can be used to view, maintain, and manage 3ware controllers, disks, and units.
*
Configuration. The RAID level set for a unit.
*
Controller. The physical card from 3ware that you insert into a computer system and connect to your disk drives or enclosure. The controller contains firmware that provides RAID functionality. 3ware makes a number of different models of SATA RAID controllers. (See System Requirements.)
*
Controller ID number. Unique number assigned to every 3ware controller in a system, starting with zero.
*
Create an array. The process of selecting individual disk drives and selecting a RAID level. The array will appear to the operating system as a single unit. Overwrites any existing unit configuration data on the drives. Note that in 3ware software tools, arrays are referred to as units.
*
DCB. Disk configuration block. This is 3ware proprietary RAID table information that is written to disk drives that are in a RAID unit, single disk, or spare. The DCB includes information on the unit type, unit members, RAID level, and other important RAID information.
*
Delete an array. Deleting an array (or unit) is the process of returning the drives in a unit to individual drives. This erases the DCB information from the drives and deletes any data that was on them. When a unit is deleted from a controller, it is sometimes referred to as being “destroyed.” If you want to remove a unit without deleting the data on it, do not delete it; instead use the Remove feature in 3DM, and then physically remove the drives.
*
Destroying. Same as deleting a unit.
*
Degraded unit. A redundant unit that contains a drive that has failed.
*
Disk roaming. When moving a unit from one controller to another, refers to putting disks back in a different order than they initially occupied, without harm to the data.
*
Distributed parity. Parity (error correction code) data is distributed across several drives in RAID 5 and RAID 50 configurations. Distributing parity data across drives provides both protection of data and good performance.
*
Drive ID. A unique identifier for a specific drive in a system. Also called a port ID.
*
Drive Number. The SCSI number, or channel number, of a particular drive.
*
ECC. Error correction code. ECC Errors are grown defects that have occurred on a drive since it was last read.
*
ECC Error policy. Determines whether an error detected during a rebuild stops the rebuild or whether the rebuild can continue in spite of the error. Specified by the Continue on Source Error During Rebuild unit policy.
*
EMS (Enclosure Management Services). Chassis-monitoring functions for environmental, power, mechanical monitoring, and control using the I²C (chassis control) bus port.
*
Export a unit. To remove the association of a unit with a controller. Does not affect the data on the drives. Used for array roaming, when you want to swap out a unit without powering down the system, and move the unit to another controller. Compare to Delete, which erases all unit configuration information from the drive.
*
Exportable unit or drive. In 3BM (BIOS), exportable units and drives are those that will be available to the operating system when you boot your computer.
*
Fault tolerant. A RAID unit which provides the ability to recover from a failed drive, either because the data is duplicated (as when drives are mirrored) or because of error checking (as in a RAID 5 unit).
*
Firmware. Computer programming instructions that are stored in a read-only memory on the controller rather than being implemented through software.
*
Grown defect. Defects that arise on a disk from daily use.
*
Hot spare. A drive that is available, online, and designated as a spare. When a drive fails in a redundant unit, causing the unit to become degraded, a hot spare can replace the failed drive automatically and the unit will be rebuilt.
*
Hot swapping. The process of removing a disk drive from the system while the power is on. Hot swapping can be used to remove units with data on them, when they are installed in hot-swap carriers. This is referred to as array roaming. Hot swapping can also be used to remove and replaced failed drives when a hot-swap carrier is used.
*
I²C-(or Inter-IC) bus. A two-wire serial bus solution used as a control, diagnostic, environmental, and power management for EMS (enclosure management services).
*
Import a unit. Attach a set of disk drives with an existing configuration to a controller and make the controller aware of the unit. Does not affect the data on the drives.
*
Initialize. For 3ware SATA RAID controllers, initialize means to put the redundant data on the drives of redundant units into a known state so that data can be recovered in the event of a disk drive failure. For RAID 1 and 10, initialization copies the data from the lower port to the higher port. For RAID 5 and 50, initialization calculates the RAID 5 parity and writes it to disk (background initialization). This is sometimes referred to as resynching, and does not erase user data. Note: If foreground initialization is done before the operating system has loaded, zeroes are written to all of the drives in the unit. This process, done through the 3ware BIOS, does erase existing data.
*
Logical Units. This term is used in the 3ware CLI. It is usually shortened to “units.” These are block devices presented to the operating system. A logical unit can be a one-tier, two-tier, or three-tier arrangement. JBOD, Spare, and Single logical units are examples of one-tier units. RAID 1 and RAID 5 are examples of two-tier units and as such will have sub-units. RAID 10 and RAID 50 are examples of three-tier units and as such will have sub-sub-units.
*
JBOD. An unconfigured single drive. The acronym is derived from “just a bunch of disks.” Note that earlier versions of the 3ware RAID controller exported JBODs to the OS. By default, this is not the case in 9000-series controllers. However, a policy in 3BM can be set that allows JBODs to be seen by the OS. By default, individual drives must be configured as Single Disks in order to be made available to the OS.
*
Migration. The process of changing the characteristics of a unit. The change can be to expand the capacity of the unit (OCE), change the stripe size of the unit, change the unit from redundant to non-redundant, or to change the unit from non-redundant to redundant.
*
Mirrored disk array (unit). A pair of drives on which the same data is written, so that each provides a backup for the other. If one drive fails, the data is preserved on the paired drive. Mirrored disk units include RAID 1 and RAID 10.
*
NCQ (Native Command Queuing). A feature designed to improve performance of SATA hard disks in some applications that require a lot of random access of data, such as server-type applications. When NCQ is enabled, the commands are reordered on the drive itself.
NCQ must be supported by the drive. NCQ must be turned on in both the drive and the RAID controller. By default, the RAID unit’s queue policy is disabled when creating a unit.
*
Non-redundant units. A disk array (unit) without fault tolerance (RAID 0, single disk, or JBOD.).
*
OCE (Online Capacity Expansion). The process of increasing the size of an existing RAID unit without having to create a new unit. See also migration.
*
Parity. Information that the controller calculates using an exclusive OR (XOR) algorithm and writes to the disk drives in RAID 5 and RAID 50 units. This data can be used with the remaining user data to recover the lost data if a disk drive fails.
*
PCB. Printed circuit board.
*
P-Chip. PCI interface chip that connects the PCI bus to the high-speed internal bus and routes all data between the two using a packet switched fabric. There is one P-chip per controller card.
*
Port. A controller has one or many ports (typically 4, 8, 12, 16). Each port can be attached to a single disk drive. On a controller with a multilane serial port connector, one connector supports four ports.
*
Port ID. A unique identifier for a specific port in a system. Also called a drive ID.
*
RAID. Redundant array of inexpensive disks, combined into a unit (array), to increase your storage system’s performance and provide fault tolerance (protection against data loss).
*
Rebuild task schedule. The specification for when rebuilding, may occur, including start time and duration.
*
Rebuild a unit. To generate data on a new drive after it is put into service to replace a failed drive in a fault tolerant unit (for example, RAID 1, 10, 5, or 50).
*
Redundancy. Duplication of data on another drive or drives, so that it is protected in the event of a drive failure.
*
Remove a drive. The process of making a drive unavailable to the controller.
*
Remove a unit. The process of making a unit unavailable to the controller and the operating system. After a unit is removed it can be hot swapped out of the system. This is sometimes referred to as exporting a unit.
*
RLM (RAID Level Migration). The process of using an existing unit of one or more drives and converting it to a new RAID type without having to delete the original unit. For example, converting a single disk to a mirrored disk or converting a RAID 0 unit to a RAID 5 unit.
*
Self-test. A test that can be performed on a scheduled basis. Available self-tests include Upgrade UDMA mode and Check SMART Thresholds.
*
Stagger time. The delay between drive groups that will spin up, at one time, on a particular controller.
*
Stripe size. The size of the data written to each disk drive in RAID unit levels that support striping. The size of stripes can be set for a given unit during configuration. In general, smaller stripe sizes are better for sequential I/O, such as video, and larger strip sizes are better for random I/O (such as databases). The stripe size is user-configurable at 64KB, 128KB, or 256KB.
This stripe size is sometimes referred as a “minor” stripe size. A major stripe size is equal to the minor stripe size times the number of disks in the unit.
*
Striping. The process of breaking up files into smaller sizes and distributing the data amongst two or more drives. Since smaller amounts of data are written to multiple disk drives simultaneously, this results in an increase in performance. Striping occurs in RAID 0, 5, 10 and 50.
*
Subunit. A logical unit of storage that is part of another unit. For example, the mirrored pairs (RAID 1) in a RAID 10 unit are subunits of the RAID 10 unit.
*
UDMA mode. UDMA mode is a protocol that supports bursting data up to 133 MB/sec with PATA disk drives and 1.5Gb/sec and 3.0 Gb/sec with SATA disk drives.
*
Unit ID. A unique identifier for a specific unit in a system.
*
Unit Number. The SCSI number, or channel number, of a particular unit.
*
Unit. A logical unit of storage, which the operating system treats as a single drive. A unit may consist of a single drive or several drives. Also known as an array.
*
Verify. A process that confirms the validity of the redundant data in a redundant unit. For a RAID 1 and RAID 10 unit, a verify will compare the data of one mirror with the other. For RAID 5 and RAID 50, a verify will calculate RAID 5 parity and compare it to what is written on the disk drive.

Table of Contents Previous Next Index
www.3ware.com
1-408-542-8800
1-800-840-6055

Copyright AMCC 2004-2006
To contact Support, log in at https://www.3ware.com