GLOSSARY

This glossary describes abbreviations, acronyms, and terms used in MSU.

A | B | C | D | E | F | G | H | I | K | L | M | N | O | P | R | S | V | W

A

Adapter

Adapter refers to any board that hosts the RAID controller ASIC.

Note: MSU executes commands on the RAID controller, which is hosted on either an HBA (Host Bus Adapter), evaluation board, or OEM motherboard.

See Controller.

AES

Acronym for Advanced Encryption Standard.

Note: This feature requires hardware and software support.

AES Link

Refers to AES encryption being applied to data at the link level (transmission from one network location to another).

Note: This feature requires hardware and software support.

Array

An Array is a combination of two or more physical disks that are presented to the operating system as a single storage device comprising of one (or more) virtual disk(s).

See Virtual Disk.

ASIC

Acronym for Application Specific Integrated Circuit.

Auto-Rebuild

 

Auto-Rebuild is disabled by default. When enabled, degraded virtual disks (and arrays) are automatically rebuilt if a spare drive of suitable size is available.

Note: The size of the spare drive must be equal to or greater than that being replaced. The Gigabyte Rounding for the virtual disk may provides some flexibility in size.

See Spare, Gigabyte Rounding, and Rebuild.

 

 

B

Background Initialization

Refers to Full Initialization running as a background activity.

See Full Initialization.

BBU

Abbreviation for Battery Backup Unit.

BGA

Abbreviation for Background Activity.

 

 

C

Consistency Check

Consistency Check checks and reports the integrity of parity/mirror data on the virtual disk.

Notes:

  • Consistency Check is recommended as a regular maintenance procedure for virtual disks.

  • Aside from regular maintenance runs, also run Consistency Check when a virtual disk is not performing as expected.

  • Perform Full Initialization on the virtual disk before running Consistency Check.

Consistency Fix

Consistency Fix checks, reports, and fixes parity/mirror data on virtual disks.

Note: Perform Full Initialization on the virtual disk before running Consistency Fix.

Controller

Refers to the RAID controller ASIC that MSU executes commands on.

Note: Although MSU and CLI use the terms Adapter and Controller interchangeably, MSU is executing commands on the RAID controller ASIC.

See Adapter.

Copy Back

 

Copy Back replaces a functioning physical disk on an operating array by copying all the contents of the functioning physical disk to a replacement physical disk, without degrading the array. This is useful when the functioning disk shows signs of impending failure.

Note: Copy Back is only available for RAID-On-Chip (ROC) controllers.

 

 

D

Dedicated Spare Drive

A dedicated spare drive is a spare that is assigned for exclusive use by a specific array.

Note: Dedicated Spare Drive is only available for RAID-On-Chip (ROC) controllers.

See Global Spare Drive, Spare.

Degraded Virtual Disk/Array

A fault-tolerant virtual disk/array becomes degraded when one or more member physical disks fail. In this state, the fault tolerance of the virtual disk/array is compromised to the extent determined by the RAID level and number of failed member physical disks. Rebuild the disk to ensure continued fault tolerance.

See Rebuild.

 

 

E

Erase Foreign Configuration

Erase Foreign Configuration erases the RAID configuration metadata on a foreign virtual disk. This is useful when using a foreign physical disk in a new virtual disk.

Note: If the foreign physical disk was originally part of another virtual disk, erasing RAID configuration data might damage that virtual disk, depending on the virtual disk’s fault tolerance capabilities.

 

 

F

Fast Initialization

In this mode, MSU erases the Master Boot Record (MBR) in all physical disks constituting the virtual disk. This is a destructive process that erases all data on the virtual disk.

Note: Fast Initialization is available only when creating a virtual disk, except for RAID-On-Chip Controllers (ROCs).

Fault Tolerance

Refers to the ability of a virtual disk/array to remain operational even when one or more member physical disks fail. The extent of fault tolerance is determined by the RAID level of the virtual disk/array.

See Redundancy.

Foreground Initialization

Foreground Initialization writes 0 to the entire virtual disk. As with Fast Initialization, all data on the virtual disk is destroyed.

Full Initialization

In this mode, mirror (or parity) data is updated to ensure consistency of data on the physical disks constituting the virtual disk. This non-destructive method of initialization is performed in the background while the operating system continues to access the virtual disk.

Note: Full Initialization is recommended as a regular maintenance procedure for virtual disks.

 

 

G

Gigabyte Rounding

Gigabyte Rounding is applicable to RAID levels using the disk mirroring concept (RAID 1 and RAID 10). It defines the flexibility in size for spare drives used to rebuild degraded virtual disks (and arrays).

Available options for Gigabyte Rounding are None, 1G, and 10G.

Note: For example, 120.5 GB and 115.7 GB physical disks can create a 115.7 GB RAID 1 virtual disk. If Gigabyte Rounding is set to 1G, then the controller floors the size of the physical disks to the nearest lower GB size. The 120.5 GB and 115.7 GB disks are treated as 120 GB and 115 GB physical disks, resulting in a 115 GB RAID 1 virtual disk. If the 115.7 GB physical disk fails, the virtual disk can be rebuilt with a spare drive sized between 115 and 115.7 GB, providing flexibility in size up to 0.7 GB.

Global Spare Drive

A global spare drive is a spare that is available for use by all virtual disks (and arrays).

See Dedicated Spare Drive, Spare.

 

 

H

Hot Spare

Controllers supporting this feature allow MSU to automatically replace a failed or offline physical disk with a spare drive connected to the controller. The automatic replacement is performed while the virtual disk (or array) is still operating.

See Auto-Rebuild.

Hot Swap

Controllers supporting this feature allow users to manually hot-swap (replace) a failed or offline physical disk with a spare drive at the same location while the virtual disk (or array) is still operating.

 

 

I

Import

Controllers supporting this feature allow you to import a virtual disk created with another RAID controller.

Initialization

See Fast Initialization and Full Initialization.

Initialization Rate

The controller assigns priority to Background Initialization based on the Initialization Rate. Default is 50% (Medium).

Note: When background activities are in progress, the controller is still available to the OS for normal operations. However, the response time may be lower depending on the background activity control rate.

IO Controller (IOC)

IO Controllers (IOCs) such as the Marvell 88SE63xx/88SE64xx SAS controllers can create and manage RAID 0, 1, 5, and 10 virtual disks.

Note: Marvell Storage controllers support different sets of RAID levels depending on the hardware model and OEM software package. Some software packages, depending on OEM selections, support limited RAID levels by design. Check with the OEM vendor for information specific to your controller.

IO Processor (IOP)

An IOP is similar to an IOC, except that it contains an on-board CPU that enables hardware RAID functionality. The Marvell 88SE91xx 6 Gbps SATA RAID IOP can create and manage RAID 0 and RAID 1 virtual disks.

Note: Marvell Storage controllers support different sets of RAID levels depending on the hardware model and OEM software package. Some software packages, depending on OEM selections, support limited RAID levels by design. Check with the OEM vendor for information specific to your controller.

 

 

K

Key-phrase

Refers to the alpha-numeric string used to generate the AES key for data encryption.

Note: MSU uses an algorithm to combine the key-phrase with internal 128/256-bit AES keys to generate the AES key used for encryption.

 

 

L

Locate

Locate turns the LED on a physical disk ON or OFF. This is useful when identifying a physical disk in an enclosure or backplane.

Note: The enclosure or backplane must support the Locate feature.

 

 

M

Media Patrol

Media Patrol checks and fixes media errors on physical disks.

For a configured physical disk, Media Patrol checks and fixes media errors immediately. For spare drives, Media Patrol logs all media errors and fixes them when the spare drive is used to rebuild a virtual disk/array.

Media Patrol Rate

The controller assigns priority to Media Patrol based on the Media Patrol Rate. Default is 100% (High).

Note: When background activities are in progress, the controller is still available to the OS for normal operations. However, the response time may be lower depending on the background activity control rate.

Metadata

Metadata, stored by MSU on physical disks, contains RAID configuration data about the virtual disk and/or array.

Migrate

Migrate allows you to increase the disk capacity and change the RAID level of an operating RAID array

Module Consolidate

Enable Module Consolidate to consolidate all requests sent from the OS. Enabling this feature can improve sequential read/write performance. Default is On (Enable).

 

 

N

No Initialization

No initialization is performed on the virtual disk.

No Read Ahead

Read Ahead Cache is disabled.

See Read Ahead.

 

 

O

Online Capacity Expansion

Refers to disk capacity expansion of an operating RAID array.

Note: The RAID level does not change. See Migrate.

 

 

P

PD

Abbreviation for Physical Disk.

Poll S.M.A.R.T. Status

Enable Poll S.M.A.R.T. Status to periodically check S.M.A.R.T. status of physical disks. Default is Off (Disable).

If MSU’s request for S.M.A.R.T status fails or if the physical disk is no longer in S.M.A.R.T status, MSU logs the event in Event Logs.

 

 

R

RAID

RAID is an acronym for Redundant Array of Independent Disks. For more information, see Selecting a RAID Level.

RAID-On-Chip Controller (ROC)

RAID-On-Chip (ROC) controllers such as the Marvell 88RC8180 can create and manage RAID 0, 1, 1E, 5, 10, 50, and 60 arrays. ROCs contain an on-board CPU which enables hardware RAID functionality.

Note: Marvell Storage controllers support different sets of RAID levels depending on the hardware model and OEM software package. Some software packages, depending on OEM selections, support limited RAID levels by design. Check with the OEM vendor for information specific to your controller.

Read Ahead

In this read cache mode, the controller’s read performance is configured for sequential access of data. It reads extra blocks from the physical disk to cache memory, in anticipation of the next read command from the application.

Note: For applications requiring random access of data, enabling Read Ahead might degrade read performance due to the overhead of advance read operations.

Rebuild

Rebuild is the process of reconstructing data on a degraded virtual disk/array. Data is reconstructed on the replacement physical disk(s) using parity or mirror data on the virtual disk.

Note: Virtual disks/arrays can be rebuilt only as long as they are fault tolerant. All data is lost on the virtual disk/array if the number of failed member physical disks is greater than its fault tolerance capabilities.

Rebuild Rate

The controller assigns priority to Rebuild based on the Rebuild Rate. Default is 50% (Medium).

Note: When background activities are in progress, the controller is still available to the OS for normal operations. However, the response time may be lower depending on the background activity control rate.

Redundancy

Refers to the concept of duplicating data either through mirroring or storing parity information, in order to enable rebuilding of virtual disks/arrays in the event of the failure of one or more member physical disks.

See Fault Tolerance.

Relearn

The Relearn procedure stabilizes/optimizes BBU performance. During Relearn, the BBU is discharged completely and then re-charged to full capacity. If MSU detects that BBU capacity/voltage is low and/or unstable, it automatically initiates the Relearn procedure. Relearn may also be scheduled as a maintenance task for the BBU to ensure stable/optimal performance.

 

 

S

Set Free

Set Free releases an offline physical disk from a virtual disk.

For information on offline physical disks, see Icons in MSU.

Spare

Refers to physical disk assigned as spare drive for use in rebuilding  virtual disks/arrays.

See Dedicated Spare Drive, Global Spare DriveHot Spare, and Hot Swap.

Stripe Size

Stripe Size specifies the size of single data block on the virtual disk.

Available stripe sizes are 16K, 32K, 64K (default), 128K, 256K, 512K,  and 1024K. For RAID 5 and 50, the stripe size is limited to 64K.  Support for different stripe sizes varies with controllers.

Note: The larger the stripe size, the longer it takes for the controller to read  from and write to data blocks on the physical disks. For best read/write performance, a larger stripe size is recommended for applications requiring large data transfers such as audio, video, and graphics. A smaller stripe size enhances read/write performance for applications with content much smaller in size, such as e-mails, and documents.

Synchronization Rate

 

The controller assigns priority to Consistency Check and Consistency  Fix based on the Synchronization Rate. Default is 50% (Medium).

Note: When background activities are in progress, the controller is still  available to the OS for normal operations. However, the response time may  be lower depending on the background activity control rate.

 

 

V

VD

Abbreviation for Virtual Disk.

Virtual Disk

A virtual disk is a combination of two or more physical disks that are  presented to the operating system as a single storage device.

See Array.

 

 

W

Write-Back,

Performance 

In this write cache mode, the controller is configured for best write  performance. Write data is stored temporarily in DDR cache memory  and flushed to the hard disk at the appropriate time. In the event of  power failure, there is a risk of losing data integrity if the transfer of  data from the DDR cache to hard disk is not completed.

Note: A BBU (Battery Backup Unit) is recommended when using Write-Back  cache mode. When a BBU is available and optimally charged, MSU  automatically turns on Write-Back (Performance) cache.

Write-Through, Reliable 

This is the default write cache mode in MSU. In this mode, all write  operations are performed on the hard disk before a completion status  is returned to the operating system.