OceanStor Dorado 6.1.x SmartDedupe and SmartCompression Feature Guide For File
OceanStor Dorado 6.1.x SmartDedupe and SmartCompression Feature Guide For File
6.1.x
SmartDedupe and
SmartCompression Feature Guide
for File
Issue 07
Date 2023-07-15
and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and
the customer. All or part of the products, services and features described in this document may not be
within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements,
information, and recommendations in this document are provided "AS IS" without warranties, guarantees
or representations of any kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.
Website: https://e.huawei.com
Purpose
This document describes SmartDedupe and SmartCompression features, including
their working principles and application scenarios.
The following table lists the product models applicable to this document.
NOTICE
This document is updated periodically with the software version. The operations
described in this document use the latest version as an example. Note that the
supported functions and features vary according to the software version. The
content in this document is for reference only.
Intended Audience
This document is intended for:
Symbol Conventions
The symbols that may be found in this document are defined as follows.
Symbol Description
Change History
Changes between document issues are cumulative. The latest document issue
contains all the changes made in earlier issues.
Issue 07 (2023-07-15)
This issue is the seventh official release. The updates are as follows:
Optimized descriptions about some operations.
Issue 06 (2023-04-20)
This issue is the sixth official release. The updates are as follows:
Optimized descriptions about some operations.
Issue 05 (2022-11-15)
This issue is the fifth official release. The updates are as follows:
Optimized descriptions about some operations.
Issue 04 (2022-08-25)
This issue is the fourth official release.
Issue 03 (2022-04-15)
This issue is the third official release.
Issue 02 (2022-01-25)
This is the second official release. The updates are as follows:
Optimized descriptions about some operations.
Issue 01 (2021-09-30)
This issue is the first official release.
Contents
C Glossary................................................................................................................................... 24
D Acronyms and Abbreviations............................................................................................. 40
1 Feature Description
1.1 Overview
This section introduces the background, definition, and benefits of SmartDedupe
and SmartCompression.
Background
For any company, data is the core asset. With the demand to find value in the
massive volumes of data that rises with rapid business growth, companies will
need an ever-larger storage capacity. Fast data growth poses the following
challenges:
● Business data may overlap, causing redundant copies and unnecessary storage
device procurement expenses.
● Device maintenance and management become complex as more devices are
added, including huge expenses spent on equipment rooms, power supply,
and cooling systems.
The consequent maintenance costs of redundant data have far exceeded its value.
Therefore, storage system administrators are looking for ways to reduce redundant
data while ensuring data integrity and accuracy.
For all-flash storage systems, the amount of data that is written into SSDs and
data write count must be reduced in the service operating process due to the
erase/write cycle and service life issues of SSDs.
Definition
SmartDedupe and SmartCompression are developed by Huawei to provide state of
the art functions, including smart data deduplication and data compression.
Benefits
Table 1-1 describes the benefits of SmartDedupe and SmartCompression.
Benefit Description
Reduced initial Fewer storage devices are required to store the same
procurement costs amount of data.
Prolonged service The amount of data that is written into SSDs and the
life of SSDs data write count are reduced. Therefore, SSDs are less
worn and their service life is prolonged.
License Requirements
SmartDedupe and SmartCompression are two value-added features.
Both of them are available after you have purchased the effective capacity license
or SmartDedupe & SmartCompression software license.
Adaptive dedupe
algorithm
Similarity-based dedupe
Inline dedupe Calculate fingerprints for
similarity-based dedupe.
Compression Key technique:
Inline Similarity-based dedupe
Key techniques:
compression Opportunity Dedupe Fingerprint
1. Compression
table processing table
algorithm
Fingerprint
2. Data compaction
Data
compaction
Data
1. When a user writes data, the adaptive deduplication algorithm identifies data
suitable for inline deduplication based on data characteristics and directly
performs inline deduplication.
2. The adaptive deduplication algorithm identifies data suitable for similarity-
based deduplication based on data characteristics, calculates similar
fingerprints (SFPs), and adds the SFPs to the similarity-based deduplication
opportunity table. Then the system compresses the user data, writes the
compressed data to the storage pool, and returns a success message.
3. The background deduplication task finds similar data in the opportunity table
and reads the data from disks for similarity-based deduplication. After the
deduplication is complete, the fingerprint table is updated.
Basic Concepts
The basic concepts of SmartDedupe and SmartCompression are as follows:
● Deduplication data block size: Specifies the granularity of data that will be
deduplicated in a storage system.
● Compression data block size: Specifies the granularity of data that will be
compressed in a storage system.
● Similarity-based deduplication: The system divides data into blocks of a fixed
size and analyzes the similarity among the blocks. Then, the system
deduplicates the identical data blocks and performs combining compression
on the similar data blocks.
● Fingerprint: The fingerprint is a fixed-length binary numeric value. In a
storage system, all the mappings between data block fingerprints and data
storage locations are stored in the fingerprint table.
● Similar fingerprint (SFP): Specifies the similarities among data. If two pieces
of data have the same SFP, the contents of the two pieces of data are
partially or completely the same.
● Gradient fingerprint (GFP): A piece of data may be similar to multiple pieces
of other data, and therefore has multiple SFPs. To ensure that fingerprints
with high similarities are preferentially processed during deduplication, the
system also records the GFPs for describing data similarities when calculating
SFPs.
NOTE
SmartDedupe
Figure 1-2 shows the similarity-based deduplication process.
NOTE
In 6.1.5 and later versions, you can run the change disk_domain general
disk_domain_id=? dedup_method=? command on the CLI in developer mode to change
the deduplication mode, which can be inline (inline deduplication) or offline (similarity-
based deduplication).
Start
Start
Divide the written data
into blocks.
Check whether similar data blocks are the
same based on byte-by-byte comparison.
Calculate SFPs and GFPs If yes, perform fixed-length deduplication.
Periodically check whether Yes
of the new data blocks. If no, perform delta compression, record
SFPs exist in the opportunity fingerprint information to the fingerprint
table. table, update fingerprint information to
Write the data blocks to the metadata of the data blocks, and
disks and record the reclaim the space of these data blocks.
data blocks' fingerprint No
and location
information in the
opportunity table.
End
Step 1:
1. The storage system divides newly-written data into blocks. The Application
Request Size set on the file system is the block size.
2. The storage system uses a similar fingerprint algorithm to calculate the SFPs
and GFPs of the new data blocks.
NOTE
Only 6.1.2 and later versions support GFPs. For 6.1.0, only SFPs of the newly-written
data blocks are calculated.
3. The storage system writes the data blocks to disks and records data blocks'
fingerprint and location information in the opportunity table.
Step 2:
1. The storage system periodically checks whether SFPs exist in the opportunity
table.
– If yes, go to 2.
– If no, continue the periodic check.
2. The storage system checks whether similar data blocks are the same based on
byte-by-byte comparison.
– If yes, the storage system considers the new data block redundant and
deletes it. Then, the storage system points the fingerprint and storage
location of the new data block to that of the existing one in the
fingerprint table.
– If no, the storage system performs combining compression on the new
data block, records its fingerprint in the fingerprint table, updates the
fingerprint to the metadata of the data block, and reclaims the storage
space of the data block.
For example, file system 1, file system 2, and file system 3 in a storage system
have the same application request size and enabling status of SmartDedupe and
SmartCompression. Table 1-2 lists the existing data blocks in file system 1, file
system 2, and file system 3 as well as the results of comparison between new data
blocks J, K, and L in file system 1 and the existing data blocks.
Figure 1-3 shows the data deduplication results when SmartDedupe is disabled
and enabled respectively.
A B C D E F G H I
Data delivered A B C D E F G H I
by the host
File
system 1
File
system 2
File
system 3
... File
system 1
File
system 2
File
system 3
...
J
Data block J is the same as data block E.
J
Data is directly written to Data block K is similar to data block I.
K
K
L persistent storage media without Delta compression is performed.
L
fingerprint calculation. Data block L is written and its fingerprint
is not found in the opportunity table.
Data written J K L L
to the A B C D E F G H I A B C D E F G H I K
storage
system
File
system 1
File
system 2
File
system 3
... File File File ...
system 1 system 2 system 3
SmartCompression
The storage systems support inline compression. If the SmartCompression feature
is enabled for a file system when the file system is created, the storage system will
compress all the data written to the file system.
SmartCompression
● Databases
Databases typically have large volumes of stored data. With
SmartCompression, over 65% of storage space can be saved with only a
negligible dip in performance.
● File services
SmartCompression is also widely applied in file services.
● Engineering, natural disaster, and geological data
Similar to database backup, this data has a unified storage format but with
considerable variations between data sets. Using SmartCompression helps
save storage space.
Precautions
The application type of a file system cannot be modified once specified. You are
advised to specify the application type based on the service I/O model.
Configuration Method
Some application types are preset for typical application scenarios of file systems.
For the preset application types, the application request size and the SmartDedupe
and SmartCompression status have been configured. When creating a file system
NOTE
To specify the application type when creating a file system on the CLI, run the
create file_system general name=? pool_id=? capacity=? alloc_type=?
workload_type_id=? command. After creation, run the show file_system general
file_system_name=? command. Workload Type Name in the command output
indicates the application type of the file system. For details, visit Command/
Event/Error Code Query.
Follow-up Procedure
To query the details (including the SmartDedupe and SmartCompression status)
about an application type, run the show workload_type detail_info command in
developer mode.
Precautions
The application type of a file system cannot be modified once specified. You are
advised to specify the application type based on the service I/O model.
Configuration Method
Some application types are preset for typical application scenarios of file systems.
For the preset application types, the application request size and the SmartDedupe
and SmartCompression status have been configured. When creating a file system
on the CLI or DeviceManager, you only need to specify an appropriate application
type.
The preset application types include NAS_Default, NAS_Virtual_Machine,
NAS_Database, NAS_Large_File, Office_Automation, and NAS_EDA.
When a file system is created on DeviceManager, the default application type is
NAS_Default. You can select an application type that matches your service I/O
model by referring to the following application type details. Table 2-2 lists the
application request size and the SmartDedupe and SmartCompression status for
each application type.
NOTE
If the preset application types do not meet service requirements, run the create
workload_type general name=? io_size=? dedup_enabled=? compression_enabled=?
template_type=file_system fs_layer_distribution_algorithm=? id=? command in
developer mode to create an application type.
Follow-up Procedure
To query the details (including the SmartDedupe and SmartCompression status)
about an application type, run the show workload_type detail_info command in
developer mode.
Precautions
The application type of a file system cannot be modified once specified. You are
advised to specify the application type based on the service I/O model.
Configuration Method
Some application types are preset for typical application scenarios of file systems.
For the preset application types, the application request size and the SmartDedupe
and SmartCompression status have been configured. When creating a file system
on the CLI or DeviceManager, you only need to specify an appropriate application
type.
NOTE
NOTE
Follow-up Procedure
To query the details (including the SmartDedupe and SmartCompression status)
about an application type, run the show workload_type detail_info command in
developer mode.
NOTE
● The screenshot is for reference only and the actual displayed information may vary.
● If both the effective capacity license and the SmartDedupe & SmartCompression license
are imported to the storage system, the DeviceManager home page displays the data
reduction information as the SmartDedupe & SmartCompression license takes effect,
and does not display the effective capacity.
2. Choose System > Storage Pools. Click the name of the desired storage pool.
On the page that is displayed, view the data reduction ratio and total space
saving ratio.
Figure 3-4 Viewing detailed storage pool information (applicable to 6.1.2 and
later versions)
NOTE
● The screenshot is for reference only and the actual displayed information may vary.
● In admin mode on the CLI, run the show storage_pool general pool_id=? command
and query Data Reduction Ratio of the storage pool in the command output.
● In developer mode on the CLI, run the show storage_pool data_info command and
query the data reduction ratio of the storage pool. For details about the command, visit
Command/Event/Error Code Query.
NOTE
The screenshot is for reference only and the actual displayed information may vary.
This section provides some CLI commands for configuring and managing
SmartDedupe and SmartCompression (for file services).
NOTE
Table A-3 Commands when the SmartDedupe & SmartCompression license is used
Operation Command
C Glossary
A
AC power module The module that transfers the external AC power
supply into the power supply for internal use.
Application server A service processing node (a computer device) on the
network. Application programs of data services run
on the application server.
Asynchronous remote A kind of remote replication. When the data at the
replication primary site is updated, the data does not need to be
updated synchronously at the mirroring site to finish
the update. In this way, performance is not reduced
due to data mirroring.
Air baffle It optimizes the ventilation channels and improves
the heat dissipation capability of the system.
Audit log guarantee A mode for recording audit logs. This mode
mode preferentially ensures that the audit log function is
normal and no audit log is missing.
Audit log non- A mode for recording audit logs. In this mode,
guarantee mode services are running properly. Audit logs may be
missing.
B
Backup A collection of data stored on (usually removable)
non-volatile storage media for purposes of recovery
in case the original copy of data is lost or becomes
inaccessible; also called a backup copy. To be useful
for recovery, a backup must be made by copying the
source data image when it is in a consistent state.
The act of creating a backup.
C
Cache hit ratio The ratio of the number of cache hits to the number
of all I/Os during a read task, usually expressed as a
percentage.
Captive screw Specially designed to lock into place on a parent
board or motherboard, allowing for easy installation
and removal of attached pieces without release of
the screw.
Challenge Handshake A password-based authentication protocol that uses a
Authentication challenge to verify that a user has access rights to a
Protocol system. A hash of the supplied password with the
challenge is sent for comparison so the cleartext
password is never sent over the connection.
Compliance mode A protection mode of WORM. In compliance mode,
files within their protection period cannot be changed
or deleted by either the file user or by the system
administrator. Files with expired protection periods
can be deleted but not changed by the file user or
the system administrator.
Controller The control logic in a disk or tape that performs
command decoding and execution, host data transfer,
serialization and deserialization of data, error
detection and correction, and overall management of
device operations. The control logic in a storage
subsystem that performs command transformation
and routing, aggregation (RAID, mirroring, striping, or
other), high-level error recovery, and performance
optimization for multiple storage devices.
Controller enclosure An enclosure that accommodates controllers and
provides storage services. It is the core component of
a storage system and generally consists of
components, such as controllers, power supplies, and
fans.
Copying A pair state. The state indicates that the source LUN
data is being synchronized to the target LUN.
Container root Space used to store the metadata for running
directory container images and container instances.
Container image An image is a special file system, which provides the
programs, libraries, resources, and configuration files
required for running containers. It also contains
configuration parameters, for example, for
anonymous disks, environment variables, and users.
The image does not contain dynamic data, and its
content will not be modified after construction.
Containerized An image can start multiple containers, and an
application application can contain one or a group of containers.
D
Data compression The process of encoding data to reduce its size. Lossy
compression (i.e., compression using a technique in
which a portion of the original information is lost) is
acceptable for some forms of data (e.g., digital
images) in some applications, but for most IT
applications, lossless compression (i.e., compression
using a technique that preserves the entire content of
the original data, and from which the original data
can be reconstructed exactly) is required.
Data flow A process that involves processing data extracted
from the source system. These processes include:
filtering, integration, calculation, and summary,
finding and solving data inconsistency, and deleting
invalid data so that the processed data meets the
requirements of the destination system for the input
data.
Data migration A movement of data or information between
information systems, formats, or media. Migration is
performed for reasons such as possible decay of
storage media, obsolete hardware or software
(including obsolete data formats), changing
performance requirements, the need for cost
efficiencies etc.
Data source A system, database (database user; database
instance), or file that can make BOs persistent.
Deduplication The replacement of multiple copies of data — at
variable levels of granularity — with references to a
shared copy in order to save storage space and/or
bandwidth.
Dirty data Data that is stored temporarily on the cache and has
not been written onto disks.
E
eDevLUN Logical storage array space created by a third-party
storage array.
Expansion module A component used for expansion.
Expansion Connects a storage system to more disk enclosures
through connection cables, expanding the capacity of
the storage system.
F
Field replaceable unit A unit or component of a system that is designed to
be replaced in the field, i.e., without returning the
system to a factory or repair depot. Field replaceable
units may either be customer-replaceable or their
replacement may require trained service personnel.
Firmware Low-level software for booting and operating an
intelligent device. Firmware generally resides in read-
only memory (ROM) on the device.
Flash Translation Layer Flash Translation Layer (FTL) organizes and manages
host data, enables host data to be allocated to NAND
flash chips of SSDs in an orderly manner, maintains
the mapping relationship between logical block
addresses (LBAs) and physical block addresses
(PBAs), and implements garbage collection, wear
leveling, and bad block management.
Front-end port The port that connects the controller enclosure to the
service side and transfers service data. Front-end port
types are Fibre Channel and iSCSI.
Front-end interconnect On a storage device, all controllers share the front-
I/O module (FIM) end interface modules.
G
Garbage collection The process of reclaiming resources that are no
longer in use. Garbage collection has uses in many
aspects of computing and storage. For example, in
flash storage, background garbage collection can
improve write performance by reducing the need to
perform whole block erasures prior to a write.
Gateway A device that receives data via one protocol and
transmits it via another.
Global garbage With a view to defragmentation of storage arrays
collection and garbage collection of disks, global garbage
collection reduces garbage of disks by enabling
storage arrays to inform disks of not implementing
invalid data relocation and of controlling space
release so that disks and controllers consume less
space, reducing costs and prolonging the useful life
of storage arrays.
H
Hard disk tray The tray that bears the hard disk.
Heartbeat Heartbeat supports node communication, fault
diagnosis, and event triggering. Heartbeats are
protocols that require no acknowledgement. They are
transmitted between two devices. The device can
judge the validity status of the peer device.
Hit ratio The ratio of directly accessed I/Os from the cache to
all I/Os.
Hot swap The substitution of a replacement unit (RU) in a
system for a defective unit, where the substitution
can be performed while the system is performing its
normal functioning normally. Hot swaps are physical
operations typically performed by humans.
HyperMetro A value-added service of storage systems.
HyperMetro means two datasets (on two storage
systems) can provide storage services as one dataset
to achieve load balancing among applications and
failover without service interruption.
HyperMetro domain A HyperMetro configuration object generally; made
up of two storage arrays and one quorum server.
HyperMetro services can be created on a HyperMetro
domain.
HyperMetro vStore A HyperMetro vStore pair consists of two vStores,
pair that is, two tenants. After a HyperMetro relationship
is set up for a pair of vStores, the datasets in the two
vStores work in redundancy mode and provide
storage services in one dataset view, achieving hitless
service failover.
I
In-band management The management control information of the network
and the carrier service information of the user
network are transferred through the same logical
channel. In-band management enables users to
manage storage arrays through commands.
Management commands are sent through service
channels, such as I/O write and read channels. The
advantages of in-band management include high
speed, stable transfer, and no additional
management network ports required.
Initiator The system component that originates an I/O
command over an I/O interconnect. The endpoint
that originates a SCSI I/O command sequence. I/O
adapters, network interface cards, and intelligent I/O
interconnect control ASICs are typical initiators.
I/O Shorthand for input/output. I/O is the process of
moving data between a computer system's main
memory and an external device or interface such as a
storage device, display, printer, or network connected
to other computer systems. This encompasses
reading, or moving data into a computer system's
memory, and writing, or moving data from a
computer system's memory to another location.
L
Load balance A method of adjusting the system, application
components, and data to averagely distribute the
applied I/Os or computing requests to physical
resources of the system.
Logical unit The addressable entity within a SCSI target that
executes I/O commands.
Logical unit number The SCSI identifier of a logical unit within a target.
Industry shorthand, when phrased as "LUN", for the
logical unit indicated by the logical unit number.
LUN formatting The process of writing 0 bits in the data area of the
logical drive and generating related parity bits so that
the logical drive can be in the ready state.
LUN mapping A storage system maps LUNs to application servers
so that application servers can access storage
resources.
LUN migration A method for the LUN data to migrate between
different physical storage spaces while ensuring data
integrity and uninterrupted operation of host
services.
LUN snapshot A type of snapshot created for a LUN. This snapshot
is both readable and writable and is mainly used to
provide a snapshot LUN from point-in-time LUN
data.
Lever A lever resides on the structural part of a module. It
is used to insert or remove a module into or from a
chassis, saving efforts.
Local image repository A private repository used to store the container
images and Helm charts imported by users. It is
different from the standard image repository. The
imported images and Helm charts must meet the
compatibility requirements of the system.
M
Maintenance terminal A computer connected through a serial port or
management network port. It maintains the storage
system.
Management interface The module that integrates one or more
module management network ports.
Management network An entity that provides means to transmit and
process network management information.
Management network The network port on the controller enclosure
port connected to the maintenance terminal. It is provided
for the remote maintenance terminal. Its IP address
can be modified with the change of the customer's
environment.
N
NVM Express A host controller interface with a register interface
and command set designed for PCI Express-based
SSDs.
NVMe SSD A solid state disk (SSD) with a non-volatile memory
express (NVMe) interface. Compared with other
SSDs, such SSDs can deliver higher performance and
shorter latency.
O
Out-of-band A management mode used during out-of-band
management networking. The management and control
information of the network and the bearer service
information of the user network are transmitted
through different logical channels.
P
Power failure When an external power failure occurs, the AC PEM
protection depends on the battery for power supply. This
ensures the integrity of the dirty data in the cache.
Pre-copy When the system monitors a failing member disk in a
RAID group, the system copies the data from the disk
to a hot spare disk in advance.
Q
Quorum server A server that can provide arbitration services for
clusters or HyperMetro to prevent the resource access
conflicts of multiple application servers.
Quorum Server Mode A HyperMetro arbitration mode. When a HyperMetro
arbitration occurs, the quorum server decides which
site wins the arbitration.
R
RAID level The application of different redundancy types to a
logical drive. A RAID level improves the fault
tolerance or performance of the logical drive but
reduces the available capacity of the logical drive.
You must specify a RAID level for each logical drive.
Ransomware file When launching attacks, ransomware usually
interception generates encrypted files with special file name
extensions. In light of this, the system intercepts the
write to files with specific file name extensions to
block the extortion from known ransomware and
protect file systems in the storage system.
Real-time ransomware Ransomware has similar I/O behavior characteristics.
detection By analyzing file I/O behavior characteristics, the
system quickly filters out abnormal files and
performs deep content analysis on the abnormal files
to detect files attacked by ransomware. Then, secure
snapshots are created for file systems where files
have been attacked, and alarms are reported to
notify the data protection administrator, limiting the
impact of ransomware and reducing losses.
Reconstruction The regeneration and writing onto one or more
replacement disks of all of the user data and check
data from a failed disk in a mirrored or RAID array. In
most arrays, a rebuild can occur while applications
are accessing data on the array's virtual disks.
Redundancy The inclusion of extra components of a given type in
a system (beyond those required by the system to
carry out its function) for the purpose of enabling
continued operation in the event of a component
failure.
S
Script A parameterized list of primitive I/O interconnect
operations intended to be executed in sequence.
Often used with respect to ports, most of which are
able to execute scripts of I/O commands
autonomously (without policy processor assistance).
A sequence of instructions intended to be parsed and
carried out by a command line interpreter or other
scripting language. Perl, VBScript, JavaScript and Tcl
are all scripting languages.
Serial port An input/output location (channel) that sends and
receives data (one bit at a time) to and from the CPU
of a computer or a communications device. Serial
ports are used for serial data communication and as
interfaces for some peripheral devices, such as mouse
devices and printers.
Service data The user and/or network information required for the
normal functioning of services.
Service network port The network port that is used to store services.
Simple network An IETF protocol for monitoring and managing
management protocol systems and devices in a network. The data being
monitored and managed is defined by an MIB. The
functions supported by the protocol are the request
and retrieval of data, the setting or writing of data,
and traps that signal the occurrence of events.
Single point of failure One component or path in a system, the failure of
which would make the system inoperable.
Smart disk enclosure Being compared with traditional disk enclosures, the
smart disk enclosures are equipped with Arm chips
and DDR memories or other computing modules to
achieve powerful computing capabilities. With such
capabilities, the smart disk enclosures can help
controllers to share some computing loads,
accelerating data processing.
Share authentication During vStore configuration synchronization, the
share authentication information (including the share
information and domain controller configuration) is
synchronized to the secondary end.
T
Target The endpoint that receives a SCSI I/O command
sequence.
Target LUN The LUN on which target data resides.
Thin LUN A logic disk that can be accessed by hosts. It
dynamically allocates storage resources from the thin
pool according to the actual capacity requirements of
users.
Topology The logical layout of the components of a computer
system or network and their interconnections.
Topology deals with questions of what components
are directly connected to other components from the
standpoint of being able to communicate. It does not
deal with questions of physical location of
components or interconnecting cables. The
communication infrastructure that provides Fibre
Channel communication among a set of PN_Ports
(e.g., a Fabric, an Arbitrated Loop, or a combination
of the two).
Trim A method by which the host operating system may
inform a storage device of data blocks that are no
longer in use and can be reclaimed. Many storage
protocols support this functionality via various
names, e.g., ATA TRIM and SCSI UNMAP.
U
User interface The space where users interact with a machine.
W
Wear leveling A set of algorithms utilized by a flash controller to
distribute writes and erases across the cells in a flash
device. Cells in flash devices have a limited ability to
survive write cycles. The purpose of wear leveling is
to delay cell wear out and prolong the useful life of
the overall flash device.
Write amplification Increase in the number of write operations by the
device beyond the number of write operations
requested by hosts.
Write amplification The ratio of the number of write operations on the
factor device to the number of write operations requested
by the host.
Write back A caching technology in which the completion of a
write request is signaled as soon as the data is in the
cache. Actual writing to non-volatile media occurs at
a later time. Write back includes inherent risks: an
application will take action predicated on the write
completion signal, and a system failure before the
data is written to non-volatile media will cause
media contents to be inconsistent with that
subsequent action. For these reasons, sufficient write
back implementations include mechanisms to
preserve cache contents across system failures
(including power failures) and a flushed cache at
system restart time.
Write Once Read Many A type of storage, designed for fixed content, that
preserves what is written to it in an immutable
fashion. Optical disks are an example of WORM
storage.
Write through A caching technology in which the completion of a
write request is not signaled until data is safely
stored on non-volatile media. Write performance
equipped with the write through technology is
approximately that of a non-cached system. However,
if the written data is also held in a cache, subsequent
read performance may be dramatically improved.
Z
Zone A collection of Fibre Channel N_Ports and/or
NL_Ports (i.e., device ports) that are permitted to
communicate with each other via the fabric. Any two
N_Ports and/or NL_Ports that are not members of at
least one common zone are not permitted to
communicate via the fabric. Zone membership may
be specified by: 1) port location on a switch, (i.e.,
Domain_ID and port number); or, 2) the device's
N_Port_Name; or, 3) the device's address identifier;
or, 4) the device's Node_Name. Well-known
addresses are implicitly included in every zone.
C
CLI Command Line Interface
CPU Central Processing Unit
F
FC Fibre Channel
I
I/O Input / Output
IOPS Input/Output Operations Per Second
L
LUN Logical Unit Number
R
RAID Redundant Arrays of Independent Disks
T
TCO Total Cost of Ownership
V
VAAI vStorage APIs for Array Integration
VDI Virtual Desktop Infrastructure