Microsoft, its logo, Windows, Windows Server and MS-DOS are worldwide registered trademarks of
Microsoft Corporation of the U.S.A.
Promise Technology, Inc., its logo, FastTrak, FastBuild Utility, and Web-based Promise Array
Manager (WebPAM) are registered trademarks of Promise Technology Inc. of the U.S.A.
NEC ESMPRO and NEC EXPRESSBUILDER are registered trademarks of NEC Corporation.
All company names and product names mentioned herein are trademarks or registered trademarks
(1) No part of this manual may be reproduced in any form without the prior written permission of
NEC Corporation.
(2) The contents of this manual may be revised without prior notice.
(3) The contents of this manual shall not be copied or altered without the prior written permission of
NEC Corporation.
(4) All efforts have been made to ensure the accuracy of all information in this manual. If you
notice any part unclear, incorrect, or omitted in this manual, contact the sales agent where you
purchased this product.
(5) NEC assumes no liability arising from the use of this product, nor any liability for incidental or
consequential damages arising from the use of this manual regardless of Item (4).
(6) If you find any missing pages or pages out of order in this manual, please contact your dealer
なお、本書は和英併記となっております。日本語での説明は i ページから 52 ページを、英
語での説明は i ページから xvi ページおよび、53 ページから 106 ページを参照してくださ
い。
Congratulations for your purchase of the Disk Array Controller.
The User’s Guide describes how to install and use the N8103-89 Disk Array Controller
(SATA) correctly and safely. Read the guide thoroughly before handling it. In addition, refer
to this manual when you want to know how to use it or some malfunction occurs. Always
keep the manual at hand so that you can see it as soon as possible if necessary.
For the server in which the disk array controller is installed, refer to the User’s Guide of the
server. Read "Notes on Use" carefully before handling the disk array controller.
This User's Guide is written in both Japanese and English. For Japanese, refer to pages i
to 52. For English, refer to pages i to xvi and 53 to 106.
ii
このユーザーズガイドは、必要なときすぐに参照できるよう、お手元に置いておくようにしてください。
「使用上のご注意」を必ずお読みください。
Keep this User's Guide at hand for quick reference at anytime necessary.
Be sure to read this section carefully.
使用上のご注意~必ずお読みください~
NOTES ON USE
本製品を安全に正しくご使用になるために必要な情報が記載されています。
The following includes information necessary for proper and safe operation of the product.
This section provides notes on using your product safely. Read this section carefully to
ensure proper and safe use of the product. For symbols, see "SAFETY INDICATIONS"
provided earlier.
<全般的な注意事項>
General
人命に関わる業務や高度な信頼性を必要とする業務には使用しない
Do not use the product in life-critical applications or applications requiring
high reliability.
The product is not intended for integration with or control of facilities or equipment
that may affect human life or that require a high degree of reliability, such as
medical equipment, nuclear power facilities or instruments, aerospace instruments,
transportation facilities or instruments. NEC does not assume any liability for
accidents resulting in injury or death, or for any damages to property that may occur
as a result of using the product in such facilities, equipment, or control systems.
煙や異臭・異音がしたまま使用しない
Do not use the server if any smoke, odor, or noise is present.
If smoke, odor, or noise is present, immediately turn off the server and disconnect
the power plug from the AC outlet, then contact your service representative. Using
the server in such conditions may cause a fire.
針金や金属片を差し込まない
Keep needles or metal objects away from the server.
Do not let any form of liquid (water etc.) or foreign matter (e.g., pins or paper clips)
enter the server. Failure to follow this warning may cause an electric shock, a fire, or
a failure of the server. When such things accidentally enter the server, immediately
turn off the power and disconnect the power plug from the AC outlet. Do not
disassemble the server. Contact your service representative.
<電源・電源コードに関する注意事項>
Power Supply and Power Cord Use
vii
電源がONのまま取り付け・取り外しをしない
Disconnect the power cord(s) before installing or removing the product
in/from the server.
Make sure to power off the server and disconnect the power cord(s) from a power
outlet before installing/removing the product in/from the server, or connecting with
the peripheral devices. All voltage is removed only when the power cords are
unplugged.
Make sure the cable condition before connection. Using the damaged connector,
bent connector pin, or dirty connector may cause a fire due to short-circuit.
ぬれた手で電源コードをもたない
Do not hold the power plug with a wet hand.
When disconnecting the power cord from the server, hold the plug and pull it straight
out. Pulling the cord out by the cable portion could damage the cable to result in an
electrical shock hazard or a fire.
CAUTION
viii
<設置・移動・保管・接続に関する注意事項>
Installation, Relocation, Storage, and Connection
CAUTION
プラグを差し込んだままインタフェースケーブルの取り付けや取り外しをしない
Do not connect any interface cable with the power cord of the server plugged
to a power source.
Make sure to power off the server and unplug the power cord from a power outlet
before connecting/disconnecting any interface cable to/from the server. If the
server is off-powered but its power cord is plugged to a power source, touching a
cable or connector may cause an electric shock or a fire resulted from a short
circuit.
指定以外のインタフェースケーブルを使用しない
Do not use any unauthorized interface cable.
Use only interface cables authorized by NEC and locate a proper device and
connector before connecting a cable. Using an unauthorized cable or connecting a
cable to an improper destination may cause a short circuit, resulting in a fire.
Also, observe the following notes on using and connecting an interface cable.
Do not step on the cable.
Do not place any object on the cable.
Do not use the server with loose cable connections.
Do not use any damaged cable connector.
Make sure the cable is securely locked with screw.
ix
腐食性ガスの存在する環境で使用または保管しない
Do not use or store the product in the place where corrosive gases exist.
Make sure not to locate or use the server in the place where corrosive gases (sulfur
dioxide, hydrogen sulfide, nitrogen dioxide, chlorine, ammonia, ozone, etc) exist.
Also, do not install it in the environment where the air (or dust) includes components
accelerating corrosion (ex. sulfur, sodium chloride) or conductive metals. There is a
risk of a fire due to corrosion and shorts of an internal printed board.
Consult with your service representative for the location appropriate to the server.
高温注意
Avoid installation in extreme temperature conditions.
Immediately after the server is powered off, its internal components such as hard
disk drives are very hot. Leave the server until its internal components fully cool
down before installing/removing any component.
CAUTION
x
<お手入れに関する注意事項>
Cleaning and Working wit
h the Product
自分で分解・修理・改造はしない
Do not disassemble, repair, or alter the server.
Never attempt to disassemble, repair, or alter the product on any occasion. Failure
to follow this instruction may cause an electric shock or fire as well as malfunctions
of the product.
プラグを差し込んだまま取り扱わない
Disconnect the power plug before accessing inside the server.
Make sure to power off the server and disconnect the power plug from a AC outlet
before accessing inside the server. Touching any internal device of the server with
its power cord connected to a power source may cause an electric shock even if the
server is off-powered.
Always connect the DC cable and/or interface cable firmly. An incompletely
connected cable may cause a contact failure, resulting in smoking or fire.
CAUTION
<運用中の注意事項>
During Operation
xi
CAUTION
雷がなったら触らない
Avoid contact with the server during thunderstorms.
雷が鳴りだしたら、本製品内蔵の本体装置には、触れないでください。感電するお
それがあります。
Disconnect the power plug from the outlet when a thunderstorm is approaching.
If it starts thundering before you disconnect the power plug, do not touch any part of
the server containing the product. Failure to follow this warning may cause an
electric shock.
The guide is intended for persons who are familiar with operating systems including
Windows and fundamental operations of general-purpose I/O devices including the
keyboard and mouse.
The carton contains various accessories, as well as the product itself. See the packing list
to make sure that you have everything and that individual components are not damaged. If
you find any component missing or damaged, contact your sales agent.
Be sure to take appropriate measures not to leak important data (e.g., customers'
information or companies' management information) on the removed hard disk to
any third parties.
Data seems to be erased when you empty "Recycle Bin" of Windows or execute
the "format" command of the operating system. However, the actual data remains
written on the hard disk. Data not erased completely may be restored by special
software and used for unexpected purposes.
It is strongly recommended that the software or service (both available at stores)
for data erasure should be used in order to avoid the trouble explained above. For
details on data erasure, ask your sales representative.
NEC assumes no liability for data leakage if the product is transferred to third party
without erasing the data.
To transfer or sell any software application that comes with the product to a third party, the
following requirements must be satisfied:
All provided software applications must be transferred and no backup copies must be
retained.
Software applications must be uninstalled before transferring the product.
It is the user's responsibility to completely erase or modify all the data
stored in storage device such as hard disk, backup data cartridge, floppy
disk, or any other media (CD-R/CD-RW) so that the data cannot be
restored.
The device failure due to shock or thermal changes, as well as operator's misconduct, may
cause loss of data. To avoid loss of data, NEC recommends that you should make a
back-up copy of your valuable data on a regular basis.
Fast Initialization を実施する場合は、再度<Ctrl>+<Y> キーを押します。Fast
Initialization を実施しない場合は、その他の任意のキー(<Esc>キーなど)を押します。
Fast Initialization Option has been selected.
It will erase the MBR data of the disks.
<Press Ctrl-Y Key if you are sure to erase it>
<Press any other key to ignore this option>
Read this chapter first if you use the disk array controller for the first time.
This chapter describes the notes you should always follow while you use the card, the
features of the card, and the hardware setup.
1. Notes on Use - Always Follow These Notes -
Follow the following notes to allow you to use the card safely.
1-1. Installation of WebPAM
Install the Web-based Promise Array Manager (called WebPAM hereafter), management
utility which manages the card on OS. The installation of WebPAM allows you to perform
the following operations:
Events and errors occurred on array system can be registered in the event log
and used effectively for troubleshooting and diagnosis.
WebPAM event information can be monitored by using NEC ESMPRO.
Manual rebuild, media patrol, and/or synchronization can be executed and
scheduled.
For the installation of WebPAM, refer to the online manual "N8103-89 Disk Array Controller
(SATA) Supplementary Setup Procedure" saved in the "S-ATA Array Management
Software" CD-ROM coming with the card.
1-2. Preventive Maintenance by Media
Patrol/Synchronization
Routine media patrol or synchronization is recommended as the preventive maintenance
against subsequent defects of hard disk drives (called HDDs hereafter). These features
allow subsequent defects of HDDs to be found and repaired as soon as possible. Both the
features can be done routinely by using the WebPAM scheduling feature.
For the detailed features of the media patrol and synchronization, see "Chapter 3 Features
of Disk Array Controller"
The recommended scheduling interval is once per week. Depending on the operation
status of your system, the scheduling interval should be at least once per month.
To utilize the media patrol or synchronization, WebPAM must be installed.
Check
56
2. Specification
Item Specification Remarks
Number of SATA connectors 4 internal channels SATA Signal: 7pin
Cache capacity 64 MB
PCI bus Conforming to PCI2.3
PCI connector Universal/32 bits 66 MHz
Maximum PCI bus transfer rate 266 MB/sec
Device interface
Maximum data transfer rate 150 MB/sec
RAID level 0, 1, 5, or 10
Maximum number of cards installed in
server
Maximum number of connectable HDDs4
Maximum number of logical drives 8
Outer
dimension
Weight About 0.09kg
Operating voltage 5V
Power consumption (max.) 5.3W 3.3V/1.044A
Operating environment Temperature: 10ºC to 35ºC
Full-height PCI
In use of bracket
Low-profile PCI
In use of bracket
Serial ATA Generation1
available
1
121 (width) x 180(depth) x
22 (height) mm
81 (width) x 180 (depth) x
22 (height) mm
Humidity: 20% to 80%
1 HDD connected per
channel
3.3V/0.458A(SDRAM)
1.8V/0.174A
Without condensation
57
3. Features of Disk Array Controller
The disk array controller is equipped with four channels of interface connectors conforming
to Serial-ATA Generation1. The data transfer rate per channel is up to 150 MB/sec. The
card realizes low cost and high performance.
The card can be connected to an additional battery to operate in the WriteBack mode,
which can improve the access performance further. In addition, the card can be connected
to an additional HDD cage to provide the hot-swap feature.
Features of disk array controller
Data transfer rate of up to 150 MB/sec
Installation of 64MB SDRAM
Up to 4 SATA HDDs connectable per board (1 HDD connectable per channel)
Support of RAID levels 0, 1, 5, and 10
Operation in WriteBack mode enabled by connection of additional battery
Report monitoring with use of NEC ESMPRO enabled by installation of WebPAM
Automatic detection of fault drive
Replacement of defected HDD enabled without system shutdown (hot-swap) (in
connection of additional HDD cage used exclusively)
Warning by beep available
Low profile correspondence
The card does not support the PCI hot plug feature.
Notice
Tips
The card can be connected with the following additional HDD cages:
Channel 1 - 4 (Port 1 - 4)
These channels allow the card to be connected to SATA devices.
Channel 2
Channel 4
59
2
3
4
5
6
7
Channel 1
Channel 3
HW label
Indicates the management revision of the card.
REV
XXX
Additional battery connector
The connector allows the card to be connected to an additional battery.
I2C connector
The connector is connected with an I2C cable.
N code label
Indicates the N code of the card.
N8103-89
HDD LED connector
The connector allows the card to be connected to the motherboard in the server in order
to illuminate the DISC ACCESS lamp on the server.
PCI connector
The connector allows the card to be connected to a PCI slot in the server.
60
5. Hardware Setup
Install the disk array controller in a server in the following procedure.
Before the installation, always refer to the User’s Guide of the server. The
job flow varies depending on the server type, system configuration, and
existence of additional HDD cage. Check the server type and device
Check
Start setup
Prepare for setup.
Select and install bracket.
Connect LED and SATA cables.
Install additional HDD cage.(*)
Form SATA cable.
Connect I2C cable.(*)
End setup.
(*) These steps are not required if the additional HDD cage is not installed.
configuration before the installation to follow the correct flow.
Install the disk array controller
(with connection of additional battery).
5-1. Prepare for setup
Note the following before the setup.
Only a single disk array controller can be installed in a server.
Some limitation may be imposed to the installation on the PCI slot
Notice
1. Exit from all applications and shut down OS.
2. Press the POWER switch on the server to turn off the power of the server.
depending on the type of the server. Before the installation, check the
limitation following the User’s Guide of the server.
HDDs to be connected to the card should have the same specification.
Contact your service representative for HDDs which can be connected to
the card.
The card may coexist with other PCI boards (including disk array
controller, mirroring board, and SCSI controller) only limitedly. Before
using the card together with other PCI boards, ask your service
representative whether the card can coexist with the other PCI boards.
61
3. Pull out all the power cords connected to the power unit of the server from the
receptacles.
4. Remove the side cover on the server following the procedure described in the
User’s Guide of the server.
For the installation or removal of the side cover and other components on
the server, refer to the User’s Guide of the server.
Check
62
5-2. Selecting and Installing Bracket
The card is previously equipped with a full-height PCI bracket. To install the card on a PCI
slot suit to the low profile, the full-height PCI bracket should be replaced with the
low-profile PCI bracket coming with the card.
1. Remove the screws (2) fixing the full-height PCI bracket to the card.
2. Remove the full-height PCI bracket.
3. Install the low-profile PCI bracket on the card.
4. Fix the low-profile PCI bracket with the screws (2) removed in step 1.
Full-height PCI bracket
Low-profile PCI bracket
Take the similar procedure to replace the low-profile PCI bracket with the
full-height PCI bracket.
Tips
5-3. Installing the Disk Array Controller
1. Confirm the position of the PCI slot in which the disk array controller is installed
and remove the corresponding additional slot cover.
Save the removed additional slot cover carefully. The removed screw will
be used to install the card. Do not lose them.
The card does not support the PCI hot-plug feature. Before install or
Notice
Check
2. Insert the card into the PCI slot securely and fix it. To fix the card with screws, use
the screw removed when the additional slot cover is removed.
remove the card from the server, always turn off the server and pull out
the power cord from the receptacle.
Some limitation may be imposed to the installation on the PCI slot
depending on the type of the server. Before the installation, check the
limitation following the User’s Guide of the server.
63
This card (disk array controller)
Screw
3. To connect an additional battery to the card, install the battery in the same
procedure.
When the card cannot be inserted into the PCI slot well, pull out it once
and insert it again. Note that the cared may be damaged if excess force
is given to it.
Notice
For details of how to install the additional battery, refer to the User’s
Guide of the N8103-93 Battery Backup Unit (SATA).
64
5-4. Connecting LED and SATA Cables
Connect the LED cable coming with the card to the HDD LED connector. For the
connection, see the figure below and the connection table. For the connection to the
motherboard, refer to the User’s Guide of the server. If it is hard to connect the LED cable,
pull out the card from the PCI slot once and connect the cable to the card.
Next, connect a SATA cable to the SATA connector on the card. The channel number of the
SATA cable should be the same as the channel number (port number) of the card.
LED cable
SATA cables
Port4
Port3
1 2 3 4
HDD LED connectors
Connect the black
cable to this pin.
Connect the red cable to this
pin. (This is not used for a
single-line LED cable.)
LED cable connection table
HDD LED
connector
Notice
Pin 1 Connected with red cable Not used
Pin 2 Connected with black cable Connected with black cable
Pin 3 Not used Not used
Pin 4 Not used Not used
Depending on the type of the server, the LED cable has two lines or only
a single line.
For the connection of the LED cable to the motherboard, refer to the
User’s Guide of the server.
To use the N8154-09F SATA2 HDD Cage, connect the SATA cable
coming with the cage to the card. For the detailed connection, refer to
the User’s Guide of the additional HDD cage.
LED cable (coming with server)
2-line cable 1-line cable
65
5-5. Installing Additional HDD Cage
To install an additional HDD cage in the server, see the procedure described in the User’s
Guide coming with the additional HDD cage.
5-6. Connecting I2C Cable
To use the additional HDD cage, connect the card to the cage through the I2C cable
coming with the card. Connect the 3-pin connector (white) of the I2C cable to the I2C
connector on the card and the 4-pin connector (black) to the additional HDD cage. For the
mating connector on the additional HDD card, refer to the User’s Guide of the additional
HDD cage.
Without use of the additional HDD cage, the I2C cable may not be connected.
I2C connector
3-pin connector (white)
To additional HDD cage
I2C cable
The N8154-01F S-ATA HDD Cage also contains an I2C cable in its
package. However, it is unavailable for the card. Always use the I2C cable
coming with the card.
Notice
4-pin connector (black)
66
5-7. Forming Cables
To fix the SATA and I2C cables appropriately, form the cables in the following procedure.
1. Attachment of cable clamp
Attach the cable clamp coming with the card on the server. Attach the cable clamp
at an arbitrary position where cables can be fixed securely. Some servers do no
have any area available for the attachment of the cable clamp. If so, bundle
cables so that they may not be disordered.
Cable clamp
Bottom plate of
chassis of server
Before attaching the cable clamp on the server, wipe the area where the
cable clamp will be attached with dry cloth.
Some servers require their accessories to be used instead of the cable
Check
clamp coming with the card. For details, refer to the User’s Guides of the
servers.
2. Cable forming
Release the lock of the cable clamp. Bundle the cables to fix them.
SATA cable
67
Cable clamp
I2C cable
After forming the SATA cable, make sure that the cable is not loose and
inserted to the mating connector straight.
Notice
68
Chapter 2 RAID
This chapter describes the RAID features which the disk array controller supports.
1. Overview of RAID
1-1. What is RAID (Redundant Array of Inexpensive
Disks)?
RAID is an abbreviation for "Redundant Array of Inexpensive Disks." The RAID technology
allows more than one hard disk drive (HDD) to be handled collectively.
In actual, RAID can configure more than one HDD as a single array (logical drive) to
operate the HDDs effectively. This can bring higher performance than a single HDD of a
large capacity.
For the array (logical drive) configuration, the host computer recognizes more than one
HDD as if it were a single HDD. The host computer accesses to more than one HDD
configuring an array (logical drive) in parallel.
Some RAID levels can recover data from remaining data and parity by using rebuild if an
error occurs in an HDD. This can provide high reliability for the system.
69
1-2. RAID Levels
The record mode enabling the RAID feature includes several levels. Among the levels, the
disk array controller supports the following levels; RAID 0, RAID 1, RAID 5, and RAID10.
The number of HDDs required to create a logical drive varies depending on the RAID level
as shown in the table below.
RAID level
RAID 0 1 4
RAID 1 2 2
RAID 5 3 4
RAID 10 4 4
Number of required HDDs
Min. Max.
For details of the RAID levels, see "2. RAID Levels" described later in this
chapter.
Tips
70
1-3. Logical Drive
A logical drive is configured with a group of more than one HDD. It is recognized as a
physical drive by OS. Up to eight logical drives are permitted by the card when four HDDs
are installed in the server.
The figure below shows a sample configuration in which the disk array controller is
connected with four HDDs, three of which create a logical drive and the remaining HDD is
set to a single hot-spare disk.
Disk array controller
Logical drive 1
Channel 1
(Port 1)
LD 1-1
Channel 2
(Port 2)
LD 1-2
Channel 3
(Port 3)
LD 1-3
Channel 4
(Port 4)
Spare
LD x-x : HDDs assigned to
logical drive
Spare :Hot-spare disk
1-4. Parity
The parity means redundant data. A single set of redundant data is created from the data
saved in more than one HDD.
The created redundant data is used for data recovery when a HDD is defected.
1-5. Hot-Swap
The hot-swap enables a HDD to be removed (replaced) under system operation.
For the hot-swap with the card, an additional HDD cage must be connected
to the card.
Check
71
1-6. Hot-Spare Disk
The hot-spare disk is prepared as an auxiliary HDD substituting for a defected HDD
included in a logical drive which is configured at a redundant RAID level. Detecting a HDD
fault, the system disconnects the HDD (or makes it offline) and starts rebuild using the
hot-spare disk. The hot-spare disk can be set in two types as listed in the table below.
Setting Feature
Global Spare
Dedicated Spare
To define a hot-spare disk, WebPAM, the management utility of the card,
must be installed.
Available as a hot-spare disk if a HDD in any logical drive is
defected. The disk of the type can accept media patrol.
Available as a hot-spare disk only if a HDD in the specified logical
disk is defected. The disk of the type can accept media patrol.
Check
Tips
For standby rebuild, or rebuild using hot-spare disks, see "Chapter 3
Features of Disk Array Controller."
"Free" disk
A "free" disk (unconfigured disk not assigned to a logical drive and hot-spare disk) can
substitute for defected HDD in the similar way as a global spare. In the configuration
including both hot-spare and "free" disks, the hot-spare disks are used preferentially.
However, "free" disks cannot be subject to media patrol. Thus it is recommended to define
"free" disks as hot-spare disks.
72
2. RAID Levels
This section details the RAID levels which the disk array controller can support.
2-1. Characteristics of RAID Levels
The table below lists the characteristics of the RAID levels.
Level Function Redundancy Characteristics
RAID0 Striping No Data read/write at the highest rate
Largest capacity
Capacity: (capacity of single HDD) ×
(number of HDDs)
RAID1 Mirroring Yes Two HDDs required
Capacity: capacity of single HDD
RAID5
RAID10
Striping of both data
and redundant data
Combination of striping
and mirroring
Yes Three or more HDDs required
Capacity: (capacity of single HDD) ×
((number of HDDs) - 1)
Yes Four HDDs required
Capacity: (capacity of single HDD) × 2
2-2. RAID0
In RAID 0, data to be recorded is distributed to HDDs. The mode is called "striping."
In the figure below, data is recorded in stripe 1 (disk1), stripe 2 (disk 2), and stripe 3 (disk
3)… in the order. Because RAID0 allows all HDDs to be accessed collectively, it can
provide the best disk access performance.
RAID 0 cannot have data redundancy. If a HDD is defected, the data saved
in the HDD cannot be recovered.
Notice
Disk array controller
Disk 1Disk 2
Stripe 1
Stripe 4
Stripe 2
Stripe 5
Disk 3
Stri pe 3
Stri pe 6
73
2-3. RAID1
In the RAID1 level, data saved in a HDD is recorded to another HDD without change. The
mode is called "mirroring."
When data is recorded to a single HDD, the same data is recorded to another HDD. If
either of the HDDs is defected, the other HDD containing the same data can substitute for
the defected HDD. Thus the system can continue to operate without interruption.
Disk array controller
Disk 1Disk 2
Stri pe 1
Stri pe 2
Stripe 1
Stripe 2
2-4. RAID5
In RAID5, data is distributed to HDDs by striping and, at the same time, the parity
(redundant data) is distributed to the HDDs. This mode is called "striping with distributed
parity."
Each of stripe x, stripe x+1, and parity (x, x+1) created from stripe x and stripe x+1 is
recorded to a specific HDD. Accordingly, the total capacity assigned to the parity is just the
same as the capacity of a single HDD. If any one of the HDDs configuring a logical drive is
defected, data is still available without problems.
Disk array controller
Disk 1Disk 2Disk 3
Stripe 1
Stripe 4
Parity (5, 6)
Stripe 2
Parity (3, 4)
Stripe 5
Parity (1, 2)
Stripe 3
Stripe 6
74
2-5. RAID10
Data is distributed to HDDs by striping and the stripes are recorded by mirroring. Therefore,
RAID10 can realize the disk access performance as high as RAID0 and the reliability as
high as RAID1 at a time.
Disk array controller
Disk 1Disk 2
Stripe 1
Stripe 3
Stripe 5Stripe 6Str ipe 5
RAID0RAID0
Stripe 2
Stripe 4
RAID1
Disk 3
Stripe 1
Stripe 3
Disk 4
Stri pe 2
Stri pe 4
Stri pe 6
75
Chapter 3 Features of Disk Array
Controller
This chapter describes the features of the disk array controller.
1. Rebuild
If a HDD is defected, the rebuild feature can recover the data in the defected HDD. The
rebuild can be applied to redundant logical drives in the RAID1, RID5, or RAID10 level.
1-1. Manual Rebuild
The manual rebuild can be done by using WebPAM, the management utility of the disk
array controller. Select a HDD and start the rebuild manually.
For the detailed operation, refer to the online manual "Web-based Promise Array Manager
User’s Guide" saved in the "S-ATA Array Management Software" CD-ROM coming with the
card.
1-2. Auto Rebuild
The disk array controller can automatically start the rebuild without use of any utility such
as WebPAM.
The auto rebuild includes two types as follows:
Standby rebuild
Automatic rebuild by using hot-spare disks. In the configuration including
hot-spare disks, the rebuild is done automatically if a HDD assigned to a logical
drive is defected.
Hot-swap rebuild
Automatic rebuild by hot-swapping defected HDD.
76
Notice
Note the following for the rebuild:
The HDD to be used for rebuild should have the same capacity, number
of revolutions, and standard as the defected HDD.
During rebuild, the processing rate is decreased due to much load.
During rebuild, do not shut down or reboot the server. If the server is
shut down by an unforeseen accident such as power interruption, turn on
the power again as soon as possible. The rebuild is automatically
restarted.
For the hot-swap rebuild, an additional HDD cage is required.
The interval from the removal of the defected HDD to the installation of a
substitute HDD should be 90 sec or longer.
If the hot-swap rebuild does not operate, perform the manual rebuild.
77
2. Media Patrol
The media patrol gives the read & verify test in the entire area of HDDs. It can be done for
all HDDs assigned to logical drives and hot-spare disks.
The media patrol allows subsequent defects of HDDs to be detected and repaired.
Accordingly, routine media patrol is recommended as preventive maintenance. Routine
media patrol can be done by scheduling.
For HDDs configuring redundant logical drives in the RAID1, RAID5, or RAID10 level or
those assigned to hot-spare disks, error sectors detected during media patrol can be
repaired. For HDDs configuring non-redundant logical drives in the RAID0 level, error
sectors are registered to the bad sector list (BSL) to be managed.
If an access to a HDD being subject to media patrol occurs, the media patrol is interrupted
temporarily and continued from the interrupted point at the completion of the access. Thus,
the media patrol can hardly decrease the system performance.
Note the following for the media patrol:
In an environment of frequent HDD accesses, media patrol cannot be
advanced well. In such environment, it is recommended to use
Notice
synchronization for preventive maintenance instead of media patrol.
To enable media patrol, WebPAM must be installed.
For the detailed operation, refer to the online manual "Web-based
Promise Array Manager User’s Guide" saved in the "S-ATA Array
Management Software" CD-ROM coming with the card.
78
3. Synchronization
The synchronization checks consistency among logical drives. It is available for redundant
logical drives in the RAID1, RAID5, or RAID10 level.
Similar to the media patrol, routine synchronization can be done by scheduling.
The synchronization can not only check consistency but also repair detected error sectors
in the same way as the media patrol. Accordingly, it can be used as preventive
maintenance.
Different from media patrol, the synchronization continues its operation at a certain rate if
targeting logical drives are accessed. Therefore, considerable load may be applied to the
system during synchronization to decrease in the processing rate. In an environment of
frequent accesses, preventive maintenance using synchronization is recommended
because media patrol cannot be advanced.
Note the following for the synchronization:
To perform synchronization, WebPAM must be installed.
For the detailed operation, refer to the online manual "Web-based
Notice
Promise Array Manager User’s Guide" saved in the "S-ATA Array
Management Software" CD-ROM coming with the card.
79
4. Expansion
The expansion adds one or more HDDs to an existing logical drive to increase the capacity
of the logical drive.
Note the following for the expansion:
Before expansion, data backup and synchronization should always be
done.
Notice
During expansion, the processing rate is decreased due to much load.
Expansion is unavailable for a logical drive in the critical state. Before
the expansion, rebuild should be done to recover logical drive.
During expansion, do not shut down or reboot the server. If the server is
shut down by an unforeseen accident such as power interruption, turn on
the power again as soon as possible. After rebooting the system, the
expansion is automatically started.
For the expansion, WebPAM must be installed.
For the detailed operation, refer to the online manual "Web-based
Promise Array Manager User’s Guide" saved in the "S-ATA Array
Management Software" CD-ROM coming with the card.
80
Ex: Expansion of logical drive in RAID5
The figure below shows an example of adding a single 80GB HDD to a RAID5 logical drive
configured with three 80GB HDDs.
Logical drive(RAID5
80GB80GB80GB
Logical drive(RAID5
80GB80GB80GB
)
[Before execution]
Capacity = 160GB
80GB
Expand
)
[After execution]
Capacity = 240GB
80GB
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.