A Storage overview. 45

Index. 51

Figures

Multiple physical drives (D1, D2, and D3) in a system. 45 Multiple physical drives (D1, D2, and D3) configured into one logical drive (L1). 45 Data striping (S1-S4) and data blocks (B1-B12) on multiple physical drives (D1, D2, and D3). 45 Two arrays (A1, A2) containing five logical drives (L1 through L5) spread cross five physical drives (D1 through D5). 46 RAID 1 array, with two physical hard drives (D1, D2). 47 RAID 1+0 array, with eight physical hard drives (D1 through D8). 47 RAID 5 array, with three physical hard drives (D1, D2, D3) showing distributed parity information (Px,y). 48 RAID 6 array, with four physical hard drives (D1, D2, D3, D4) showing distributed parity information (Px,y)(Qx,y). 49

Tables

Document conventions. CLI special keys. RAID 0 features. RAID 1, RAID 1+0 features. RAID 5 features. RAID 6 features. Summary of RAID methods. Choosing a RAID level.
.7. 11. 47. 48. 48. 49. 50. 50

About this guide

This user guide provides information to help you use the MSA Command Line Interface (CLI).

Intended audience

This book is intended for use by administrators with a moderate amount of SAN-management experience.

Related documentation

In addition to this guide, see the following related documents: MSA1500 compatibility guide HP StorageWorks Modular Smart Array 1500 installation and configuration overview (printed poster)
HP StorageWorks 1500 Modular Smart Array installation guide HP StorageWorks 1500 Modular Smart Array maintenance and service guide
MSA1000 compatibility guide HP StorageWorks Modular Smart Array 1000 installation and configuration overview (printed poster)
HP StorageWorks 1000 Modular Smart Array installation guide HP StorageWorks 1000 Modular Smart Array maintenance and service guide
HP Array Configuration Utility user guide
Configuring Arrays on HP Smart Array Controllers reference guide
These and other HP documents can be found on the HP documents website: http://www.docs.hp.com.
Document conventions and symbols
Table 1 Document conventions Element
Cross-reference links and e-mail addresses Website addresses Italics font Monospace font Key names Text typed into a GUI element, such as into a box GUI elements that are clicked or selected, such as menu and list items, buttons, and check boxes

Convention

Medium blue text: Figure 1 Medium blue, underlined text (http://www.hp.com) Bold font
Text emphasis File and directory names System output Code Text typed at the command-line Code variables Command line variables

Left arrow Right arrow Up arrow Down arrow

Keyboard keys

Ctrl+B Ctrl+F Ctrl+P Ctrl+N

Meaning

Move the cursor back one character Move the cursor forward one character Recall the previous command in the command buffer Recall the next command in the command buffer

NOTE: 0.

The CLI uses a zero-based numbering system. For example, LUN number assignments begin with
CLI commands in redundant configurations
When an MSA has two controllers, the same firmware image is run on both controllers. The two controllers communicate with each other through a PCI bus called the inter-controller link (ICL). Each controller has a serial port and has a CLI available to the users who connect the serial port to a serial terminal. Some CLI commands are entered from one controllers CLI prompt but are executed from the other controller. Command syntax instructs the controller to accept user input, pass the command to the other controller, and then display the result. The following keywords are used in the CLI to indicate a specific controller: this_controlleris included in command syntax to refer to the controller that the CLI is connected to. other_controlleris included in command syntax to refer to the other controller in the MSA.

CLI command syntax

As previously mentioned, commands are entered at the CLI prompt. Commands are not case sensitive and must be typed out in full. CLI command strings include the basic command plus specific command options, some of which are mandatory and some of which are optional. The CLI does not support line-continuation characters. If all characters of a command do not fit on one line of the CLI, let them wrap around to the next line on the screen. The maximum command length is 255 characters.

Example command

add unit 0 data=disk101-disk103 raid_level=0
This example command has three portionsthe basic command, plus two required command options.

Basic command

add unit <#> data=diskrange raid_level=r
The basic command includes a word or phrase used to instruct the controller. Commands usually contain a verb with a noun. Every CLI command must begin with a basic command.

Command options

#the number to assign to the LUN. data=diskrangethe number of an individual drive or range of drives to incorporate into the LUN. Disks are identified by box number and bay number. For example, disk101 through disk103 identifies disks 1 through 3 in box number 1. raid_level=rthe RAID fault-tolerance level to use, where r represents: = = = = RAID RAID RAID RAID (no fault tolerance) (mirroring) (distributed parity) (advanced data guarding (ADG))

An option is defined as words or phrases listed after the basic command that supply necessary information to support the command. If an option is required but not entered, the CLI command string is considered invalid. If an option is available but not required, a default value is used.
Overview of CLI storage configuration procedures
When using the CLI to initially configure the MSA controller and its storage, use the following sequence: 1. Create the LUNs. See LUN-related commands on page 25 for command descriptions. NOTE: OpenVMS environments must assign a unique ID number to each LUN. See Assigning a name or ID to a LUN on page 28 for instructions. 2. Enter array controller settings, including global parameters. See Global commands on page 35 for command descriptions. NOTE: OpenVMS environments must assign a unique ID number to each array controller. See Setting the controller ID on page 36 for instructions. 3. Enter connection information about the HBAs with access to the MSA. See Host connection commands on page 40 for command descriptions. NOTE: Each host accessing the storage must identify its operating system (profile type.) See Changing the profile of a connection on page 41 for instructions. 4. (Optional) Limit access to the storage. See Access Control List commands on page 43 for command descriptions.

Accessing the CLI

The CLI is accessed through a host computer connected to the serial port of an MSA controller. Use the following steps to establish a serial connection to the controller: 1. Make sure there is an MSA controller installed in Slot 1 of the unit (Slot 1 is located on the front right of the unit). In multipathing configurations, two MSA controllers must be installed. 2. Connect the MSA controller to a host using the custom serial cable included in the shipping carton of the MSA. An additional or replacement custom serial cable can be ordered using part number 259992-001. 3. Set up a terminal emulator. Depending on the host operating system, different emulators are available. For example, Linux uses Minicom and Microsoft Windows uses HyperTerminal. Setup procedures for emulators differthe following instructions are for setting up HyperTerminal: a. Click Start > All Programs > Accessories > Communications > HyperTerminal to open HyperTerminal.

size=syy(optional) how much of the available space on the indicated drives is to be used for the LUN, where s represents the LUN size, and yy indicates MB or GB. (When GB is entered, all return displays will be converted to MB.) If no size is specified, the maximum available space of the included disks, up to 2 TB, will be assigned to the unit. spare=diskrange(optional) disk(s) to assign as a spare to the unit, where diskrange represents the disk number. More than one disk can be assigned as a spare to a LUN. And, the same drive can be assigned as an available spare to several different LUNs. maxboot=enable|disable(optional) changes the size of the boot partition. Enable = (default) 8-GB boot partition Disable = 4-GB boot partition cache=enable|disable(optional) determines whether to use the array controller cache for the LUN. Enable = (default) use the array controller cache Disable = do not use the array controller cache
IMPORTANT: Maintain a record of the units as they are created. These unit numbers are used in other CLI commands. In addition to recording the unit number, the drives included, RAID type and size, record the order in which they are created.
Example command and response to create a single LUN from a group of drives
CLI> add unit 0 data=disk101-disk107 disk110 raid_level=6 stripe_size=64
First volume to be configured on these drives. Logical Unit size = 69460 MB RAID overhead = 0 MB Total space occupied by new unit = 69460 MB Free space left on this volume = 0 MB Unit 0 is created successfully.
Example command and response to create multiple LUNs on a group of drives
CLI> add unit 1 data=disk111-disk114 raid_level=5 stripe_size=32 size=1000mb
First volume to be configured on these drives. The logical unit size has been adjusted by 4MB for optimal performance. Logical Unit size = 996 MB RAID overhead = 498 MB Total space occupied by new unit = 1494 MB Free space left on this volume = 24533 MB Unit 1 is created successfully.
CLI> add unit 2 data=disk111-disk114 raid_level=5 stripe_size=32 size=2000mb Logical Unit size = 2000 MB RAID overhead = 1000 MB Total space occupied by new unit = 3000 MB Free space left on this volume = 21533 MB Unit 2 is created successfully. CLI> add unit 3 data=disk111-disk114 raid_level=5 stripe_size=16 size=4000mb Logical Unit size = 4000 MB RAID overhead = 2000 MB Total space occupied by new unit = 6000 MB Free space left on this volume = 15533 MB Unit 3 is created successfully.

#the number of the LUN to modify. raid_level=rthe RAID level to migrate to. RAID levels include: = = = = RAID RAID RAID RAID (no fault tolerance) (mirroring) (distributed parity) (advanced data guarding (ADG)) use use use use stripe stripe stripe stripe sizes sizes sizes sizes 8, 8, 8, 8, 16, 16, 16, 16, 32, 32, 32, 32, 64, 128, and 256 (Default: 128 KB) 64, 128, and 256 (Default: 128 KB) and 64 (Default: 16KB) and 64 (Default: 16 KB)
stripe_size=s(optional) the stripe size to use. Options include: RAID RAID RAID RAID can can can can
CLI> migrate unit 0 raid_level=5 stripe_size=32
The RAID level of Unit 0 will now be 5. Unit 0 is being migrated. Use show unit 0 to monitor progress.
Changing the cache setting for a LUN
To enable or disable the array accelerator cache for a specific LUN:
set unit <#> <cache=[enable|disable]>
#the number of the LUN to modify. In this example, LUN 0 will be modified. cache=enable|disableeither enables or disables the use of the array accelerator cache for the specified LUN.
CLI> set unit 0 cache=enable Cache for unit 0 has been enabled.
Setting the preferred path for a LUN
NOTE: This command is available only with active/active firmware (v6.x or later).
Configures an explicit preferred controller path for specified logical units. Host requests that are directed to the non-preferred controller will be delayed slightly, due to data and command transfer through the inter-controller link. If implicit path changes are enabled (see Enabling/disabling automatic path switching on page 34), controller firmware may automatically switch controller paths for unit(s) based on host request load.
set preferred_path [controller] [reset] <unit_list>
controllerindicate the controller by entering this_controller, other_controller, controller=1, or controller=2. resetrestore the last explicitly configured path for the specified LUNs. unit_listindicate the LUNs to set by entering all, the unit number of a specific LUN, or a range of unit numbers.
CLI> set preferred_path this_controller 0-9
SetPreferredPath start. change_map=0x2FF, lun_to_ctlr_map=0x3FF POST_Do_Msg: 67 CACHE HARDWARE TEMPORARILY DISABLED Quiescing host I/O for LUN map 0xFFEFF. Host/proxy I/O quiesced. Quiescing host I/O for LUN map 0xFFF00100. Host/proxy I/O quiesced. Quiescing cache flush task. Quiescing background I/O. Background quiesced. Flipping ownership(s). change_map=0x2FF, lun_to_ctlr_map=0x3FF POST_Do_Msg: 41 REDUNDANCY ACTIVE ACTIVE CONTROLLER Preferred path for unit 0 is now controller 1 (this_controller) Preferred path for unit 1 is now controller 1 (this_controller) Preferred path for unit 2 is now controller 1 (this_controller) Preferred path for unit 3 is now controller 1 (this_controller) Preferred path for unit 4 is now controller 1 (this_controller) Preferred path for unit 5 is now controller 1 (this_controller) Preferred path for unit 6 is now controller 1 (this_controller) Preferred path for unit 7 is now controller 1 (this_controller) Preferred path for unit 9 is now controller 1 (this_controller) Resuming I/O. I/O Resumed. POST_Do_Msg: 65 CACHE HARDWARE ENABLED

Enabling/disabling automatic path switching
If implicit path changes are enabled, controller firmware will automatically determine and, as needed, switch controller paths for units based on host request load, according to the specified preferences.
set auto_path_change <enable|disable> [prp=#] [mnr=#] [twp=#] [twl=#]
enable|disableenables or disables implicit path switching. prpspecify a proxied percentage (# = 51-100). This is the minimum percentage of proxied requests required to trigger implicit path change. mnrspecify the minimum number of requests that must be issued to the LUN before path changes are enabled (the number is rounded up to the next highest power of two). twpspecify the minimum number of minutes (0-255) to wait at power-up before changing LUN ownership. twlspecify the minimum number of minutes (0-255) to wait before changing LUN ownership again for same unit.
CLI> set auto_path_change disable Implicit (automatic load-based) LUN ownership changes are currently DISABLED.
CLI> set auto_path_change enable Implicit (automatic load-based) LUN ownership changes are currently ENABLED. Implicit LUN ownership change preferences are as follows: Minimum % of proxied requests to trigger implicit path change.: Minimum number of requests to LUN before path change enabled.: Minimum time to wait before changing ownership after power-up.: Minimum time between ownership changes for LUN..: 75% minutes 30 minutes
CLI> set auto_path_change twp=15 Implicit (automatic load-based) LUN ownership changes are currently ENABLED. Implicit LUN ownership change preferences are as follows: Minimum % of proxied requests to trigger implicit path change.: Minimum number of requests to LUN before path change enabled.: Minimum time to wait before changing ownership after power-up.: Minimum time between ownership changes for LUN..: 75% minutes 30 minutes

Global commands

The CLI provides the ability to configure the MSA controller, enter global settings, and perform system-wide commands, including: Setting global variables Setting the controller ID Setting hard addressing Changing the CLI prompt Updating MSA firmware Disabling a controller for removal Overriding the ACU lock Enabling or disabling debug console logging

set this_controller_id <identifier>
-orset other_controller_id <identifier>
identifierthe user-defined ID for the controller. The ID can be up to 230 alphanumeric characters or a decimal number in the range of 0 though 65535. If spaces are included in the name, quotation marks are required.
CLI> set this_controller_id MSA-1 Controller identifier MSA-1 created.

Setting hard addressing

To enable or disable Fibre Channel Arbitrated Loop hard addressing on an MSA controller:
set this_controller_hard_address <enable|disable> <loop_id>
enable|disableenables or disables hard addressing on the controller. loop_ida decimal number between 0 and 125 to assign the loop identifier.
CLI> set this_controller_hard_address enable 1 Hard Addressing is enabled. Loop ID = 1, ALPA = 0xE8

Changing the CLI prompt

To temporarily change the CLI prompt from the default of CLI>:
set prompt <new_prompt>
new_promptthe user-defined name for the prompt (up to 24 alphanumeric characters).
CLI> set prompt MSA-1 MSA-1>

Updating MSA firmware

Periodically, HP releases updated versions of MSA controller firmware. Updates may include: additional features and functions, performance enhancements, and fixes to known issues. In addition to MSA1000 and MSA1500 components, this command can also be used to update components on attached MSA20 storage enclosures. NOTE: Because this updating method uses an out-of-band serial connection to the MSA, it is substantially slower than in-band Fibre Channel updating methods. If possible, HP recommends using the Fibre-Channel updating method for your operating system. Operating system
Windows ProLiant host Linux ProLiant host HP-UX host
In-band utility (Fibre Channel)
MSA Flash Utility MSA Flash Utility SCSI Command Utility (SCU) Array Configuration Utility CLI (ACU-CLI)
Out-of-band utility (Serial)

MSA CLI MSA CLI MSA CLI

OpenVMS host Tru64 UNIX host
MSA_Util SCSI CAM Utility (SCU)
MSA CLI not tested MSA CLI not supported
Before using this method to update the firmware, make note of the following requirements:
A full backup of all data on the MSA array being updated is required. A copy of the array configuration, including all the information used to create the units, is needed for the update. If it has been more than six months since you restarted your MSA storage system, HP recommends that you power-cycle the MSA (power off, and then power on) before updating the firmware to ensure that you are working with a fresh system. The MSA must be power-cycled after updating the firmware to activate the new firmware. When performing an online update for active/passive configurations, ensure that the front-right controller is the active one. The firmware update may not work properly if the front-left controller is active during the update, especially if the update is performed via serial cable. Existing firmware on MSA1000 controllers must be v4.32 or higher. Existing firmware on MSA1500 controllers must be v4.94 or higher. Host computer must support the 1k Xmodem (Ymodem) protocol. Because firmware updates require a restart of the MSA and its attached storage enclosures, perform MSA controller firmware updates only during a scheduled maintenance window.

For newly installed MSA, do not perform a firmware update until the controller batteries are fully charged. For existing MSA, do not perform a firmware update until you have confirmed that the host mode or profile for each connection is correctly set. The host mode identifies the operating system of each HBA connection to the storage. Do not use the default setting. If the host mode is not properly set, hosts may lose access to the storage or experience other difficulties after the update. Use the show connections and set connection commands to view and change the profile. After performing an update, confirm that the host mode or profile for each connection is still set correctly. Use the show connections and set connection commands to view and change the profile. HP-UX environments must have an MSA Fibre link to a host.
IMPORTANT: If you encounter any problems during the firmware update process, stop and contact HP technical support. See HP technical support on page 9 for support contact information. 1. Schedule a maintenance window for the update. 2. Obtain the firmware files (.bin option) and save it to a temporary location on the host. Firmware updates are provided on the Software, firmware & drivers page of MSA websites: MSA1000: http://www.hp.com/go/msa1000 MSA1500: http://www.hp.com/go/msa1500cs 3. Perform a full backup of the data on the array. 4. If the MSA has not been restarted in the last six months, power-cycle the array. This ensures that you are working with a fresh system before beginning the firmware update. 5. Open an emulator session to access the CLI. 6. Enter the following command:
download firmware offline
This command performs an update and does not automatically restart the MSA. Through this updating method, the MSA controller and fan modules can be updated. This method is supported for use in both single-controller and dual-controller configurations. 7. On the emulator window menu bar, select Transfer > Send File. a. Click Browse, navigate to the directory in which you placed the firmware files, and then select the firmware file. b. Expand the Protocol drop-down box, and then select 1K Xmodem. c. Click Send. A 1K Xmodem status window is displayed. NOTE: During the updating process, the following messages are displayed on the MSA controller LCD panel: FIRMWARE FLASH STARTED FIRMWARE FLASH DONE 8. Wait for a completion message to be displayed. 9. Restart the MSA by pressing the power/standby button on the front of the MSA. Newly downloaded firmware cannot be accessed until the MSA is restarted.
NOTE: In dual-controller configurations, firmware on the two controllers is compared each time the MSA chassis is restarted. If the versions are mismatched, the system prompts to clone the firmware on the controller with the latest version over to the controller with the earlier version firmware. The following message is displayed on the LCD panel of the controller with the earlier firmware: 07 CLONE FIRMWARE ? < = NO, > = YES Press the > button on the LCD panel to clone the firmware. During the cloning process, informational messages are displayed on the controller LCD panels. When the cloning process is complete, the just-updated controller automatically restarts. 10. Before resuming I/O access to the MSA, verify that the MSA and its storage arrays are online.

ConnectionNamethe user-defined name to give the connection, up to 16 alphanumeric characters. wwpn=xxxxxxxx-xxxxxxxxthe WWPN of the active HBA inside the host attached to the MSA. The show connections command can be used to obtain the WWPN. profile=profile_namethe platform of the host. Options include: Defaultis not operating-system specific and should not be used; use a valid pre-defined name. Windows OpenVMS Tru64 Linux Solaris Netware HP offset=offset_value(default: 0) the unit offset for assigning logical volumes.
CLI> add connection MSA-1 wwpn=12345678-12345678 profile=windows Connection MSA-1 has been added successfully. Profile Windows is set for the new connection.
Changing the profile of a connection
To change the operating-system profile associated with an existing connection (or to change the HBA of a connection): NOTE: The connection profile is sometimes referred to as the Host Mode.
set connection <ConnectionName> [wwpn=xxxxxxxx-xxxxxxxx] [profile=profile_name] [offset=offset_value]
ConnectionNamethe name of the connection to modify. wwpn=xxxxxxxx-xxxxxxxxthe WWPN of the connection to modify, using the WWPN of the HBA. The show connections can be used to obtain the WWPN. profile=profile_namethe platform of the host. Options include: Defaultis not operating-system specific and should not be used; use a valid pre-defined name. Windows OpenVMS Tru64 Linux Solaris Netware HP offset=offset_value(default: 0) the unit offset for assigning logical volumes.
CLI> set connection MSA-1 profile=windows
The profile of connection MSA-1 is set to Windows successfully.
Changing the user-defined name of a connection
To change the user-defined name associated with a connection:
rename connection <old_connection_name> <new_connection_name>
old_connection_namethe name of the connection that is to be changed. new_connection_namethe new name to assign to the connection, up to 16 alphanumeric characters.
CLI> rename connection abc xyz
Connection(s) has been renamed successfully.
Deleting the user-defined name of a connection
To remove the user-defined name associated with a connection to a host:
delete connection <connection_name>
connection_namethe nickname that was assigned to the HBA within the host.
CLI> delete connection MSA-1
Connection(s) has been deleted successfully.

Creating a new profile

This command creates a new host profile.
add profile <profileName>
profileNamethe name of a new profile.
Moving to a different profile
This command copies the profile settings of a valid, pre-defined profile into a different profile. NOTE: If you are migrating from active/passive to active/active firmware and did not set your profile (it is still set to default), you can use this command to copy the settings for a valid profile (such as Windows) into the default profile. However, HP recommends using the set connection command to properly set the host connection. For details, see Changing the profile of a connection on page 41.

S1 S2 S3 S4

B1 B4 B7 B10 D1

B2 B5 B8 B11 D2

B3 B6 B9 B12 D3 15312
Figure 3 Data striping (S1-S4) and data blocks (B1-B12) on multiple physical drives (D1, D2, and D3)
For data in the logical drive to be readable, the data block sequence must be the same in every stripe. This sequencing process is performed by the array controller, which sends the data blocks to the drive write heads in the correct order. A natural consequence of the striping process is that each physical drive in a given logical drive will contain the same amount of data. If one physical drive has a larger capacity than other physical drives in the same logical drive, the extra capacity is wasted, because it cannot be used by the logical drive. The group of physical drives containing the logical drive is called a drive array, or array. Because all physical drives in an array are commonly configured into just one logical drive, the term array is often used as a synonym for logical drive. However, an array can contain several logical drives, each of a different size (Figure 4).
L3 L1 L2 D1 D2 D3 D4 DL4 L5
Figure 4 Two arrays (A1, A2) containing five logical drives (L1 through L5) spread across five physical drives (D1 through D5) Each logical drive in an array is distributed across all of the physical drives within the array. A logical drive can also extend across more than one storage enclosure attached to the array system. Drive failure, although rare, is potentially catastrophic. For example, in Figure 4, failure of any one physical drive in an array causes every logical drive in the array to suffer irretrievable data loss. To protect against data loss due to physical drive failure, logical drives are usually configured with fault tolerance.

Fault-tolerance methods

To protect against data loss due to physical drive failure, logical drives are usually configured with fault tolerance. The following configuration types are available: RAID 0Data striping only (no fault tolerance) RAID 1+0Drive mirroring RAID 5Distributed data guarding RAID 6 (ADG)Advanced data guarding For any configuration except RAID 0, further protection against data loss can be achieved by assigning a drive as an online spare. This drive contains no data and is connected to the same controller as the array. When any other physical drive in the array fails, the controller automatically rebuilds information data protection. (In the unlikely event that another drive in the array fails while data is being rewritten to the spare, the logical drive will still fail.) A spare is assigned to an array and is automatically assigned to all logical drives in the same array. You do not need to assign a separate spare to each array; you can configure one hard drive to be the spare for several arrays.

RAID 0No fault tolerance

A RAID 0 configuration (Figure 3) provides no protection against data loss when a drive fails. However, it is useful for rapid storage of large amounts of non-critical data (for printing or image editing, for example), or when cost is the most important consideration. Table 3 RAID 0 features Disadvantages All data on the logical drive is lost if a physical drive fails. Cannot use an online spare. Can only preserve data by backing it up to external drives. Advantages Highest write performance of all RAID methods. Lowest cost per unit of stored data of all RAID methods. All drive capacity is used to store data (none is needed for fault tolerance).

RAID 1+0Drive mirroring

In a RAID 1+0 configuration, data on a physical hard drive is duplicated to a second drive. NOTE: When only two hard drives are included in the array, this fault-tolerance method is called RAID 1. When more than two hard drives are included in the array, this fault-tolerance method is called RAID 1+0. RAID 1 is not supported on the MSA1510i storage system.

B1 B2 B3 B4 D1

B1 B2 B3 B4 D2 15314
Figure 5 RAID 1 array, with two physical hard drives (D1 and D2) When the array has more than two physical drives, drives are mirrored in pairs (Figure 6).
Figure 6 RAID 1+0 array, with eight physical hard drives (D1 through D8)
In each mirrored pair, the physical drive that is not busy answering other requests answers any read requests that are sent to the array. (This behavior is called load balancing.) If a physical drive fails, the remaining drive in the mirrored pair can still provide all the necessary data. Several drives in the array can fail without incurring data loss, as long as no two failed drives belong to the same mirrored pair. RAID 1+0 is useful when high performance and data protection are more important than the cost of physical drives. Table 4 RAID 1, RAID 1+0 features Disadvantages Expensive (half of the drives are used for fault tolerance). Only half of total drive capacity usable for data storage. Advantages Highest read and write performance of any fault-tolerant configuration. No loss of data as long as no failed drive is mirrored to another failed drive.
RAID 5Distributed data guarding
In this method, a block of parity data is calculated for each stripe from the data that is in all other blocks within that stripe. The blocks of parity data are distributed across every physical drive within the logical drive (Figure 7). When a physical drive fails, data that was on the failed drive can be calculated from the data on the remaining drives and the parity data. This recovered data is written to the assigned spare or to a replacement drive in a process called a rebuild.

Mirroring 50% n/Yes Only if no two failed drives are in a mirrored pair High Medium High

RAID 5

Distributed Data Guarding (DDG) 67% to 93% (n-1)/n 3 Yes No

RAID 6

Advanced Data Guarding (ADG) 50% to 96% (n-2)/n 4 Yes Yes

High High Low

High Low Medium
*Values for usable drive space are calculated with these assumptions:
All physical drives in the array have the same capacity. Online spares are not used. No more than 14 physical drives are used per array for RAID 5. No more than 56 drives are used with RAID 6.

Choosing a RAID level

Use Table 8 to help you determine the best RAID level for your environment. Table 8 Choosing a RAID level Also important
Cost effectiveness I/O performance Cost effectiveness Fault tolerance I/O performance I/O performance Cost effectiveness Fault tolerance
Most important characteristic

Fault tolerance

Suggested RAID level
RAID 6 RAID 1+0 RAID 6* RAID 5 (RAID 0 if fault tolerance is not required) RAID 5 (RAID 0 if fault tolerance is not required) RAID 1+0
Access Control Lists See ACL ACL adding to with the CLI 44 commands 43 deleting from, with the CLI 44 viewing, with the CLI 43 ADD ACL command 44 ADD CONNECTION command 40 ADD PROFILE command 42 ADD UNIT command 26 array controller configuration, with the CLI 35 arrays and logical drives 45 audience 7 authorized reseller, HP 9 prompt, changing of 37 serial cable requirements 14 serial connection, setup 14 Server Connection Commands 40 setup 14 special keystrokes 11 command line interface, see CLI commands ADD ACL 44 ADD CONNECTION 40 ADD PROFILE 42 ADD UNIT 26 CHANGE MODE 43 COPY PROFILE 42 DELETE ACL 44 DELETE CONNECTION 42 DOWNLOAD FIRMWARE 37 EXPAND UNIT 30 EXTEND UNIT 31 LOCATE 25 OVERRIDE ACULOCK 39 RENAME CONNECTION 41 SET CONNECTION 41 SET DEBUG 39 SET GLOBALS 35 SET OTHER_CONTROLLER 36 SET PROMPT 37 SET THIS_CONTROLLER 36 SET UNIT 32 SET UNIT_ID 28 SHOW ACL 43 connections adding, with the CLI 40 changing profiles of, with the CLI 41 changing the name of, with the CLI 41 deleting names of, with the CLI 42 viewing, with the CLI 40, 43 conventions document 7 text symbols 8 COPY PROFILE command 42

Replacing a fan module. Before you begin. Verifying component failure. Removing the component. Installing the component. Verifying proper operation. Replacing the SCSI I/O module. Before you begin. Verifying component failure. Removing the component. Installing the component. Verifying proper operation. Replacing the power button module Before you begin. Verifying component failure. Removing the component. Installing the component. Verifying proper operation. Replacing the MSA1000 chassis.
7 Enclosure and hard drive migrations.
Adding a new storage enclosure. Migrating existing storage enclosures. Moving hard drives or arrays.
8 Capacity expansion and extension. 9 Hard drive failures and faulted LUNs.
Recognizing hard drive failure. Effects of hard drive failure. Compromised fault tolerance. Recovering from compromised fault tolerance (enabling failed LUNs) Best practices when replacing hard drives. Automatic data recovery (rebuild). Time required for a rebuild. Failure of another drive during rebuild..

102 103

10 Array controller rmware
Determining the currently installed rmware version Updating controller rmware. Cloning controller rmware. Recovering corrupted rmware.

105 106

11 SCSI hard drive rmware. A Regulatory compliance and safety.
Regulatory compliance. Federal Communications Commission notice. Class A equipment. Class B equipment. Declaration of conformity for products marked with Modications. Cables. Regulatory compliance identication numbers. Regulatory compliance label location. International notices and statements. the FCC logo,. United States. only.

109 111

112 112
Canadian notice (avis Canadien). Class A equipment. Class B equipment. European Union notice. BSMI notice. Japanese notice. Korean notices. Safety notices. Battery replacement notice. Taiwan battery recycling notice. Power cords. Japanese power cord notice. Electrostatic discharge. Preventing electrostatic damage Grounding methods.

Index.

Tables
1.Document conventions. 2.LCD panel messages. 3.Component hot-pluggability. 67
This guide provides information about: Maintaining the MSA Servicing the MSA WARNING! To reduce the risk of personal injury from electric shock and hazardous energy levels, do not exceed the level of repairs specied in these procedures. Troubleshooting and repair procedures included in this document are detailed to allow only subassembly/module-level repair. Because of the complexity of the individual boards and subassemblies, do not attempt to make repairs at the component level or to make modications to any printed wiring board. Improper repairs can create conditions that are hazardous.

Off Off, On, or Blinking

Off On

Off, On, or Blinking

Fibre Channel I/O module LEDs
The Fibre Channel I/O module has three LEDs. The following gure and table describe their locations and meanings.

Name Status

Condition Off

Color

Meaning Power not applied to Fibre Channel I/O module, controller not plugged in, or unsuccessful controller POST Power applied and connection established between Fibre Channel I/O module and Fibre Channel controller Power is applied but there is no link to the controller or there is no c-ontroller present A 1-Gb link is established and the status is good A 1-Gb link was established but is now inactive A 2-Gb link is established and the status is good A 2-Gb link was established but is now inactive

1-Gb link status

Solid Blinking

Green Amber Green Amber

2-Gb link status
NOTE: If both link status lights are Off, a link has not been established since the application of power, initiation of a reset, hot-plug of Fibre Channel I/O module, removal of the transceiver, or hot-plug of the transceiver. If both link status lights are blinking amber, the MSA controller has been removed or a link cannot be established.
Power supply/fan assembly LEDs
Each power supply/fan assembly has one LED. The LED is illuminated when both the power supply and the fan are operational. When a power supply or fan fault occurs, the LED goes off. If the LED is off, AC power is not present or there is a complete power supply failure.
The power-supply-mounted fans cool the chassis by circulating air through the enclosure. The rate at which air moves (the airow) determines the amount of cooling. This airow is a function of fan speed (rpm). These fans, under the control of the Environmental Monitoring Unit (EMU) or the associated power supply, can operate at multiple speeds. This ensures that when the enclosure temperature changes, the fans can automatically adjust the airow. If one fan operates too slowly or completely shuts off, the other fan operates at a higher speed. At the same time, the error condition is reported to the user via the LCD panel on the front of the controller.

ACU website: http://h18000.www1.hp.com/products/servers/priliantstorage/ softwaer-management/acumatrix/index.html.
Array Conguration Utility Command Line Interface (ACU-CLI) overview
The Array Conguration Utility Command Line Interface (ACU-CLI) is a command line user interface used to congure an array controller and its storage. The ACU-CLI is supported for use in Windows, Linux, and HP-UX environments and can run locally through a browser, remotely through HP Systems Insight Manager (HP-SIM), or ofine from the MSA Support Software CD. For more information about the ACU-CLI, see the following: ACU-CLI section of the array conguration reference guide, available on the Documentation CD or the Technical documentation page of the MSA1000 website: http://www.hp.com/go/msa1000. ACU-CLI text le, available in the ACU-CLI folder on the Support Software CD.
MSA Command Line Interface (MSA-CLI) overview
The MSA Command Line Interface (MSA-CLI) is built into the MSA array controller rmware and offers a command level method of conguring, managing, and monitoring the array controller and its storage. The MSA-CLI is accessed through a host computer connected to the customized RJ-45Z serial port on the front of the MSA controller. This interface is provided for system administrators who prefer this method of system management, instead of a graphical user interface. All supported operating systems can access the CLI. Sample tasks include: conguring the storage units (LUNs), limiting access to the storage, and viewing controller and storage setup and status information.
Array Diagnostics Utility (ADU) overview
The HP Array Diagnostic Utility (ADU) is a web-based application that creates a report on all HP storage controllers and hard drives. This report provides vital information to assist in identifying faults or conditions that may require attention. Collected information includes: Hard drive errors Intermittent problems Pre-failure Notication Controller, array, and logical drive problems Third-party, possible counterfeit, or non-HP hard drives Array controller and hard drive rmware versions Length of hard drive service Physical and logical drive conguration information Serial numbers of hard drives and array controllers The ADU is installed on a server in the network and is provided on the MSA Support Software CD.

Manual and automatic (real-time) analysis capabilities on local or remote machines Filtering capabilities on error logs Automatic notication (customer or HP Support Center) Rule-based (allowing more frequent updates) Proactive indictment with the operating system Web and command line interfaces For more information, go to the WEBES website: http://h18023.www1.hp.com/support/svctools/ webes/?jumpid=reg_R1002_USEN.
5 LCD panel and message descriptions
LCD panel overview LCD message types LCD message descriptions NOTE: Some messages may specify a box number. For the MSA1000, the following box numbers are dened:
Pre-dened box number Box 1 Box 2 Box 3 Physical connection to MSA SCSI I/O module This is the MSA1000 chassis Port A of the SCSI I/O module Port B of the SCSI I/O module

LCD panel overview

Each array controller contains an integrated Liquid Crystal Display (LCD). This module is used for displaying informational and error messages, showing the status of the module, and for providing user input when required. Traditional Power-On Self-Test (POST) messages issued by the PCI-based controllers have been combined with runtime event notication messages to create a new set of controller display messages. The display module is capable of holding up to 100 messages. After this maximum size is reached, older messages are removed to make room for newer ones. When a new message is sent to the LCD, the display shows that message and ignores any previous scrolling position. This new message is now the most recent message available. To use the input buttons on the LCD panel: Press the Up navigation button to view older messages. Press the Down navigation button to view more recent messages. Press the Left navigation button to jump to the last message. Press the Right navigation button to enter the Management menu. Press the Left and Right navigation buttons at the same time to delete the currently displayed message.
The display module consists of the following components:
Fault LEDindicates an internal MSA component failure or external storage enclosure failure Two-line, twenty-column text message LCD display panel Left navigation button Up navigation button Right navigation button Down navigation button Redundancy Link LED (dual-controller congurations only) indicates that both array controllers and their corresponding Fibre Channel I/O modules are in a redundant mode of operation

LCD message types

The display module is capable of holding up to 100 messages. Once this maximum has been reached, older messages are removed to make room for newer ones. Messages can be of three types: Error messages Informational messages User input messages For a complete list of messages and their meanings, see LCD message descriptions.

12 ONLINE UPGRADE FLASHING FIRMWARE 13 ONLINE UPGRADE FLASHING STARTED 14 ONLINE UPGRADE FLASHING DONE 15 ONLINE UPGRADE FLASHING FAILED
Informational Informational Informational Error
Download new rmware from the HP website and initiate the rmware update again. If the issue persists, contact HP support.
16 ONLINE UPGRADE RESTARTING SYSTEM 17 ONLINE UPGRADE COMPLETE 18 ONLINE UPGRADE INCOMPLETE
An online rmware update is forcing an automatic restart of the system. An online rmware update has completed successfully. An online rmware update was successful, but a subsequent step failed. In dual-controller congurations, this may cause the other to not update properly. In dual-controller congurations, each controller restarts at least once during an online rmware update. During these restarts, the system is not redundant. The SCSI subsystem is being initialized as part of the power-on sequence. The MSA is scanning for any attached SCSI devices as part of the power-on sequence. The MSA is initializing all discovered SCSI devices as part of the power on sequence. Restart the system. In dual-controller congurations, if the 07 CLONE FIRMWARE? '<'=NO, '>'=YES message is displayed, select >.

Informational Error

19 ONLINE UPGRADE REDUNDANCY HALTED
20 INITIALIZING SCSI SUBSYSTEM 21 SCANNING FOR SCSI DEVICES
22 INITIALIZING SCSI DEVICES
24 BAD SCSI BUS MODE NON-LVD DEVICE FOUND Error
The MSA does not support SCSI Single Ended (SE) devices; it only supports SCSI Low Voltage Differential (LVD) devices. During a rmware update, a check is made of the hardware to ensure that it is compatible with the version of rmware being installed. This error indicates a problem. The MSA is not properly identifying an external storage enclosure.
The MSA should be powered off and then all SCSI devices attached to it should be examined. Any SE devices found should be removed and replaced with LVD devices. Contact HP support.

49 REDUNDANCY FAILED FIRMWARE LOCKUP
If the MSA is currently involved in host I/O, remove the standby array controller, wait 10 seconds, and then reinsert it, ensuring that it is fully seated in the chassis. If this does not resolve the issue, wait until downtime is available. Power off the MSA, remove both array controllers, and reinsert them, ensuring they are fully seated in the chassis. Should the issue persist, contact HP support. If the MSA is currently involved in host I/O, remove the standby array controller, wait 10 seconds, and then reinsert it ensuring that it is fully seated in the chassis. If this does not resolve the issue, wait until downtime is available. Power off the MSA, remove both array controllers, and reinsert them, ensuring they are fully seated in the chassis. Should the issue persist, contact HP support. If the MSA is currently involved in host I/O, remove the standby array controller, wait 10 seconds, and then reinsert it, ensuring that it is fully seated in the chassis. If this does not resolve the issue, wait until downtime is available. Power off the MSA, remove both array controllers, and reinsert them, ensuring they are fully seated in the chassis. Should the issue persist, contact HP support.
50 REDUNDANCY FAILED OUT OF MEMORY
While either attempting to enter redundant mode or already operating in redundant mode, one of the array controllers failed to allocate required memory. Redundancy is disabled at this time.
51 REDUNDANCY FAILED I/O REQUEST ERROR
While either attempting to enter redundant mode or already operating in redundant mode, one of the array controllers encountered an error while sending I/O between the two array controllers over the communication channel between them. Redundancy is disabled at this time.
52 REDUNDANCY FAILED PCI BUS ERROR Error
While either attempting to enter redundant mode or already operating in redundant mode, one of the array controllers encountered a PCI bus error on the communication channel used between the two array controllers. Redundancy is disabled at this time.
If the MSA is currently involved in host I/O, remove the standby array controller, wait 10 seconds, and then reinsert it, ensuring that it is fully seated in the chassis. If this does not resolve the issue, wait until downtime is available. Power off the MSA, remove both array controllers, and reinsert them, ensuring they are fully seated in the chassis. Should the issue persist, contact HP support. Reinsert the missing array controller ensuring that it is fully seated in the chassis.
53 REDUNDANCY FAILED NO SECOND CONTROLLER
While operating in redundant mode, one of the array controllers was removed. Redundancy is disabled at this time. The cache memory modules on the two controllers are not the same size. All cache memory modules must be the same size for redundancy to operate.

Replace the indicated hard drive.
100 VOLUME #<n> STATE OK
101 VOLUME #<n> STATE FAILED
The failed hard drives should be replaced as soon as possible. If necessary, restore data from backup.
102 VOLUME #<n> STATE INTERIM RECOVERY
The failed hard drives should be replaced as soon as possible.
103 VOLUME #<n> STATE REBUILDING
104 VOLUME #<n> STATE DISABLED
Power off the MSA and all attached external storage enclosures. Unplug and reinsert all SCSI hard drives, ensuring they are fully seated in their bays. Check the cables connecting the MSA to any attached storage enclosures. Power on the attached storage enclosures and then the MSA.
105 VOLUME #<n> STATE EXPANSION ACTIVE
The congured volume is currently performing a volume expansion operation.
106 VOLUME #<n> STATE WAITING TO REBUILD
The congured volume is waiting to start rebuilding data on a SCSI hard drive that replaces a previously failed drive. The rebuild may not have started yet because the array controller is already performing a rebuild on another congured volume. The congured volume is waiting to start a volume expansion operation. The expansion may have not started yet because another congured volume is undergoing expansion or a rebuild is occurring on the congured volume. The congured volume is missing too many of the SCSI hard drives that it is composed of, making it unusable. The volume is disabled.
107 VOLUME #<n> STATE WAITING TO EXPAND
108 VOLUME #<n> STATE MISSING DRIVES
Power off the MSA and then all attached storage enclosures. Unplug and reinsert all SCSI hard drives ensuring they are fully seated in their bays. Check the cables connecting the MSA to any attached storage enclosures. Power on the attached enclosures and then the MSA. The MSA should be powered off and the good hard drives should be restored while the failed hard drives should be replaced.
109 VOLUME #<n> STATE WRONG DRIVE REPLACED
The congured volume appears to have had known, good SCSI hard drives replaced instead of known, failed hard drives. The volume expansion operation on the congured volume has been disabled. This may be because a rebuild operation is ongoing, another expansion is already running, or the cache memory is disabled due to a low battery. The expansion will start once the condition has been cleared. The array controller is calculating and storing parity information for the congured volume and therefore performance may be lower until it completes.
110 VOLUME #<n> EXPANSION DISABLED
111 VOLUME #<n> INITIALIZING PARITY
112 VOLUME #<n> REBUILD FAILURE Error
The rebuild operation on the congured volume has failed.
If the volume is still operating in regenerative mode, remove the new SCSI hard drive that was added as a replacement for the original failed hard drive and replace it with a different new drive. Open the Array Conguration Utility (ACU) or MSA Command Line Interface (MSA-CLI) and use it to determine the state of the volume. If the volume is still operational, it is possible to retry the operation.

Customer replaceable components
For more information about HP's customer Self Repair program, contact your local service provider. For the North American program, go to: http://www.hp.com/go/selfrepair.
Parts-only warranty service
Your HP Limited Warranty may include a parts-only warranty service. Under the terms of part-only warranty service, HP provides replacement parts free of charge. For parts-only warranty service, CSR part replacement is mandatory. If you request HP to replace these parts for you, you are charged for the travel and labor costs of this service.
Procuring the spare component
TheIllustrated parts list lists the parts that qualify for CSR, including assembly and spare numbers. Parts have a nine-character component number on their label. The rst six characters identify the element; the last three characters dene the revision level. The replacement component revision level must be the same as, or later than, the number on the failed component. The greater the revision level, the later the revision.
Returning the defective component
In the materials shipped with a CSR component, HP species whether the defective component must be returned to HP. In cases where it is required, you must ship the defective part back to HP within a dened period of time, normally ve business days. The defective part must be returned with the associated documentation in the provided shipping material. Failure to return the defective part may result in HP billing you for the replacement. With a CSR, HP will pay all shipping and part return costs and determine the courier/carrier to be used.

Recommended tools

When replacing certain components, such as the power button assembly, the following tools may be necessary: 4-mm at-blade screwdriver Phillips screwdriver

Warnings and precautions

Electrostatic Discharge Information Grounding methods Equipment Symbols Weight Warning Rack warnings and precautions Device warnings and precautions

Removing or installing a Fibre Channel interconnect blank

Before you begin

CAUTION: Before removing a component or blanking panel from an operational device, make sure that you have the replacement part available. Removing a component or blank impacts the airow and cooling ability of the device. To avoid possible overheating, insert the new or replacement component within one or two minutes. If the internal temperature exceeds acceptable limits, the device may overheat and automatically shut down or restart. Parts can be damaged by electrostatic discharge. Use proper anti-static protection.
Removing an interconnect blank
1. Loosen the thumbscrew that holds the blank in place (1). 2. Remove the blank from the back of the unit (2).
Installing an interconnect blank
Place the blank into the MSA chassis bay and tighten the thumbscrew that holds the panel in place.
Removing or installing a hard drive blank
Removing a hard drive blank
CAUTION: A blank or a hard drive must be installed in each hard drive slot. Failure to install a blank or hard drive in all slots could result in thermal failures. 1. Review all warnings, cautions, and preparation procedures as detailed in Warnings and precautions. 2. Push in the tabs to unlock the hard drive blank (1). 3. Continue to press the tabs while pulling the hard drive blank out of the chassis (2).
Installing a hard drive blank
Slide the blank into the enclosure chassis bay until it clicks into place.

Replacing a hard drive

Verifying component failure
Hard drive failure is indicated by an amber Drive Failure LED. CAUTION: When replacing a hot-pluggable hard drive, it is important that you follow the guidelines in this section. Failure to do so could result in data loss and could void your warranty.
CAUTION: See the following for more information about replacing hard drives: Hard drive LEDs Hard drive failure and faulted LUNs The HP ProLiant Servers Troubleshooting Guide, available on the Management CD

Removing the component

Follow these guidelines when replacing hard drives: Never remove more than one hard drive at a time (two if using RAID 6). When you replace a hard drive, the controller uses data from the other hard drives in the array to reconstruct data on the replacement hard drive. If you remove more than one hard drive, a complete data set is not available to reconstruct data on the replacement drive(s) and permanent data loss could occur. Never remove a working hard drive. The amber Drive Failure LED indicates a hard drive that has been failed by the controller. Permanent data loss will occur if a working hard drive is removed while replacing a failed drive. Never remove a hard drive while another hard drive is being rebuilt. A drive's Online LEDs ash green (once per second) while it is being rebuilt. A replaced hard drive is rebuilt from data stored on the other hard drives. If the system has an online spare drive, wait for it to complete rebuilding before replacing the failed drive. When a hard drive fails, the online spare becomes active and begins rebuilding as a replacement drive. After the online spare has completed Automatic Data Recovery (the online LEDs will be continuously lit), replace the failed hard drive with a new replacement drive. Do not replace the failed hard drive with the online spare. The system will automatically rebuild the replacement hard drive and reset the spare hard drive to an available state.

If you replace a hard drive while the system is off, it may be necessary to rebuild the replaced hard drive. 1. Review all warnings, cautions, and preparation procedures as detailed in Warnings and precautions. 2. Be sure that the Online and Drive Access LED LEDs are both Off and that the Fault LED is On. 3. Press the ejector button (1) and pull the release lever to the full open position (2). 4. Wait approximately 10 seconds for the internal disk to stop spinning before continuing to the next step. Handling a hard drive while the internal media is spinning can damage the media. 5. Pull the hard drive out of the enclosure bay (3).

Installing the component

1. Press the hard drive ejector button (1) and pull the release lever to the full open position (2).
2. Slide the replacement hard drive into the same bay from which the old hard drive was just removed, making sure that the lever is in the full open position to ensure a correct latch. Then, push the hard drive into the bay as far as it will go, pressing rmly on the hard drive to make sure that it is seated properly (1).
3. Close the ejector lever against the front of the hard drive (2). 4. Verify that the hard drive is rmly engaged and that the ejector lever is latched. CAUTION: Data can be lost if the hard drive is not rmly seated. 5. As the hard drive begins to spin, the drive LEDs will illuminate, indicating that the system has recognized the new drive. In fault-tolerant congurations, allow the replacement hard drive to be reconstructed automatically from parity data stored on the other hard drives. While the data is being rebuilt, the LEDs will ash.
Verifying proper operation
The hard drive LEDs light one at a time and then turn off together to indicate that the system has recognized the new drive. In fault tolerant congurations, allow the replacement hard drive to be reconstructed automatically with data from the other hard drives. While reconstruction is in progress, the online LEDs ash.
Replacing the controller, controller cache, or controller battery
The following sections detail procedures for replacing an MSA array controller, controller cache module, or controller cache battery pack: Before you begin Verifying component failure Removing the controller Removing the controller cache module Removing the controller cache battery pack Installing the controller cache battery pack Installing the controller cache module Installing the controller Verifying the replacement Perform only the steps required for your service event. For example, re-seating an array controller includes Removing the controller and Installing the controller.

4. Swing the battery pack away from the cache module to about a 30-degree angle.
5. Lift the pack upward to unhook the top of the battery pack.
6. Wait approximately 15 seconds after removing the old battery to allow the battery charge monitor to reset. NOTE: If you do not wait 15 seconds after removing the old battery, full charge of the new cache battery is delayed. 7. Repeat step 3 through step 6 to remove the second battery pack on this cache module.
Installing the controller cache battery pack
1. Remove the array controller as instructed in Removing the controller. 2. Remove the controller cache module as instructed in Removing the controller cache module. 3. Remove the controller cache battery pack as instructed in Removing the controller cache battery pack. 4. Hook the top of the new NiMH battery pack to the top of the module, with the pack held at a 30-degree angle to the plane of the module.
5. After the battery pack is hooked in position, swing the pack downward, ensuring that the bottom clip and two pegs line up with the holes in the cache module. 6. Verify that the top hook (1) and bottom clip (2) on the battery pack are securely attached to the cache module.
7. Repeatstep 4 through step 6 to install the second battery pack on the module. 8. Install the serviced cache module in the controller as instructed in Installing the controller cache module. NOTE: The cache is disabled while the batteries are charging and is automatically re-enabled after the batteries are fully charged.
Installing the controller cache module
1. Slide the cache module into the controller (1), until the side latches are fully engaged (2).
2. Close the controller cover. 3. Insert the serviced controller into the chassis as instructed in Installing the controller.
Installing the controller
1. Insert the new or serviced array controller into the MSA chassis controller bay (1). 2. Push the controller in as far as it will go, making sure that the controller is seated, and then press the latch handle inward until it is ush against the front panel (2).
3. If necessary, power on the MSA as instructed in Connecting the Power and Applying power to the MSA. NOTE: Each time a system with two controllers is powered on or restarted, or if a second controller is hot-plugged into a single-controller conguration, the rmware versions on the controllers are compared. If the rmware versions on the two controllers are not the same, the controller LCD panel prompts to clone the rmware from the controller with the latest version onto the other controller. For information, see Cloning controller rmware.
After replacing the array controller, controller cache module, or cache module battery pack, verify that: The controller Heartbeat LED is blinking. The controller Fault LED is off. No new error messages are displayed on the array controller LCD display panel. After installing a new array controller, verify that the latest available rmware is installed. See Array controller rmware for the following instructions: Determining the currently-installed rmware version Updating controller rmware Cloning controller rmware Recovering corrupted rmware

4. Attach SCSI cables to the MSA expansion SCSI connectors. 5. Attach the SCSI cables to the expansion storage enclosures. 6. Attach the power cords to the power supplies at the rear of the enclosures. 7. Power on each enclosure and wait for it to complete its startup routine. 8. Power on the MSA array controller system using the power button on the front of the system. 9. Wait for the MSA to complete its startup routine. 10. Power on the host servers and allow them to boot. 11. Run the ACU or the MSA-CLI to verify that the current conguration is maintained and new volumes are identied. If the migrated volumes are not identied, power down the system and verify the connections. 12. Run the ACU or the MSA-CLI to add LUNs. Some systems may require rebooting to use the new LUNs. For more information on migrations, go to the HP storage website: http://www.hp.com/go/storage.
Moving hard drives or arrays
CAUTION: Back up all data before removing hard drives or changing congurations. Failure to do so could result in permanent loss of data. Before moving hard drives and arrays, run the ACU or the MSA-CLI. Before moving hard drives, the following conditions must be met: No hard drives are failed, missing, or degraded. No more than 32 LUNs will be congured for a controller. The array should be in its original conguration with no active spare hard drives. Capacity expansion is not running. Controller rmware is the latest version (recommended). Before moving an array from one controller to another, the following additional conditions must be met: All hard drives in the array must be moved at the same time. Positions of hard drives on the destination controller must remain the same during relocation of the array. When the appropriate conditions are met, perform the following procedure to move the hard drives: 1. Power Off the MSA storage system as detailed in Removing power from the MSA. 2. Move the hard drives. 3. Power On the MSA storage system as detailed in Applying power to the MSA. 4. The following message should be displayed on the LCD front panel of the controller 86 DRIVE POSITION CHANGE DETECTED. IMPORTANT: If the following LCD message is displayed: 121 NO VOLUMES DETECTED, turn Off system immediately to avoid data loss, and return the hard drives to their original locations. 5. Check the new conguration by running the ACU or MSA-CLI.
Enclosure and hard drive migrations
8 Capacity expansion and extension
Array expansion is the addition of physical hard drives to an array that has already been congured. The capacity of these added physical hard drives may then be added to an existing LUN on the array or congured into a new LUN. LUN extension is the enlargement of an existing LUN after the corresponding array has undergone capacity expansion. Expansions and extensions are performed through the ACU or MSA-CLI. If you are using hot-pluggable hard drives, expansion can be performed online (that is, without shutting down the operating system). Online extension can only be performed if supported by the operating system. CAUTION: Back up all data before removing hard drives or changing congurations. Failure to do so could result in permanent loss of data.

BSMI notice

Japanese notice

Korean notices

Safety notices
Battery replacement notice
Your computer is equipped with a lithium manganese dioxide, a vanadium pentoxide, or an alkaline internal battery or battery pack. There is a danger of explosion and risk of personal injury if the battery is incorrectly replaced or mistreated. Replacement is to be done by an HP authorized service provider using the HP spare part designated for this product. For more information about battery replacement or proper disposal, contact an HP authorized reseller or HP authorized service provider. WARNING! Your computer contains an internal lithium manganese dioxide, a vanadium pentoxide, or an alkaline battery pack. There is risk of re and burns if the battery pack is not properly handled. To reduce the risk of personal injury: Do not attempt to recharge the battery. Do not expose to temperatures higher than 60C. Do not disassemble, crush, puncture, short external contacts, or dispose of in re or water. Replace only with the HP spare part designated for this product.
Batteries, battery packs, and accumulators should not be disposed of together with the general household waste. To forward them to recycling or proper disposal, please use the public collection system or return them to HP, an authorized HP Partner, or their agents.
For more information about battery replacement or proper disposal, contact an HP authorized reseller or service provider.
Taiwan battery recycling notice
The Taiwan EPA requires dry battery manufacturing or importing rms in accordance with Article 15 of the Waste Disposal Act to indicate the recovery marks on the batteries used in sales, giveaway or promotion. Contact a qualied Taiwanese recycler for proper battery disposal.

Power cords

The power cord set must meet the requirements for use in the country where the product was purchased. If the product is to be used in another country, purchase a power cord that is approved for use in that country. The power cord must be rated for the product and for the voltage and current marked on the product electrical ratings label. The voltage and current rating of the cord should be greater than the voltage and current rating marked on the product. In addition, the diameter of the wire must be a minimum of 1.00 mm2 or 18 AWG, and the length of the cord must be between 1.8 m (6 ft) and 3.6 m (12 ft). If you have questions about the type of power cord to use, contact an HP authorized service provider. NOTE: Route power cords so that they will not be walked on and cannot be pinched by items placed upon or against them. Pay particular attention to the plug, electrical outlet, and the point where the cords exit from the product.