Dell PowerVault MD3000 Technical manual

www.dell.com | support.dell.com
Dell™ PowerVault™
MD3000 Systems
Information Update
信息更新
Mise à jour des informations
Aktuelle Informationen
アップデート情報
설명서 갱신본
Actualización de información

www.dell.com | support.dell.com
Dell™ PowerVault™
MD3000 Systems
Information Update

Notes, Notices, and Cautions
NOTE: A NOTE indicates important information that helps you make better use of your computer.
NOTICE: A NOTICE indicates either potential damage to hardware or loss of data and tells you how to avoid
the problem.
CAUTION: A CAUTION indicates a potential for property damage, personal injury, or death.
____________________
Information in this document is subject to change without notice.
© 2006 Dell Inc. All rights reserved.
Reproduction in any manner whatsoever without the written permission of Dell Inc. is strictly forbidden.
Trademarksusedinthistext:Dell,theDELLlogo,OpenManage,andPowerVaultaretrademarksofDellInc.;Microsoft,Windows,andDirectX
are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries; SUSE is a registered
trademark of Novell, Inc., in the United States and other countries; Red Hat is a registered trademark of Red Hat, Inc.
Other trademarks and trade names may be used in this document to refer to either the entities claiming the marks and names or their products.
Dell Inc. disclaims any proprietary interest in trademarks and trade names other than its own.
November 2006 P/N FF455 Rev. A01

Information Update 3
This document provides updated information regarding the Dell
®
PowerVault
®
MD3000 Storage
Enclosure. The following topics are included:
• Modular Disk Storage Manager
• Physical disks
• Virtual disks
• Virtual disk migration
• Configuring the RAID controller module network
• Password resets
• Issues and workarounds in Red Hat
®
Enterprise Linux (Version 3)
• MD3000 configuration considerations
• MD3000 maintenance considerations
Modular Disk Storage Manager
The following updates apply to the Modular Disk Storage Manager (MD Storage Manager):
• The MD Storage Manager graphical user interface (GUI) lists RAID 3 as an option for virtual disks.
This option is not supported for this product. If the RAID 3 option is selected, the MD3000 system
creates a RAID 5 RAID group. The level of this RAID group is erroneously reported as RAID 3. The
RAID 3 option will be removed in a subsequent release.
• In order to run MD Storage Manager on a Microsoft
®
Windows
®
2000 operating system, the DirectX
®
8.1 Runtime for Windows application programming interface must be installed.
• If you are running the Windows 2003 operating system, install the Microsoft Hotfix B916048 software
on the host before installing the MD Storage Manager software. The hotfix is located in the
\windows\Windows_2003_Server\hotfixes directory on the MD3000 Resource CD.
• The Microsoft VSS and VDS installer services for storage provisioning are available on the MD3000
Resource CD in the
\windows\VDS_VSS directory.
•
When the MD Storage Manager is being installed on Red Hat Enterprise Linux systems or SUSE
®
Linux Enterprise Servers
, or when the MD Storage Manager is run under SUSE Linux Enterprise
Servers, Japanese and Simplified Chinese are not supported.
• If a Host Context Agent refreshes the topology to an attached array when that array is being serviced,
a
Resolve Topology Conflicts message might be displayed in the Status portlet of the
Summary page. This can occur when RAID Controller Module FW is being downloaded, or when a
RAID Controller Module is being replaced or taken offline. If the
Resolve Topology
Conflicts
message is displayed during a service operation, complete the service and then either
wait 30 minutes or restart the Host Context Agent. Refer to the
Modular Disk Storage Manager User’s
Guide
for more information on the Host Context Agent.
NOTE: Do not attempt any corrective actions within the Resolve Topology Conflicts window in this situation.
Conflicts eventually resolve themselves.

4Information Update
Physical Disks
If a disk drive rebuild fails because of a source drive failure or because the drive is too small, the user
interface reports a failure of the physical disk even though the LED state on the drive indicates that
the rebuild was aborted.
Virtual Disks
If a Microsoft Windows operating system is the host server operating after virtual disks are created on the
array, you must partition and format the virtual disk using the disk management Microsoft Management
Console (MMC).
Go to Settings→Control Panel→Administrative Tools, and select the Disk Management MMC
option to format and partition the disk.
A reboot might be necessary in order for a Windows operating system to correctly recognize a target
virtual disk following a virtual disk copy. In particular, if the properties of a Windows-based volume
indicate a RAW file system (one that has not been formatted), reboot the system so that the correct
virtual disk is recognized.
When re-creating a Snapshot Virtual Disk using a Windows operating system, remove the drive letter
from the Snapshot Virtual Disk before performing the re-creation operation, and then reassign the drive
letter to the Snapshot Virtual Disk after the snapshot has been created. This sequence ensures that the
latest data is preserved on the Snapshot Virtual Disk.
Virtual disks that do not have all their member drives at controller startup are reported as optimal.
The firmware reports this to prevent disks failing from disconnected enclosures.
For Non-Redundant MSCS Cluster configurations, download an updated NVSRAM file to avoid
spurious Virtual Disk Not on Preferred Path conditions. The NVSRAM file is prefixed
with non-redundant-MSCS and can be found on the MD3000 Resource CD in the Utility/NVSRAM
directory. From the MD Storage Manager GUI, select Support→Download firmware→Download
RAID Controller Module NVSRAM and select the non-redundant-MSCS NVSRAM file.
Virtual Disk Migration
Virtual disk migration describes the action of moving drives and drive modules within the same storage
array or to different storage arrays. All data on the virtual disks is preserved during the migration process.
NOTE: Only disk groups and associated virtual disks with all member physical disks present can be migrated from
one storage array to another. Dell recommends that you only migrate disk groups that have all their associated
member virtual disks in an optimal state.
NOTE: The number of physical disks and virtual disks that a storage array supports limits the scope of the
migration.

Information Update 5
Hot and Cold Migration
Use either of the following methods to move disk groups and virtual disks:
• Hot virtual disk migration — Disk migration with the destination storage array power turned on.
• Cold virtual disk migration — Disk migration with the destination storage array power turned off.
NOTE: To ensure that the migrating disk groups and virtual disks are correctly recognized when the target storage
array has an existing physical disk, use hot virtual disk migration.
Moving Physical Disks to the Destination Array for Migration
When inserting drives into the destination storage array during hot virtual disk migration, wait for the
inserted physical disk to be displayed in the MD Storage Manager before inserting the next physical disk.
NOTICE: Without the delay between drive insertions, the storage array can become unstable and manageability is
temporarily lost.
Migrating Virtual Disks from Multiple Storage Arrays into a Single Storage Array
When migrating virtual disks from multiple, different storage arrays into a single destination storage
array, move all of the physical disks from the same storage array as a set into the new destination storage
array. Ensure that all of the physical disks from a storage array are migrated to the destination storage
array before starting migration from the next storage array.
NOTE: If the drive modules are not moved as a set to the destination storage array, the newly relocated disk groups
might not be accessible.
Migrating Virtual Disks to a Storage Array with No Existing Physical Disks
When migrating disk groups or a complete set of physical disks from a storage array to another storage
array that has no existing physical disks, turn off the destination storage array. After the destination
storage array has been turned on and has successfully recognized the newly migrated physical disks,
migration operations can continue.
NOTE: Disk groups from multiple storage arrays should not be migrated at the same time to a storage array
that has no existing physical disks. Use cold virtual disk migration for the disk groups from one storage array.
Virtual Disk Migration and Premium Features
Before migrating disk groups and virtual disks, enable the required premium features on the destination
storage array. If a disk group is migrated from a storage array that has a premium feature enabled and the
destination array does not have the feature enabled, an Out of Compliance error message can be
generated. For specific procedures to correct the error, refer to the Recovery Guru.

6Information Update
Configuring the RAID Controller Module Network
RAID controller module network configurations can be assigned using a DHCP server (the default
setting). However, if a DHCP server is not available and the 10-second selection period times out, the
RAID controller modules use the 192.168.128.101 static IP address for controller 0. For controller 1,
the RAID controller modules use the 192.168.128.102 static IP address.
Password Resets
If a password reset is ever required, refer to the password reset procedure for the Dell PowerVault
MD3000 Storage Enclosure at http://support.dell.com. If you require help with this procedure,
contact Dell for technical assistance. For more information on contacting Dell, see the "Getting Help"
chapter of the Hardware Owner’s Manual.
Issues and Workarounds using Red Hat Enterprise Linux (Version 3)
Users running Red Hat Enterprise Linux 3.0 should note the following issues and workarounds.
Skipped LUNs
The Red Hat Enterprise Linux SCSI layer does not support skipped (sparse) LUNs. If the mapped LUNs
are not contiguous, the kernel does not scan any of the remaining LUNs after the interruption.
Consequently, all LUNs after the skipped LUN are not available to the host.
The skipped LUNS can be accounted for by running hot_add to map the missing LUN numbers
each time the host is rebooted.
In-Band Management
Because the Linux 2.4 kernel does not support skipped LUNs, in-band management is not supported
on Red Hat 3.0. Use out-of-band management instead.
MPP Failover Driver Loading
The mpp_vhba failover driver component does not load unless a storage array is connected and a LUN
is mapped to the host server.
Follow these steps to ensure proper MPP driver loading and LUN mapping:
1
Ensure that the storage array is connected to the host server and that all installation instructions have
been completed, including the connection of the out-of-band (Ethernet) connections to each RAID
controller module.
2 Create a virtual disk and map it to LUN 0.
3 Reboot the host.
After the host is rebooted, the LUN is properly mapped and the MPP driver is loaded. Subsequently,
all other sequential LUNs can be created and registered to the OS by running hot_add.

Information Update 7
MD3000 Configuration Considerations
When configuring the MD3000, consider the following updates.
Changing Controller Ownership of the Virtual Disk
For hosts that have a single data path to the MD3000 storage array, the virtual disk must be owned by the
controller to which the host is connected. Configure this storage array before starting I/O operations.
To assign ownership of the virtual disk to the connected controller, select Modify→Change Virtual Disk
Ownership/Preferred Path from the MD Storage Manager GUI after the virtual disk is created.
Advanced Configuration Considerations
Some configuration options are best explored with the aid of Dell technical assistance. In particular,
the Advanced option of the MD Storage Manager’s Tools→Change Network Configuration GUI
should only be used with the guidance of Dell technical assistance.
Command Line Interface Updates
Users of the command line interface should note the following information:
1
The diagnose controller command’s testID parameter takes the following options:
• 1 — Reads the test
• 2 — Performs a data loop back test
• 3 — Writes the test
2 If configuration of the MD3000 storage array is cleared using the clear storagearray configuration
command, restart all attached hosts prior to reconfiguring the storage array.
3 When using the command-line interface to create host ports, ensure that the host type is specified in
the command. If no host is specified, the default host is Windows 2000/Server 20003 Non-Clustered.
4The SMcli -d parameter shows all storage arrays that are currently discovered.
SAS Connections
OpenManage software users who want to install a SAS 5/E connected to an MD3000 storage array must
first download and install the OM5.1.0.1 patch from the Dell Support website at support.dell.com.
This is a prerequisite in order for the OpenManage software to function properly.
A SAS Device Miswire Recovery Guru condition can be generated by connecting the host port of one
controller to the unused expansion port on the other controller in a MD3000 enclosure. This miswire
was not specified in the recovery procedures included with the MD3000 documentation.

8Information Update
Host Access Configuration
To configure hosts for the MD3000 storage array, use the Configure Host Access (Automatic) GUI.
Host access that is manually configured requires special attention to ensure that the correct SAS host
port World Wide IDs are selected for each host. If any incorrect IDs are configured, an inaccurate
topology results and a Topology Conflict is flagged on the Summary page. You can use the SAS 5/E
HBA BIOS setup program to identify the World Wide IDs of the SAS host ports.
NOTE: Verify that your host or management station network configuration—including station IP address, subnet
mask, and default gateway—is correct before adding a new storage using the Automatic option. For all Linux
kernels, set the default gateway when adding a storage array using the Automatic option so that broadcast
packets are sent to 255.255.255.255.
The Configure Host Access (Automatic) interface has changed. To ensure that only specified hosts get
added, select the hosts from the Available Hosts window that you want configured and click Add to move
them to the Selected Hosts window. After the specified hosts have been added to the Selected Hosts
window, click OK to run the automatic host configuration.
MD3000 Maintenance Considerations
For Linux kernels, stop and then restart the SMagent after performing either of the two following
maintenance tasks:
• Moving a controller offline or replacing a controller.
• Removing SAS cables from or attaching SAS cables to Red Hat Enterprise Linux (version 4),
SUSE Linux Enterprise Server 9, or SUSE Linux Enterprise Server 10 host servers.
NOTE: If a Resolve Topology Conflicts message appears in the Status portlet of the Summary page after restarting
the SMagent, it might be necessary to restart the host server to clear this message. The Resolve selection in the
Topology Conflict wizard should not be selected. The MD3000 will continue to service I/O requests; however, the
array could become partially managed if only in-band management is used. Stop all I/O operations on any host
servers that were involved in the maintenance operation and perform a restart.

www.dell.com | support.dell.com
Dell™ PowerVault™
MD3000 系统
信息更新

注、注意和警告
注：注表示可以帮助您更好地使用计算机的重要信息。
注意：注意表示可能会损坏硬件或导 数据丢失，并告诉您如何避免此类问题。
警告：警告表示可能会导致财产损失、人身伤害甚至死亡。
____________________
本说明文件中的信息如有更改，恕不另行通知。
© 2006 Dell Inc. 版权所有，翻印必究。
未经 Dell Inc. 书 许可，严禁以任何形式进行复制。
本文中使用的商标：Dell、DELL 徽标、OpenManage 和PowerVault 是Dell Inc. 的商标； Microsoft、Windows 和DirectX 是
Microsoft Corporation 在美国和／或其它国家和地区的注册商标或商标； SUSE 是Novell, Inc. 在美国和／或其它国家和地区的注
册商标； Red Hat 是Red Hat, Inc. 的注册商标。
本文件中述及的其它商标和产品名称是指拥有相应商标和名称的公司或其制造的产品。Dell Inc. 对本公司的商标和产品名称之外
的其它商标和产品名称不拥有任何专有权。
2006 年11 月P/N FF455 修订版 A01

信息更新 11
本说明文件提供有关
Dell
®
PowerVault
®
MD3000
存储设备的更新信息。包含以下主题：
•模块化磁盘存储管理器
•物理磁盘
•虚拟磁盘
•虚拟磁盘迁移
•配置 RAID 控制器模块网络
•密码重设
•Red Hat®Enterprise Linux（第 3 版）中的 题和解决方法
•MD3000 配置注意事项
•MD3000 维护注意事项
模块化磁盘存储管理器
以下更新适用于模块 磁盘存储管理器（
MD
存储管理器）：
•MD 存储管理器图形用户界面 (GUI) 将RAID 3 作为虚拟磁盘的一个选项列出。本产品不支持
此选项。如果选择 RAID 3 选项，MD3000 系统将创建一个 RAID 5 RAID 组。这一级别的 RAID
组被错误地报告为 RAID 3。在后续版本中将删除该 RAID 3 选项。
•为了在 Microsoft®Windows®2000 操作系统上运行 MD 存储管理器，必须安装适用于 Windows
应用程序编程接口的 DirectX®8.1 Runtime。
•如果您运行的是 Windows 2003 操作系统，则在安装 MD 存储管理器软件前，先在主机上
安装 Microsoft 热修补程序 B916048 软件。该热修补程序位于 MD3000 Resource CD 上的
\windows\Windows_2003_Server\hotfixes 目录下。
•MD3000 Resource CD 中的 \windows\VDS_VSS 目录下提供了用于进行存储预配置的 Microsoft
VSS 和VDS 安装程序服务。
•如果将 MD 存储管理器安装在 Red Hat Enterprise Linux 系统或 SUSE®Linux Enterprise Server
上，或者 MD 存储管理器在 SUSE Linux Enterprise Server 下运行，则不支持日文和简体中文。
• 如果所连接的某个阵列正在提供服务，而此时主机环境代理
(Host Context Agent)
将拓扑刷新到该
阵列，则在“摘要”页的“状态”门户中可能显示
Resolve Topology Conflicts （解决拓扑
冲突）信息。当正在下载
RAID
控制器模块
FW
时，或者更换
RAID
控制器模块或使其脱机时，
可能出现该 题。如果在服务运行过程中显示 Resolve Topology Conflicts（解决拓扑
冲突）信息，则完成服务，然后等待 30 分钟或重新启动主机环境代理 (Host Context Agent)。
有关主机环境代理
(Host Context Agent)
的详情，请参阅《模块 磁盘存储管理器用户指南》。
注：在此情况下，请勿在 Resolve Topology Conflicts（解决拓扑冲突）窗口内尝试任何纠正措施。
冲突最终可自行解决。

12 信息更新
物理磁盘
如果由于源驱动器故障或驱动器太小而导致磁盘驱动器重建失败，则即使驱动器上的
LED
状态表示已
中止重建，用户界面也会报告物理磁盘故障。
虚拟磁盘
如果在阵列上创建虚拟磁盘后，
Microsoft Windows
操作系统是主机服务器操作系统，则必须使用磁盘
管理
Microsoft
管理控制台
(MMC)
对虚拟磁盘进行分区和格式 。
转至
Settings
（设置） →
Control Panel
（控制面板） →
Administrative Tools
（管理工具），然后选择
Disk Management MMC
（磁盘管理
MMC
）选项，对磁盘进行格式 和分区。
为了使
Windows
操作系统能够根据虚拟磁盘副本正确地识别目标虚拟磁盘，可能需要重新引导。尤其
是当基于 Windows 的卷的属性指示 RAW 文件系统（尚未进行格式化的文件系统）时，应重新引导
系统，
以便识别正确的虚拟磁盘。
当使用
Windows
操作系统重新创建快照虚拟磁盘时，请删除快照虚拟磁盘的驱动器号后再执行重新创
建操作，然后在创建快照之后将驱动器号重新分配给快照虚拟磁盘。此序列可确保保留快照虚拟磁盘
上的最新数据。
在控制器启动时，不具有其所有成员驱动器的虚拟磁盘被报告为处于最佳状态。固件将报告这些信息，
以防止当存储设备断开连接时致使磁盘出现故障。
对于非冗余
MSCS
群集配置，请下载更新的
NVSRAM
文件，以免出现虚假的 Virtual Disk Not
on Preferred Path
（虚拟磁盘不在首选路径上）情况。
NVSRAM
文件带有前缀
non-redundant-
MSCS
，可以在
MD3000 Resource CD
的
Utility/NVSRAM
目录中找到该文件。从
MD
存储管理器
GUI
中，选择
Support
（支持） →
Download firmware
（下载固件） →
Download RAID Controller
Module NVSRAM
（下载
RAID
控制器模块
NVSRAM
），然后选择
non-redundant-MSCS NVSRAM
文件。

信息更新 13
虚拟磁盘迁移
虚拟磁盘迁移描述在同一存储阵列内或不同存储阵列之间移动驱动器和驱动器模块的操作。在迁移过
程中，将保留虚拟磁盘上的所有数据。
注：仅带有所有成员物理磁盘的磁盘组和关联的虚拟磁盘可以从一个存储阵列迁移到另一个存储阵列。
Dell 建议您仅迁移其所有关联的成员虚拟磁盘均处于最佳状态的磁盘组。
注：存储阵列支持的物理磁盘和虚拟磁盘数限制了迁移范围。
热迁移和冷迁移
请使用以下一种方式移动磁盘组和虚拟磁盘：
•热虚拟磁盘迁移 —在目标存储阵列接通电源的情况下进行磁盘迁移。
•冷虚拟磁盘迁移 —在目标存储阵列断开电源的情况下进行磁盘迁移。
注：如果目标存储阵列现有一个物理磁盘，要确保正确地识别所迁移的磁盘组和虚拟磁盘，请使用热虚
拟磁盘迁移。
将物理磁盘移到目标阵列以进行迁移
在热虚拟磁盘迁移过程中，当将驱动器插入目标存储阵列时，请等待
MD
存储管理器中显示所插入的
物理磁盘，然后插入下一个物理磁盘。
注意：如果驱动器插入操作之间没有延迟，存储阵列可能变为不稳定，且可能暂时失去可管理性。
将多个存储阵列中的虚拟磁盘迁移到单一存储阵列
当将多个不同存储阵列中的虚拟磁盘迁移到单一目标存储阵列时，请将同一存储阵列中的所有物理磁
盘作为一个组移到新的目标存储阵列。请确保存储阵列中的所有物理磁盘已迁移到目标存储阵列，
然后开始从下一个存储阵列进行迁移。
注：如果驱动器模块不是作为一个组移动到目标存储阵列，则可能无法访问新近重新放置的磁盘组。
将虚拟磁盘迁移到现不具 物理磁盘的存储阵列
当将磁盘组或一个完整的物理磁盘组从一个存储阵列迁移到另一个现不具备物理磁盘的存储阵列时，
请关闭目标存储阵列。在开启目标存储阵列并成功识别新近迁移的物理磁盘后，可以继续迁移操作。
注：请勿同时将多个存储阵列中的磁盘组迁移到现不具备物理磁盘的存储阵列。对于一个存储阵列中的
磁盘组，请使用冷虚拟磁盘迁移。
虚拟磁盘迁移和高级功能
在迁移磁盘组和虚拟磁盘前，请在目标存储阵列上启用所需的高级功能。如果从已启用高级功能的存
储阵列迁移磁盘组，而目标阵列未启用该功能，则可能生成 Out of Compliance（不兼容）错误
信息。
有关纠正此错误的特定过程，请参阅恢复
Guru (Recovery Guru)
。

14 信息更新
配置 RAID 控制器模块网络
通过使用
DHCP
服务器 默认设置），可以指定
RAID
控制器模块的网络配置。然而，如果
DHCP
服务器不可用并且超过了
10
秒的选择时间，则对于控制器
0
，
RAID
控制器模块将使用静态
IP
地址
192.168.128.101 。对于控制器
1
，
RAID
控制器模块将使用静态
IP
地址 192.168.128.102 。
密码重设
如果需要重设密码，请参阅 http://support.dell.com 中有关 Dell PowerVault MD3000 存储
设备的密码重设过程。如果您需要有关该过程的帮助，请与 Dell 联系获取技术支持。有关与 Dell
联系的详情，
请参阅《硬件用户手册》的“获得帮助”一节。
Red Hat Enterprise Linux 第 3 版）中的问题和解决方法
运行
Red Hat Enterprise Linux 3.0
的用户应注意以下问题和解决方法。
跳过的 LUN
Red Hat Enterprise Linux SCSI
层不支持跳过的（稀疏的）
LUN
。如果映射的
LUN
不连续，则在中断后，
内核将法扫描任何其它
LUN
。因此，跳过的
LUN
之后的所有
LUN
将不能用于主机。
在每次重新引导主机后，运行
hot_add
以映射缺失的
LUN
号，可以补上跳过的
LUN
。
带内管理
因为
Linux 2.4
内核不支持跳过的
LUN
，所以
Red Hat 3.0
不支持带内管理。但可以使用带外管理。
载入 MPP 故障转移驱动程序
除非已连接存储阵列并且
LUN
已映射至主机服务器，否则将不载入
mpp_vhba
故障转移驱动程序组件。
请执行以下这些步骤，以确保正确地载入
MPP
驱动程序和映射
LUN
：
1
确保存储阵列连接到主机服务器并且已完成所有安装说明，包括以带外（以太网）连接方式连接
到每个
RAID
控制器模块。
2
创建一个虚拟磁盘并将其映射到
LUN 0
。
3
重新引导主机。
重新引导主机后，
LUN
已正确地映射，并载入了
MPP
驱动程序。接下来，可以创建所有其他后续
LUN
，
并通过运行
hot_add
向操作系统注册。

信息更新 15
MD3000 配置注意事项
配置
MD3000
时，请考虑以下更新信息。
更改虚拟磁盘的控制器所有权
对于与
MD3000
存储阵列之间具有单一数据路径的主机，虚拟磁盘必须由与主机连接的控制器所拥有。
请在启动
I/O
操作前配置此存储阵列。
要将虚拟磁盘的所有权分配给所连接的控制器，在创建虚拟磁盘后，从
MD
存储管理器
GUI
中选择
Modify
（修改） →
Change Virtual Disk Ownership/Preferred Path
（更改虚拟磁盘所有权
/
首选路径）。
高级配置注意事项
某些配置选项最好在
Dell
技术支持人员的帮助下使用。尤其是
MD
存储管理器的
Tools
（工具） →
Change Network Configuration
（更改网络配置）
GUI
中的
Advanced
（高级）选项，只限在
Dell
技术
人员的指导下使用。
命令行界面更新
命令行界面的用户应注意以下信息：
1
诊断控制器命令的
testID
参数采用以下选项：
•1 —
Reads the test
（
1
— 读取测试）
•2 —
Performs a data loop back test
（
2
— 执行数据环回测试）
•3 —
Writes the test
（
3
— 写入测试）
2
如果使用
clear storagearray configuration
（清除存储阵列配置）命令来清除
MD3000
存储阵列的
配置，则在重新配置该存储阵列前，请先重新启动所有连接的主机。
3
当使用命令行界面创建主机端口时，请确保已在命令中指定主机类型。如果未指定主机，
则默认主机为
Windows 2000/Server 20003
（非群集模式）。
4
SMcli -d 参数显示当前查找到的所有存储阵列。
SAS 连接
要安装
SAS 5/E
（连接到
MD3000
存储阵列）的
OpenManage
软件用户首先必须从
Dell
支持
Web
站点
support.dell.com
下载并安装
OM5.1.0.1
增补软件。这是使
OpenManage
软件正常工作的前提条件。
在MD3000 存储设备中，通过将一个控制器的主机端口连接到另一个控制器上未使用的扩充端口，
可以生成 SAS Device Miswire Recovery Guru（SAS 设备错接恢复 Guru）状况。MD3000 说明文件
所包含的恢复过程中并未指出这种错接。

16 信息更新
主机访问配置
要为
MD3000
存储阵列配置主机，请使用
Configure Host Access (Automatic)
（配置主机访问
[
自动
]
）
GUI。手动配置主机访 时需特别小心，以确保为每个主机选择正确的 SAS 主机端口全球 ID
(World Wide ID)。如果配置了任何不正确的 ID，则会在“摘要”页上生成不准确的拓扑结果，
并标记 Topology Conflict （拓扑冲突）。可以使用
SAS 5/E HBA BIOS
设置程序以标识
SAS
主机
端口的全球
ID (World Wide ID)
。
注：在使用 Automatic （自动） 选项添加新的存储设备前，请验证主机或管理站网络配置（包括站 IP 地址、
子网掩码和默认网关）是否正确。对于所有 Linux 内核，在使用
Automatic （自动） 选项添加存储设备时，
请设置默认网关，以便将广播信息包发送到 255.255.255.255。
Configure Host Access (Automatic)
（配置主机访问
[
自动
]
）界面已更改。要确保只添加指定的主机，
请从
Available Hosts
（可用主机）窗口中选择要配置的主机，然后单击
Add
（添加），将所选主机移
到
Selected Hosts
（所选主机）窗口中。在将指定的主机添加到
Selected Hosts
（所选主机）窗口后，
单击
OK
（确定）以运行自动主机配置。
MD3000 维护注意事项
对于
Linux
内核，在执行以下两种维护任务之一后，请停止并重新启动
SMagent
：
•脱机移动控制器或更换控制器。
•从Red Hat Enterprise Linux（第 4 版）、SUSE Linux Enterprise Server 9 或SUSE Linux
Enterprise Server 10 主机服务器卸下 SAS 电缆，或将 SAS 电缆连接到这些主机服务器。
注：如果重新启动 SMagent 后，在“摘要”页的“状态”门户中显示 Resolve Topology Conflicts（解决拓
扑冲突）消息，则可能需要重新启动主机服务器以清除此信息。而不应选择 Topology Conflict（拓扑冲突）
向导中的 Resolve（解决）选项。MD3000 将继续处理 I/O 请求，但如果只使用带内管理，则阵列可能变为
不完全管理方式。在维护操作中涉及的任何主机服务器上停止所有 I/O 操作，然后执行重新启动。

www.dell.com | support.dell.com
Systèmes Dell™
PowerVault™ MD3000
Mise à jour des informations

Remarques, avis et précautions
REMARQUE : une REMARQUE indique des informations importantes qui peuvent vous aider à mieux utiliser
votre ordinateur.
AVIS : un AVIS vous avertit d'un risque de dommage matériel ou de perte de données et vous indique comment éviter
le problème.
PRÉCAUTION : une PRÉCAUTION indique un risque potentiel d'endommagement du matériel, de blessure corporelle
ou de mort.
____________________
Les informations contenues dans ce document peuvent être modifiées sans préavis.
© 2006 Dell Inc. Tous droits réservés.
La reproduction de ce document de quelque manière que ce soit sans l'autorisation écrite de Dell Inc. est strictement interdite.
Marques utilisées dans ce document : Dell, le logo DELL, OpenManage et PowerVault sont des marques de Dell Inc. ; Microsoft, Windows et
DirectX sont des marques ou des marques déposées de Microsoft Corporation aux États-Unis et/ou dans d'autres pays ; SUSE est une marque
déposée de Novell, Inc. aux États-Unis et/ou dans d'autres pays ; Red Hat est une marque déposée de Red Hat, Inc.
Tous les autres noms de marques et marques commerciales utilisés dans ce document se rapportent aux sociétés propriétaires de ces marques et de
ces noms ou à leurs produits. Dell Inc. décline tout intérêt dans l'utilisation des marques déposées et des noms de marques ne lui appartenant pas.
Novembre 2006 P/N FF455 Rev. A01