“十一”前最后一个周末了,照例送资料。
还没想好过节送什么福利,有需要大家也可以提~~
目录
1 文档版本
2 DS8100安装说明
3 DS8100物理安装
3.1 检查并使用最新的安装指南
3.2 检查并确认机器供电情况
3.3 检查并确认工作计划表填写完毕
3.4 安装电源线
3.5 检查SWITCH设定
3.6 DS8000上电
3.7 管理控制台配置
3.8 上电并检验Storage Facility
3.9 升级DS8000微码
3.10 连接光纤电缆
3.11 备份管理控制台的配置信息
4 DS8100逻辑配置
4.1 关于虚拟化存储的几个重要概念
4.2 安装DSCLI
4.3 逻辑配置
4.3.1 创建Array
4.3.2 创建Rank
4.3.3 创建Extendpool
4.3.4 将Rank加入到Extendpool
4.3.5 创建Volume
4.3.6 创建Volume Group
4.3.7 将Volume加入到Volume Group中
4.3.8 建立host connect
4.3.9 主机端识别硬盘
5 DS8000简单维护
5.1 DS8000下电步骤
5.2 停止DS8000对外服务
5.3 检查 LR/ZLR开关设置
5.4 关掉2107 RACK 1(RACK不会断电)
5.5 RACK 下电
5.6 扩展柜下电
5.7 关闭机器电源模块供电开关
5.8 打开机器电源模块供电开关
5.9 打开紧急电源开关
5.10 HMC 上电
5.11 Power on SF
5.12 检查确认工作正常
5.13 DS8000故障检查
5.14 DS8000数据收集(PE Package)
1 文档版本
(略)
2 DS8100安装说明
本安装指导作为XX项目的配套文档,旨在为XX系统建设提供技术指导。本文档以DS8100系统安装及逻辑配置为主要内容。
3 DS8100物理安装
3.1 检查并使用最新的安装指南
(略)
3.2 检查并确认机器供电情况
检查并确认机器供电的电源电压正常,符合安装要求
3.3 检查并确认工作计划表填写完毕
检查并确认工作计划表填写完毕,该系列表格在DS8000安装将被使用。
3.4 安装电源线
实施此步骤前,请确认空气开关柜的380V供电电源开关处于关闭状态。
拆除2个螺栓(下图标号1),转动到208VDC bus bar gate 的service position (下图标号2);
拆除下图中两个螺栓2,拧松顶部的螺栓1,将后挡板拆除。
按照下图,将380V电源线连接到电源模块上。拆除下图中两个螺栓2,拧松顶部的螺栓1,将后挡板拆除。
1:longer cable
2:shorter cable
3:bracket
4:208VDC bus bar gate
3.5 检查SWITCH设定
参照IBM TotalStorage DS8000: Service Guide SY27-7638-01 Page 50 “Checking the 2107 Rack switch settings”检查DS8000的所有SWITCH设置正确。
3.6 DS8000上电
按照下图,确认Local/Remote开关置于Remote,Battery电源开关CB0/CB1置于ON,打开380V主电源开关CB00,将DS8000主机柜上电,并观察有无异常现象,如无异常,可以进入下一步配置。
3.7 管理控制台配置
管理控制台(HMC)如下图,当DS8000机柜上电后,该HMC将自动上电,所有对DS8000的配置及管理将由HMC完成。
拉出HMC的液晶显示器,打开其电源,在提示的界面上键入1,将该HMC做为主控制台;稍后HMC提示登陆用户及口令,HMC缺省的用户及口令如下:
ID:CE #一般操作使用,可以完成绝大多数任务
Password:serv1cece
ID:hscroot #HMC的root用户,为IBM工程师使用
Password:abc123
使用《DS8100安装计划书》中的附录A:《公司信息工作表》,附录B:《管理控制台网络设置工作表》,附录C:《远程支持工作表》按照下表步骤对HMC进行客户化配置,配置完成后,退出HMC,重新启动HMC,使新配置生效。
按照下图所示,将客户提供的用以连接internet的网线、电话线,连接到HMC的相应端口:
3.8 上电并检验Storage Facility
按照下列步骤Powering on and verifying the storage facility,该步骤将耗时1.5小时左右。
Display the storage facility Install State.
a. In the HMC Navigation area, open Management Environment, select and open the HMC.
b. Open Service Applications → Service Focal Point.
c. In the right content area, select Install/Add/Remove Hardware. A window displays a list of storage facilities.
d. Observe the displayed Install State for the 2107 being installed.
e. Is the Machine Type and Serial Number for the 2107 being installed listed on the panel as “Ready for Install” or “New”?
Yes, select the 2107 to be installed. Go to the next step.
No, go to MAP1400 Repair a problem found during install storage facility process.
f. Click on Selected, and then from the drop-down select Storage Facility Field Install.
g. Follow the on-screen instructions to perform DDM certify
3.9 升级DS8000微码
此次安装将升级DS8000的微码至版本6.1.600.59。
3.10 连接光纤电缆
按照下图所示,将光纤电缆做好标记后,按照《SAN系统设计规划书》中连接要求,将SAN交换机与DS8000的相应端口连接到一起。
3.11 备份管理控制台的配置信息
按照下面的步骤,备份管理控制台的重要配置信息:
Perform the following steps to back up critical console data.
a. In the HMC Navigation area, open Management Environment, select and open the HMC.
b. Open Licensed Internal Code → HMC Code Update.
c. In the right content area, select Back up Critical Console Data.
d. Insert a formatted DVD-RAM media into the drive.
e. Select Backup to store your critical console data on the DVD-RAM.
f. Click OK.
备份完毕后,贴好标签,请客户保存,备用。
4 DS8100逻辑配置
4.1 关于虚拟化存储的几个重要概念
4.2 安装DSCLI
为提高配置的速度,本次划分DS8000采用DSCLI(DS Command Line Interface),在笔记本上安装DSCLI后,修改其C:\Program Files\IBM\dscli\profile\ dscli.profile,如下:
# Management Console/Node IP Address(es)
# hmc1 and hmc2 are equivalent to -hmc1 and -hmc2 command
# options.
hmc1: 172.16.0.1
#hmc2: 127.0.0.1
#
# Default target Storage Image ID
# "devid" and "remotedevid" are equivalent to
# "-dev storage_image_ID" and "-remotedev storeage_image_ID"
# command options, respectively.
devid: IBM.2107-75ABAX0
其中hmc1为DS8000的HMC地址,此处为HMC缺省IP172.16.0.1; devid为DS8000的storage facility image ID,对应ERP项目中的DS8000,为IBM.2107-75ABAX1(DS8000-1)。
安装后,创建了下面的具有admin权限的用户,在今后的配置中,我们将使用该用户,以保护admin用户:
ID:admin
Password:(此处故意隐藏)
Notes:请注意不要在24小时内3次输入错误的用户密码,3次的错误输入将导致该用户的帐户被锁。
4.3 逻辑配置
4.3.1 创建Array
项目中的DS8100配置为16组128块硬盘,因此需要创建16个array,所有array均为RAID5
dscli> mkarray -raidtype (5 or 10) -arsite S1
成功后结果如下:
4.3.2 创建Rank
根据DS8100的配置,创建了16个array,基于此,创建16个Rank:
dscli>mkrank -array A0 -stgtype fb
dscli>mkrank -array A1 -stgtype fb
dscli>mkrank -array A2 -stgtype fb
dscli>mkrank -array … -stgtype fb
dscli>mkrank -array A16 -stgtype fb
成功后结果如下:
dscli> lsrank
Date/Time: May 24, 2006 5:13:42 PM CST IBM DSCLI Version: 5.0.6.240 DS: IBM.2107-75ABAX1
ID Group State datastate Array RAIDtype extpoolID stgtype
===========================================================
R0 0 Normal Normal A0 10 P0 fb
R1 1 Normal Normal A1 10 P1 fb
R2 0 Normal Normal A2 10 P2 fb
R3 1 Normal Normal A3 10 P3 fb
R4 0 Normal Normal A4 10 P4 fb
R5 1 Normal Normal A5 10 P5 fb
R6 0 Normal Normal A6 10 P6 fb
R7 1 Normal Normal A7 10 P7 fb
R8 0 Normal Normal A8 10 P8 fb
R9 1 Normal Normal A9 10 P9 fb
R10 0 Normal Normal A10 10 P10 fb
R11 1 Normal Normal A11 10 P11 fb
R12 0 Normal Normal A12 10 P12 fb
R13 1 Normal Normal A13 10 P13 fb
R14 0 Normal Normal A14 10 P14 fb
R15 1 Normal Normal A15 10 P15 fb
R16 0 Normal Normal A16 5 P16 fb
4.3.3 创建Extendpool
创建16个Extendpool:
dscli>mkextpool -rankgrp 0 -stgtype fb P0
dscli>mkextpool -rankgrp 1 -stgtype fb P1
dscli>mkextpool -rankgrp 0 -stgtype fb P2
dscli>mkextpool -rankgrp 1 -stgtype fb P3
dscli>mkextpool -rankgrp 0 -stgtype fb P4
dscli>mkextpool -rankgrp 1 -stgtype fb P5
成功后结果如下:
4.3.4 将Rank加入到Extendpool
dscli>chrank -extpool P0 R0
dscli>chrank -extpool P1 R1
dscli>chrank -extpool P2 R2
dscli>chrank -extpool P3 R3
dscli>chrank -extpool P4 R4
dscli>chrank -extpool P5 R5
……
成功后结果如下:
dscli> lsrank
Date/Time: May 24, 2006 5:13:42 PM CST IBM DSCLI Version: 5.0.6.240 DS: IBM.2107-75ABAX1
ID Group State datastate Array RAIDtype extpoolID stgtype
===========================================================
R0 0 Normal Normal A0 5 P0 fb
R1 1 Normal Normal A1 5 P1 fb
R2 0 Normal Normal A2 5 P2 fb
R3 1 Normal Normal A3 5 P3 fb
R4 0 Normal Normal A4 5 P4 fb
R5 1 Normal Normal A5 5 P5 fb
R6 0 Normal Normal A6 5 P6 fb
R7 1 Normal Normal A7 5 P7 fb
R8 0 Normal Normal A8 5 P8 fb
R9 1 Normal Normal A9 5 P9 fb
R10 0 Normal Normal A10 5 P10 fb
R11 1 Normal Normal A11 5 P11 fb
R12 0 Normal Normal A12 5 P12 fb
R13 1 Normal Normal A13 5 P13 fb
R14 0 Normal Normal A14 5 P14 fb
R15 1 Normal Normal A15 5 P15 fb
R16 0 Normal Normal A16 5 P16 fb
4.3.5 创建Volume
执行命令如下:
dscli>mkfbvol -extpool p0 -cap 100 -name fb_vol_2000
dscli>mkfbvol -extpool p0 -cap 93 -name fb_vol_2000
受Host Pair盘的分配影响,每个Extend Pool里的有效容量不是一致的。所以为了有效利用昂贵的磁盘空间,每一个Extend Pool里的逻辑磁盘大小是不一致的。
4.3.6 创建Volume Group
为方便管理,我们尽量将同种类型的volume放到统一个VG中,命令如下:
dscli>mkvolgrp -type scsimask VG1
dscli>mkvolgrp -type scsimask VG2
4.3.7 将Volume加入到Volume Group中
下图是DS8000的IO port端口图。
4.3.8 建立host connect
在pSeries主机上登陆到AIX,执行lscfg –vl fcsx 以显示所使用的光纤卡的WWPN,主机将通过定义好的host connect通道,经由选定的DS8000上的IO ports访问Volume Group中的磁盘,命令如下:
dscli> mkhostconnect -wwname 10000000C9428122 -profile "IBM pSeries - AIX" -volgrp v0 -ioport i0200,i0300,i0032,i0102 p595-2-fcs2
4.3.9 主机端识别硬盘
在pSeries主机上安装如下SDDPCM软件及DS8000磁盘驱动:
devices.sddpcm.53.2.1.1.1.bff
devices.fcp.disk.ibm.mpio.rte
devices.sddpcm.53.rte
将以上3个文件复制到操作系统任意目录,然后执行smitty installp命令按照菜单指示安装以上三个文件包,完成后重新启动主机后,分别在所有需要识别该DS8000规划出的磁盘的主机上运行cfgmgr后,便可以正常识别到所有的磁盘hdisk。
查询该主机对应DS8100所划分的LUN请参照以下命令
# pcmpath query essmap
磁盘hdisk2是操作系统识别DS8100的一个LUN, 它共有4个通路path0 – 4
该磁盘大小约为41G, 该磁盘所对应的DS8100的LUN ID 为2002,前边字符75DLDZ1为该存储的序列号,该磁盘通过主机端光纤卡fcs0 及fcs1 连接到ds8100的光纤端口0030,0100 和 0230,0300. 所以基于以上命令就可以找到主机相关磁盘与存储磁盘的对应关系,综合来说就是 DS8100中每8块磁盘一个RAID, 该RAID可定义为5或10,每个RANK即代表一个RAID, 每个RANK可以划分若干个LUN, 每个LUN有自己的LSS ID,将各个RANK中的LUN定义到一个LUN组中,将主机光纤卡与DS8100 I/O端口绑定到该LUN组中即完成 磁盘分布工作。
5 DS8000简单维护
说明:因DS8000为高端存储设备,为用户最核心的设备,我们建议用户打开DS8000的call home(回拨)功能,即当DS8000出现故障时,可以通过电话线或者internet(为保证安全,DS8000通过连接到IBM内部网络),向IBM公司报修,IBM接到设备报修后,会立即联系责任工程师,以保证您的设备在最短的时间内得到维护。
5.1 DS8000下电步骤
注意:切勿直接使用机器上的红色紧急开关,该开关仅用于紧急情况,如火灾等!直接按下该开关,讲可能导致用户数据丢失!
如果您在系统投入生产运行以后,想对系统彻底下电,如UPS维护,请提前联络IBM工程师!
5.2 停止DS8000对外服务
Quiesce SFI /SFI’s to destage customer data from cache to disk.
1) Select: Select: Service Applications
2) Select : Service Focal Point
3) Select: Service Utilities
4) System Selection and Action Invocation
Highlight the SF: ( 2107-9xx - - - )
Choose: Selected from the tool bar.
Select: Get Storage Facility Images
5) Select an SFI menu displays, Highlight: (2107-9xx - - -)
Select: Okay
6) Choose: Selected from the tool bar
Select: Change/Show SFI State
7) From SFI server menu Select: Quiesce SFI
Confirm the Quiesce Select: Yes
Confirm the Quiesce Started Select: OK
When this completes both LPAR’s should say quiesced.
If this fails call next level of support. Other actions could loose customer data
5.3 检查 LR/ZLR开关设置
该开关在正常情况下,不需检查其设置,用户可以忽略,直接进行下一步。
Open the rear left machine cover of the base rack in the subsystem.Ensure that the local/Remote switch on the LR/ZLR card is set to Remote.
For z/Series remote power shutdown follow local procedure and refer to the 2107 Service Guide for guidance.
5.4 关掉2107 RACK 1(RACK不会断电)
Power off 2107 Rack 1 - This will shutdown SF/ lpars safely.
This will quiesce the lpars.
1) In the management console (HMC) navigation area, open the Management Environment.
2) Select Service Applications
3) Select Service Focal Point
4) Select Service Utilities.
5) Select Machine type-model/serial. Click Selected.
6) From the top tool bar, select Manage Power Control.
Ensure Power Control Mode is set to Manual Power Mode.
Change Storage Facility Power State to Power Off Storage Facility.
Click OK.
Answer Yes on the confirm screen.
7) Allow up to 5 to 10 minutes for the storage facility to power off.
8) Verify that the 2107 is now powered off.
From the Service / WebSM GUI, select: Service and Partition, then Server
Management.
View the state of the two storage facility systems that you requested to be powered
off.
Do both of the storage facility systems show as 'Powered Off'.
9) Shutdown the HMC
From HMC Service GUI / WebSM: console -> exit -> exit now -> shutdown
5.5 RACK 下电
Once HMC is shutdown successfully you can press the RACK OP Panel Power (Red).
5.6 扩展柜下电
If other Rack(s) are installed, press the appropriate Op Panel Power switch(s) to OFF.
5.7 关闭机器电源模块供电开关
打开机器右后门,将电源模块上,黄色箭头所指的开关拨下
5.8 打开机器电源模块供电开关
打开机器右后门,将电源模块上,黄色箭头所指的开关拨到上部,并检查电源模块的LED灯,确认没有黄灯亮。
5.9 打开紧急电源开关
在机器前部的控制面板的红色开关上边,有一个直径约1mm的小孔,请用一根曲别针插入后,按下紧急开关,机柜开始上电。
如果有扩展柜安装,请等待几分钟使主机柜上电自检完毕,之后打开其余机柜的红色电源开关。
5.10 HMC 上电
按下HMC的电源开关,等待HMC上电结束,并使用CE/serv1cece登陆到DS8000管理控制台。
5.11 Power on SF
From Service GUI / WebSM:
Service Utilities
Mark the SF an press "selected" - Select the correct SF if more than one is
available.
Manage Power controls Mode
Confirm Power Mode is set to Manual
Local/Remote Switch setting is set to Remote
check Power on SF.
Apply, then press the 'OK' to exit the menu
等待约20~30分钟,视您的DS8000配置,进入下步骤检查启动是否结束。
5.12 检查确认工作正常
From WebSM,
1) Select: Service Applications
2) Select: Service Focal Point
3) Select: Service Utilities
Highlight the SF: (2107-9xx---)
Choose: “Selected” from the tool bar
4) Select: View Storage Facility State
All sections should report with passed in the status column.
Failed status can be viewed by highlighting the System Check Type and then
select details.
5.13 DS8000故障检查
当DS8000发生故障时,其控制面板上的橙黄色故障灯会点亮,如果您的DS8000使用了call home功能,IBM报修中心将与您联系。如果您的DS8000没有使用call home功能,请安装如下菜单,进行故障检查:
A. Select: Service Applications
B. Select : Service Focal Point
C. Select: Manage Serviceable Events
D. From: selection screen select OK
选择OPEN的问题,您将看到DS8000发生的故障。
5.14 DS8000数据收集(PE Package)
Select : Service Applications
Select : Service Focal Point
Select : Service Utilities
Select a "StorageFacility
Click on "Selected" pull-down menu, and click on "Get Storage Facility Images
Select one of the MTMS's,如下图:
Click on the "Selected" pull-down menu, and click on "Perform Data Collection On-Demand"
Click on 'OK' to start collection process. (Note that process can take 15 to 30+ minutes.)
When the collection from the LPARs and the HMC has completed, a list of "sub-packages" is displayed. Select one or more of these sub-packages, then click on Offload button. The screen allows you to repeat the selection/Offload process for the remaining sub-packages.如下图:
When the Offload button is pressed, the package files are moved to Virtual Retain on HMC. A PMH is being opened in Retain. The package files are Called Home....
Go back to "Service Focal Point" and select "Manage Serviceable Events" Task.
Go to "Reference Code" field and select BEB00010 refcode.
Select the problem:
Click on "Selected" pull-down menu, and click on "Manage Problem Data"
You have the option to " Call Home" again or "Save to DVD":
长按下图二维码关注
也可以直接搜索公众号名称“AIX专家俱乐部”或微信号“AIXChina”关注