Google

Hacmp 5.1 的配置方法(1)

2007-06-26 21:12 来源: jackworkonline.cublog.cn 作者:jackworkonline 网友评论 0 条 浏览次数 190
 
Hacmp 自从出了5.1 版本后, 很多客户向IBM技术支持中心打电话咨询配置方法。
的确,hacmp5.1 较以前的版本无论是从界面还是从配置方法都有了很大的改变。具体
内容我们可以从IBM hacmp5。1 红皮书上获知。但对于大部分hacmp 软件使用者来说,
最重要的是要有一个快速入门的步骤。因此我会在这里介绍一下hacmp 的入门安装与
配置。
一,Hacmp 的软件安装。
1.    安装前提
如果您的操作系统是AIX5.1 ,那么您要安装维护补丁包ML03以上,目前
最高的补丁版本是ML05. 并且您还要安装RSCT 2.2.1.30 或更高版本。
以下的包也是必须要安装的:
 
• bos.adt.lib
• bos.adt.libm
• bos.adt.syscalls
• bos.net.tcp.client
• bos.net.tcp.server
• bos.rte.SRC
• bos.rte.libc
• bos.rte.libcfg
• bos.rte.libcur
• bos.rte.libpthreads
• bos.rte.odm
     如果您要安装并行的资源组,还要安装下面的包:
• bos.rte.lvm.rte5.1.0.25 or higher
• bos.clvm.enh.
2.     开始安装
一般基本上除了haview ,netwiew ( Tivoli),的 包以外,所有的hacmp的包都要安装。
3.     打补丁。
 注意,客户总是忽略给hacmp打补丁这一步骤。其实对hacmp来说,补丁是十分重要的。很多
 发现的缺陷都已经在补丁中被解决了。有的客户严格的按照正确步骤安装和配置完hacmp的软件后,发现takeover 有问题,ip接管有问题,机器自动宕机等等千奇百怪的问题,其实都与补丁
有关。所以客户一定要注意打补丁这个环节。现在hacmp最新的补丁是:
IY53044 - Latest HACMP for AIX R510 Fixes as of January 2004
大家可以从IBM网站上下载,或者打800-810-1818 热线电话索要。
4.     重启机器。 在hacmp 5。1 中 为了安全起见,不再使用/.rhosts 文件来控制两台机器
之间的命令和数据交换,而是引进的一个新的进程clcomd 。 如果你编辑/etc/inittab文件
就会发现安装完hacmp后,在最后添加了一行:clcomdES:2:once:startsrc -s clcomdES >/dev/console 2>&1 。因此重新启机后, ps –ef |grep clomd ,会发现:root 12908  6478   0   Apr 12      -  0:21 /usr/es/sbin/cluster/clcomd –d ,证明该进程启动了。Hacmp5.1使用
/usr/es/sbin/cluster/etc/rhosts 文件来代替 /.rhosts 文件的功能。
 
注意:如果两个节点间的通讯发生了什么问题,可以检查rhots 文件,或者编辑rhosts文件
加入两个节点的网络信息。
二, hacmp5.1 的配置
我们以两台机器为例:test1 和test2 , 共享三块7133 硬盘。
1.    首先配置两台机器的ip 和vg , 以及/etc/hosts  和application 启动/停止脚本
 test1:/>netstat -in
Name  Mtu   Network     Address            Ipkts Ierrs    Opkts Oerrs  Coll
en0   1500  link#2      0.4.ac.49.f2.d5      77960     0    47805     0     0
en0   1500  100.1       100.1.0.1            77960     0    47805     0     0
en1   1500  link#3      0.6.29.ec.44.d6         33     0       11     0     0
en1   1500  192.168.0   192.168.0.1             33     0       11     0     0
 
test2:/>netstat -in
Name  Mtu   Network     Address            Ipkts Ierrs    Opkts Oerrs  Coll
en0   1500  link#2      0.4.ac.49.60.23      31138     0    82582     0     0
en0   1500  100.1       100.1.0.2            31138     0    82582     0     0
en1   1500  link#3      0.4.ac.3e.b9.4b         36     0       13     0     0
en1   1500  192.168.0   192.168.0.2             36     0       13     0     0
 
test1 :
hdisk0          0004383268b07574                    rootvg          active
hdisk3          000438325e22bca7                    test1vg
hdisk4          00043832125e5aa8                    None
hdisk5          000438323d0e4487                    None
 
test2 :
hdisk0          000d29574085126d                    rootvg          active
hdisk5          000438325e22bca7                    test1vg
hdisk6          00043832125e5aa8                    None
hdisk7          000438323d0e4487                    None
 
/etc/hosts 
100.1.0.2       test2_boot1  test2
100.1.0.1       test1_boot1  test1
192.168.0.1     test1_boot2
192.168.0.2     test2_boot2
10.1.0.1        test1_svc
10.1.0.2        test2_svc
10.1.0.5        test1_per
10.1.0.6        test1_per
 
test2:/ha51>ls –l
-rwxr-xr-x   1 root     system           65 Apr 13 13:51 start
-rw-r--r--   1 root     system           31 Apr 13 11:49 start.log
-rwxr-xr-x   1 root     system           66 Apr 13 14:01 start1
-rw-r--r--   1 root     system           31 Apr 13 14:01 start1.log
-rwxrwxrwx   1 root     system           64 Apr 13 11:48 stop
-rw-r--r--   1 root     system           31 Apr 13 11:48 stop.log
-rwxr-xr-x   1 root     system           66 Apr 13 14:01 stop1
-rw-r--r--   1 root     system           31 Apr 13 14:01 stop1.log
 
vi start
date >> /ha51/start.log
banner " start app1 " >> /tmp/hacmp.out
 
vi stop
date >> /ha51/stop.log
banner "stop app1 " >> /tmp/hacmp.out
 
vi start1
date >> /ha51/start1.log
banner " start app2 " >> /tmp/hacmp.out
 
vi stop1
date >> /ha51/stop1.log
 banner "stop app2 " >> /tmp/hacmp.out
 
注意:在两个节点要保证hosts 和 启动/停止脚本要一样存在。
 
2. 用smitty hacmp 来配置hacmp
 
添加cluster 和node
smitty hacmp
Initialization and Standard Configuration
  Extended Configuration
  System Management (C-SPOC)
  Problem Determination Tools
 
Add Nodes to an HACMP Cluster
  Configure Resources to Make Highly Available
  Configure HACMP Resource Groups
  Verify and Synchronize HACMP Configuration
Display HACMP Configuration
 
                                                       
* Cluster Name                                       [ha51tsc]
 New Nodes (via selected communication paths)[m       [test2_boot1 test1_boot1]                                            
 Currently Configured Node(s)
 
这一部很重要,一般我们都是把每个节点的boot1 作为communication path . New node
可以一起加,也可以一个一个的加。当回车以后,系统会自己discover hacmp 的资源
显示如下:
。。。。。。。。。。。。。。。。。。。。
IP Network Discovery completed normally
Current cluster configuration:
 
No resource groups defined
Cluster Description of Cluster: ha51tsc
Cluster Security Level: Standard
There are 2 node(s) and 1 network(s) defined
 
NODE test1:
        Network net_ether_02
                test1_boot1     100.1.0.1
                test1_boot2     192.168.0.1
 
NODE test2:
        Network net_ether_02
                test2_boot1     100.1.0.2
                test2_boot2     192.168.0.2
。。。。。。。。。。。。。。。。。。。。。。
 
添加高可用资源(service ip , application server , vg and jfs
 
 添加服务ip地址
 Add Nodes to an HACMP Cluster
  Configure Resources to Make Highly Available
  Configure HACMP Resource Groups
  Verify and Synchronize HACMP Configuration
  Display HACMP Configuration
 
 
  Configure Service IP Labels/Addresses
  Configure Application Servers
  Configure Volume Groups, Logical Volumes and Filesystems
  Configure Concurrent Volume Groups and Logical Volumes
 
  Add a Service IP Label/Address
  Change/Show a Service IP Label/Address
  Remove Service IP Label(s)/Address(es)
 
* IP Label/Address                                   [test1_svc   ]                                                          
Network Name                                    [net_ether_02 ]
 
* IP Label/Address                                   [test2_svc   ]                                                          
Network Name                                    [net_ether_02 ]
 
 
 
添加application server
 Configure Service IP Labels/Addresses
  Configure Application Servers
  Configure Volume Groups, Logical Volumes and Filesystems
  Configure Concurrent Volume Groups and Logical Volumes
 
 
  Add an Application Server
  Change/Show an Application Server
  Remove an Application Server
 
* Server Name                                        [app1]
* Start Script                                       [/ha51/start]
* Stop Script                                     [/ha51/stop]
 
* Server Name                                        [app2]
* Start Script                                       [/ha51/start1]
* Stop Script                                     [/ha51/stop1]
 
添加共享vg ,  jfs
注意在前面的步骤中我们看到已经有一个共享VG test1vg 存在了,它使用传统的方法:
1.  在test1 节点上创建test1vg , lv,jfs
2.  Varyoffvg
3.  在test2 上 importvg
4.  Varyoffvg
现在我们试着用hacmp的功能去创建test2vg
 
Configure Service IP Labels/Addresses
 Configure Application Servers
 Configure Volume Groups, Logical Volumes and Filesystems
  Configure Concurrent Volume Groups and Logical Volumes
 
 
  Shared Volume Groups
  Shared Logical Volumes
  Shared File Systems
  Synchronize Shared LVM Mirrors
  Synchronize a Shared Volume Group Definition
 
 
  List All Shared Volume Groups
  Create a Shared Volume Group
  Create a Shared Volume Group with Data Path Devices
  Set Characteristics of a Shared Volume Group
  Import a Shared Volume Group
  Mirror a Shared Volume Group
  Unmirror a Shared Volume Group
 
 在选择菜单中同时用F7 选择test1 和 test2
Ø       test1
Ø       test2
选中PVID 00043832125e5aa8
  Node Names                                          test1,test2
  PVID                                                00043832125e5aa8
 VOLUME GROUP name                                [test2vg ]
  Physical partition SIZE in megabytes             4                                                                  
Volume group MAJOR NUMBER                          [49]
 
test2:/ha51>lspv
hdisk0          000d29574085126d                    rootvg          active
hdisk5          000438325e22bca7                    test1vg
hdisk6          00043832125e5aa8                    test2vg
hdisk7          000438323d0e4487                    None
 
test1:/ha51>lspv
hdisk0          0004383268b07574                    rootvg          active
hdisk3          000438325e22bca7                    test1vg
hdisk4          00043832125e5aa8                    test2vg
hdisk5          000438323d0e4487                    None
 
同样方法你可以在两个节点上同时创建ljfs
Shared Volume Groups
  Shared Logical Volumes
  Shared File Systems
  Synchronize Shared LVM Mirrors
Synchronize a Shared Volume Group Definition
 
 
  Journaled File Systems
  Enhanced Journaled File Systems
 
  Add a Journaled File System
  Add a Journaled File System on a Previously Defined Logical Volume
  List All Shared File Systems
  Change / Show Characteristics of a Shared File System
  Remove a Shared File System
 
  Add a Standard Journaled File System
  Add a Compressed Journaled File System
  Add a Large File Enabled Journaled File System
 
test1vg             test1,test2
test2vg             test1,test2
 
Node Names                                     test1,test2
 Volume group name                                test1vg
* SIZE of file system                                [10 ]
* MOUNT POINT                                   [/test1jfs]
  PERMISSIONS                                   read/write                                                          
  Mount OPTIONS                                  []                                                                   
  Start Disk Accounting?                              no                                                                  
  Fragment Size (bytes)                               4096                                                                
  Number of bytes per inode                           4096                                                                
  Allocation Group Size (MBytes)                       8
 
系统会自动在test1上添加test1jfs 文件系统,并且自动会在两个
节点上作update . 但是根据我自己的经验,最好还是用传统的方式在
一个结点上创建vg ,lv, jfs . 然后再import 到另一个节点上。
这里有一个tips , 如果在这里创建共享jfs 遇到问题,可以先手工把vg
在一个结点上varyon , 然后再创建就可以了。
 
创建资源组
 
 
  Initialization and Standard Configuration
  Extended Configuration
  System Management (C-SPOC)
  Problem Determination Tools
 
Add Nodes to an HACMP Cluster
  Configure Resources to Make Highly Available
  Configure HACMP Resource Groups
  Verify and Synchronize HACMP Configuration
  Display HACMP Configuration
 
Add a Resource Group
  Change/Show a Resource Group
  Remove a Resource Group
  Change/Show Resources for a Resource Group (standard)
 
 
Cascading
Rotating                                                               Concurrent                                                         Custom
 
                                                  
* Resource Group Name                               [res1]
* Participating Node Names / Default Node Priority         [test1 test2]
 
同样方法可以添加res2 
 
接下来可以配置资源组,当然也可以在Extended Configuration 中去详细配置。
我们姑且先在Initialization and Standard Configuration 中配置。
 
Smitty   cm_config_hacmp_resource_groups_menu_dmn
 
Add a Resource Group
  Change/Show a Resource Group
  Remove a Resource Group
Change/Show Resources for a Resource Group (standard)
 
选择res1
 
Resource Group Name                                 res1
  Participating Node Names (Default Node Priority)    test1 test2
 
* Service IP Labels/Addresses                        [test1_svc]                                                        Volume Groups                                      [mtest1vg]                                                            Filesystems (empty is ALL for VGs specified)                    [/test1jfs]                                                          Application Servers                                         [mapp1]
 
同样的方法配置res2
 
检查和同步hacmp配置
 
Initialization and Standard Configuration
  Extended Configuration
  System Management (C-SPOC)
  Problem Determination Tools
 
 
 
 
 
Add Nodes to an HACMP Cluster
  Configure Resources to Make Highly Available
  Configure HACMP Resource Groups
  Verify and Synchronize HACMP Configuration
  Display HACMP Configuration
 
 
Cluster Description of Cluster: ha51tsc
Cluster Security Level: Standard
There are 2 node(s) and 1 network(s) defined
 
NODE test1:
        Network net_ether_02
                test2_svc       10.1.0.2
                test1_svc       10.1.0.1
                test1_boot2     192.168.0.1
                test1_boot1     100.1.0.1
 
NODE test2:
        Network net_ether_02
                test2_svc       10.1.0.2
                test1_svc       10.1.0.1
                test2_boot1     100.1.0.2
                test2_boot2     192.168.0.2
 
Resource Group res1
        Behavior                 cascading
        Participating Nodes      test1 test2
        Service IP Label                 test1_svc
 
Resource Group res2
        Behavior                 cascading
        Participating Nodes      test2 test1
        Service IP Label                 test2_svc
 
注意nodetest1 的ip 地址排列,虽然test_boot2 排在test_boot1
前面,但是实验证明,service 地址依然会绑定在communication
path 上。
 
现在就可以做Initialization and Standard Configuration 〉〉  Verify and Synchronize HACMP Configuration  .
然后start 一下hacmp ,看看take over 是否都正常。
 
注意,很多客户是把所有的hacmp 包括应用都配好后再试起hacmp
, 作takeover 测试,这是很不好的一种习惯。因为融入的可能因素太多
了,一旦有了问题,我们还要隔离问题,先把hacmp配置简化,再一步步
作pd , 那么先前的配置就白做了。所以建议客户阶段性的监测一下hacmp
 
 
3. 到此为止我们的hacmp已经基本配置完成了。剩下的要在Extended Configuration中配置了。
在Extended Configuration中我们还可以配置 tty 心跳 网络  , hdisk 心跳网络,Persistent Node IP ,application monitor 等等。
 
下面我们先介绍一下配置hdisk 心跳网络,这也是hacmp5.1 里的一个新的功能。
 
首先我们要一个 Enhanced concurrent VG , 这个vg 不需要一定是放在concurrent 资源组里的vg , 当然也可以用concurrent 资源组里的硬盘来做
心跳网络。这个concurrent vg 可以通过传统方法建立。
1. Mkvg –c convg 
2. 在一个节点上varyoffvg , 另一个节点上importvg
 
现在我们介绍用hacmp 来创建concurrent vg .
 
 
  Initialization and Standard Configuration 
  Extended Configuration
  System Management (C-SPOC)
Problem Determination Tools
 
 
  Add Nodes to an HACMP Cluster
  Configure Resources to Make Highly Available
  Configure HACMP Resource Groups
  Verify and Synchronize HACMP Configuration
  Display HACMP Configuration
 
Configure Service IP Labels/Addresses
  Configure Application Servers
  Configure Volume Groups, Logical Volumes and Filesystems
  Configure Concurrent Volume Groups and Logical Volumes
 
 
   Concurrent Volume Groups
   Concurrent Logical Volumes
   Synchronize Concurrent LVM Mirrors
 
 
  List All Concurrent Volume Groups
  Create a Concurrent Volume Group
  Create a Concurrent Volume Group with Data Path Devices
  Set Characteristics of a Concurrent Volume Group
  Import a Concurrent Volume Group
  Mirror a Concurrent Volume Group
  Unmirror a Concurrent Volume Group
 
  选中test1 and test2
  选中共享硬盘
 
Node Names                                          test1,test2
      PVID                                                000438323d0e4487
      VOLUME GROUP name                                   convg
      PhysicalpartitionSIZEinmegabytes                         4                                      Volume group MAJOR NUMBER                          [49]                                                                 Enhanced Concurrent Mode                                   true
 
下面看一下两个节点的硬盘状况:
 
test1:
hdisk0          0004383268b07574               rootvg     active
hdisk3          000438325e22bca7               test1vg
hdisk4          00043832125e5aa8                test2vg
hdisk5          000438323d0e4487               convg
test2 :
hdisk0          000d29574085126d               rootvg          active
hdisk5          000438325e22bca7                test1vg
hdisk6          00043832125e5aa8                test2vg
hdisk7          000438323d0e4487               convg
 
现在检查hdisk 网络的状况,在一个节点上向hdisk 写数据,从另一个节点
上读数据,很像 我们在配置tty 网络之前,检查一下tty 是否连通。
 
注意:我原来的操作系统是aix5.2 01 , 安装了hacmp5.1 打了最新的补丁。
但是在/usr/sbin/rsct/bin 下找不到dhb_read 命令。 它应该是属于rsct 的,后来我把aix5.2 打倒ml02 , rsct 所有的包都生级了,reboot 机器后,找到
了dhb_read 命令。
 
1.  Add  /usr/sbin/rsct/bin/ to /etc/environment 里的path 中
2.  重新login test1 和test2 使path 生效
3. 在test1 上运行:dhb_read -p hdisk5 –r
4. 在test2 上运行:dhb_read -p hdisk7 –t
 在test1 上:test1:/>dhb_read -p hdisk5 -r
Receive Mode:
Waiting for response . . .
Link operating normally
                在test2上:
test2:/usr/sbin/rsct/bin>dhb_read -p hdisk7 -t
Transmit Mode:
Detected remote utility in receive mode.  Waiting for response . . .
Link operating normally
 
证明通讯正常。
 
添加hdisk heart beat 网络和设备
 
Initialization and Standard Configuration
  Extended Configuration
  System Management (C-SPOC)
  Problem Determination Tools
 
Discover HACMP-related Information from Configured Nodes
  Extended Topology Configuration
  Extended Resource Configuration
  Extended Event Configuration
  Extended Performance Tuning Parameters Configuration
  Security and Users Configuration
Snapshot Configuration
Extended Verification and Synchronization
 
Configure an HACMP Cluster
  Configure HACMP Nodes
  Configure HACMP Sites
  Configure HACMP Networks
  Configure HACMP Communication Interfaces/Devices
  Configure HACMP Persistent Node IP Label/Addresses
  Configure HACMP Global Networks
  Configure HACMP Network Modules
  Configure Topology Services and Group Services
  Show HACMP Topology
 
Add a Network to the HACMP Cluster
  Change/Show a Network in the HACMP Cluster
  Remove a Network from the HACMP Cluster
 
# Pre-defined Serial Device Types                                      
diskhb                                                               
rs232                                                                 
tmscsi                                                             
tmssa
                            
* Network Name[m                                       [net_diskhb_01]
* Network Type                                            diskhb
 
上一篇: 下一篇:

相关主题:hacmp  hacmp5.1

网友评论