RedHat EL5 安装Oracle 10g RAC之--CRS 安装

系统环境：

操作系统：RedHat EL5

Cluster： Oracle CRS 10.2.0.1.0

Oracle： Oracle 10.2.0.1.0

如图所示：RAC 系统架构

二、CRS 安装

Cluster Ready Service是Oracle 构建RAC,负责集群资源管理的软件，在搭建RAC中必须首先安装.

安装需采用图形化方式，以Oracle用户的身份安装(在node1上）：

注意：修改安装配置文件，增加redhat-5的支持

[oracle@node1 install]$ pwd

/home/oracle/cluster/install

[oracle@node1 install]$ ls

addLangs.sh images oneclick.properties oraparamsilent.ini response

addNode.sh lsnodes oraparam.ini resource unzip

[oracle@node1 install]$ vi oraparam.ini

[Certified Versions]

Linux=redhat-3,SuSE-9,redhat-4,redhat-5,UnitedLinux-1.0,asianux-1,asianux-2

[oracle@node1 cluster]$./runInstaller

欢迎界面

注意安装CRS的主目录，不能和Oracle软件的目录一致，需单独在另一个目录

[oracle@node1 ~]$ ls -l /u01

total 24

drwxr-xr-x 3 oracle oinstall 4096 May 5 17:04 app

drwxr-xr-x 36 oracle oinstall 4096 May 7 11:08 crs_1

drwx------ 2 oracle oinstall 16384 May 4 15:59 lost+found

[oracle@node1 ~]$

添加节点（如果主机间信任关系配置有问题，这里就无法发现node 2）

修改public 网卡属性（public 网卡用于和Client 通讯）

OCR必须采用RAW设备（Exteneral Redundancy只需一个RAW,安装后可以添加mirror)

VOTE DISK必须采用RAW设备（Exteneral Redundancy只需一个RAW,安装后可以添加多个raw构成冗余)

开始安装（并将安装软件传送到node2）

安装提示分别在两个节点按顺序执行script

node1：

<code>[root@node1 ~]# /u01/app/oracle/oraInventory/orainstRoot.sh </code>

<code>Changing permissions of /u01/app/oracle/oraInventory to </code><code>770.</code>

<code>Changing groupname of /u01/app/oracle/oraInventory to oinstall.</code>

<code>The execution of the script is complete</code>

node2：

<code>[root@node2 ~]# /u01/app/oracle/oraInventory/orainstRoot.sh </code>

[root@node1 ~]# /u01/crs_1/root.sh

<code>WARNING: directory </code><code>'/u01'</code> <code>is </code><code>not</code> <code>owned by root</code>

<code>Checking to see </code><code>if</code> <code>Oracle CRS stack is already configured</code>

<code>/etc/oracle does </code><code>not</code> <code>exist. Creating it now.</code>

<code>Setting the permissions </code><code>on</code> <code>OCR backup directory</code>

<code>Setting up NS directories</code>

<code>Oracle Cluster Registry configuration upgraded successfully</code>

<code>assigning default hostname node1 </code><code>for</code> <code>node </code><code>1.</code>

<code>assigning default hostname node2 </code><code>for</code> <code>node </code><code>2.</code>

<code>Successfully accumulated necessary OCR keys.</code>

<code>Using ports: CSS=</code><code>49895</code> <code>CRS=</code><code>49896</code> <code>EVMC=</code><code>49898</code> <code>and</code> <code>EVMR=</code><code>49897.</code>

<code>node <nodenumber>: <nodename> <</code><code>private</code> <code>interconnect name> <hostname></code>

<code>Creating OCR keys </code><code>for</code> <code>user </code><code>'root'</code><code>, privgrp </code><code>'root'</code><code>..</code>

<code>Operation successful.</code>

<code>Now formatting voting device: /dev/raw/raw2</code>

<code>Format of </code><code>1</code> <code>voting devices complete.</code>

<code>Startup will be queued to </code><code>init</code> <code>within </code><code>90</code> <code>seconds.</code>

<code>Adding daemons to inittab</code>

<code>Expecting the CRS daemons to be up within </code><code>600</code> <code>seconds.</code>

<code>CSS is active </code><code>on</code> <code>these nodes.</code>

<code>CSS is inactive </code><code>on</code> <code>these nodes.</code>

<code>Local node checking complete.</code>

<code>Run root.sh </code><code>on</code> <code>remaining nodes to start CRS daemons.</code>

node1执行成功！

[root@node2 ~]# /u01/crs_1/root.sh

<code>WARNING: directory </code><code>'/u01'</code> <code>is not owned by root</code>

<code>/etc/oracle does not exist. Creating it now.</code>

<code>Setting the permissions on OCR backup directory</code>

<code>clscfg: EXISTING configuration version 3 detected.</code>

<code>clscfg: version 3 is 10G Release 2.</code>

<code>assigning </code><code>default</code> <code>hostname node1 </code><code>for</code> <code>node 1.</code>

<code>assigning </code><code>default</code> <code>hostname node2 </code><code>for</code> <code>node 2.</code>

<code>Using ports: CSS=49895 CRS=49896 EVMC=49898 and EVMR=49897.</code>

<code>clscfg: Arguments check out successfully.</code>

<code>NO KEYS WERE WRITTEN. Supply -force parameter to override.</code>

<code>-force is destructive and will destroy any previous cluster</code>

<code>configuration.</code>

<code>Oracle Cluster Registry </code><code>for</code> <code>cluster has already been initialized</code>

<code>Startup will be queued to init within 90 seconds.</code>

<code>Expecting the CRS daemons to be up within 600 seconds.</code>

<code>CSS is active on these nodes.</code>

<code>CSS is active on all nodes.</code>

<code>Waiting </code><code>for</code> <code>the Oracle CRSD and EVMD to start</code>

<code>Oracle CRS stack installed and running under init(1M)</code>

<code>Running vipca(silent) </code><code>for</code> <code>configuring nodeapps</code>

/u01/crs_1/jdk/jre//bin/java: error while loading shared libraries: libpthread.so.0: cannot open shared object file: No such file or directory

出现以上错误，解决方法：

[root@node2 bin]# vi vipca

Linux) LD_LIBRARY_PATH=$ORACLE_HOME/lib:$ORACLE_HOME/srvm/lib:$LD_LIBRARY_PATH

export LD_LIBRARY_PATH

#Remove this workaround when the bug 3937317 is fixed

arch=`uname -m`

if [ "$arch" = "i686" -o "$arch" = "ia64" ]

then

LD_ASSUME_KERNEL=2.4.19

export LD_ASSUME_KERNEL

unset LD_ASSUME_KERNEL （添加此行信息）

#End workaround

[root@node2 bin]# vi srvctl

LD_ASSUME_KERNEL=2.4.19

export LD_ASSUME_KERNEL

unset LD_ASSUME_KERNEL（添加此行信息）

在node 2重新执行root.sh：

注意：root.sh只能执行一次，如果再次执行，需执行rootdelete.sh

[root@node2 bin]# /u01/crs_1/root.sh

WARNING: directory '/u01' is not owned by root

Checking to see if Oracle CRS stack is already configured

Oracle CRS stack is already configured and will be running under init(1M)

[root@node2 bin]# cd ../install

[root@node2 install]# ls

<code>cluster.ini install.incl rootaddnode.sbs rootdelete.sh templocal</code>

<code>cmdllroot.sh make.log rootconfig rootinstall</code>

<code>envVars.properties paramfile.crs rootdeinstall.sh rootlocaladd</code>

<code>install.excl preupdate.sh rootdeletenode.sh rootupgrade</code>

[root@node2 install]# ./rootdelete.sh

<code>CRS</code><code>-0210</code><code>: Could </code><code>not</code> <code>find resource </code><code>'ora.node2.LISTENER_NODE2.lsnr'</code><code>.</code>

<code>CRS</code><code>-0210</code><code>: Could </code><code>not</code> <code>find resource </code><code>'ora.node2.ons'</code><code>.</code>

<code>CRS</code><code>-0210</code><code>: Could </code><code>not</code> <code>find resource </code><code>'ora.node2.vip'</code><code>.</code>

<code>CRS</code><code>-0210</code><code>: Could </code><code>not</code> <code>find resource </code><code>'ora.node2.gsd'</code><code>.</code>

<code>Shutting down Oracle Cluster Ready Services (CRS):</code>

<code>Stopping resources.</code>

<code>Successfully stopped CRS resources </code>

<code>Stopping CSSD.</code>

<code>Shutting down CSS daemon.</code>

<code>Shutdown request successfully issued.</code>

<code>Shutdown has begun. The daemons should exit soon.</code>

<code>Checking to see </code><code>if</code> <code>Oracle CRS stack is down...</code>

<code>Oracle CRS stack is </code><code>not</code> <code>running.</code>

<code>Oracle CRS stack is down now.</code>

<code>Removing script </code><code>for</code> <code>Oracle Cluster Ready services</code>

<code>Updating ocr file </code><code>for</code> <code>downgrade</code>

<code>Cleaning up SCR settings </code><code>in</code> <code>'/etc/oracle/scls_scr'</code>

[root@node2 install]#

node 2 再次出错：

[root@node2 install]# /u01/crs_1/root.sh

<code>clscfg: EXISTING configuration version </code><code>3</code> <code>detected.</code>

<code>clscfg: version </code><code>3</code> <code>is 10G Release </code><code>2.</code>

<code>NO KEYS WERE WRITTEN. Supply -force parameter to </code><code>override</code><code>.</code>

<code>-force is destructive </code><code>and</code> <code>will destroy any previous cluster</code>

<code>CSS is active </code><code>on</code> <code>all nodes.</code>

<code>Waiting </code><code>for</code> <code>the Oracle CRSD </code><code>and</code> <code>EVMD to start</code>

<code>Oracle CRS stack installed </code><code>and</code> <code>running under </code><code>init</code><code>(1M)</code>

Error 0(Native: listNetInterfaces:[3])

[Error 0(Native: listNetInterfaces:[3])]

解决方法：（配置网络）

[root@node2 bin]# ./oifcfg iflist

eth0 192.168.8.0

eth1 10.10.10.0

[root@node2 bin]# ./oifcfg getif

[root@node2 bin]# ./oifcfg setif -global eth0/192.168.8.0:public

[root@node2 bin]# ./oifcfg setif -global eth1/10.10.10.0:cluster_interconnect

eth0 192.168.8.0 global public

eth1 10.10.10.0 global cluster_interconnect

并在node2上执行VIPCA:

以root身份执行vipca（在/u01/crs_1/bin)

配置信息应和/etc/hosts文件一致

开始配置

vipca配置成功后，crs服务正常工作

安装完成！

验证CRS:

[root@node2 bin]# crs_stat -t

<code>Name Type Target State Host </code>

<code>ora.node1.gsd application ONLINE ONLINE node1 </code>

<code>ora.node1.ons application ONLINE ONLINE node1 </code>

<code>ora.node1.vip application ONLINE ONLINE node1 </code>

<code>ora.node2.gsd application ONLINE ONLINE node2 </code>

<code>ora.node2.ons application ONLINE ONLINE node2 </code>

<code>ora.node2.vip application ONLINE ONLINE node2</code>

[root@node1 ~]# crs_stat -t

附：错误案例

如果在运行root.sh时出现以下错误：

在出现错误的节点上运行（root）vipca 解决！

@至此CRS安装成功！

本文转自客居天涯 51CTO博客，原文链接：http://blog.51cto.com/tiany/1408023，如需转载请自行联系原作者

RedHat EL5 安装Oracle 10g RAC之--CRS 安装

继续阅读

neo4j之cypher使用文档

GitHub连夜封杀！这份阿里 10W 字内部 Java 字面试手册到底有多强？

spark/scala关于【资源文件】加载方法概述外部文件加载方案测试资源文件打包入jar包中小结

mybatis_入门程序Mybatis入门

samba服务器的功能

AOP编程_Android优雅权限框架(1)概念基础，2021金三银四前言正文大纲正文

Effective Java 8:通用程序设计

【Linux】UDP广播报文接收速率问题

OOM三种类型

工厂模式-三种类型

【递归】高效率求2的n次幂

win10本地scala和spark安装安装scala安装spark

Linux设备模型（中）之上层容器

Oracle 批量查询传入List 返回List

scala (3) Function 和 Method

PowerPC平台 Linux移植三