RAC结构删除一个节点并删除数据库软件的解决方案

1	节点2
使用dbca删除一个节点


2	节点1
alter database disable thread 2;


3	节点1
验证是否已经删除一个节点数据库
[03:49:06 oracle(db)@rac1 ~]$ srvctl config database -d vposrac
Database unique name: prod
Database name: prod
Oracle home: /u01/app/oracle/product/11.2.0.3/db
Oracle user: oracle
Spfile: +DATA/prod/spfileprod.ora
Domain:
Start options: open
Stop options: immediate
Database role: PRIMARY
Management policy: AUTOMATIC
Server pools: prod
Database instances: prod1
Disk Groups: DATA
Mount point paths:
Services:
Type: RAC
Database is administrator managed


4 节点1
停止节点2的监听
srvctl disable listener -l LISTENER -n vposrac2-s1;
srvctl stop listener -l LISTENER -n vposrac2-s1;


5	节点2
更新节点信息
/u01/app/oracle/product/11.2.0.3/db/oui/bin/runInstaller -updateNodeList ORACLE_HOME=/u01/app/oracle/product/11.2.0.3/db "CLUSTER_NODES={vposrac2-s1}" -local


6	节点2
删除节点2的ORACLE软件
/u01/app/oracle/product/11.2.0.3/db/deinstall/deinstall -local


7	节点1
更新剩余节点的信息
/u01/app/oracle/product/11.2.0.3/db/oui/bin/runInstaller -updateNodeList ORACLE_HOME=/u01/app/oracle/product/11.2.0.3/db "CLUSTER_NODES={vposrac1-s1}"


删除oracle软件已经完成
下面来删除grid软件
8	节点1
查看ons的资源情况
[10:19:57 oracle@vposrac1-s1 ~]$ olsnodes -s -t
vposrac1-s1     Active  Unpinned
vposrac2-s1     Active  Unpinned


9	节点2
以root用户执行rootcrs.pl 脚本清楚grid的配置信息
/u01/app/11.2.0.3/grid/crs/install/rootcrs.pl -deconfig -deinstall -force


10	节点1
再次确认ons的资源情况
[10:21:04 oracle@vposrac1-s1 ~]$ olsnodes -s -t
vposrac1-s1     Active  Unpinned
vposrac2-s1     Inactive        Unpinned


11	节点1
root用户执行
删除另外一个节点的信息
 crsctl delete node -n vposrac2-s1


olsnodes -s -t
rac1 Active Unpinned


12	节点2
oracle用户执行
/u01/app/11.2.0.3/grid/oui/bin/runInstaller -updateNodeList ORACLE_HOME=/u01/app/11.2.0.3/grid/ "CLUSTER_NODES=vposrac2-s1" CRS=TRUE -silent -local


13	节点2
卸载grid软件
oracle用户执行
/u01/app/11.2.0.3/grid/deinstall/deinstall -local
root 用户下执行
rm -rf /etc/oraInst.loc
rm -rf /opt/ORCLfmap
rm -rf /etc/oratab


14	节点1
更新剩余节点的信息
oracle用户执行：
/u01/app/11.2.0.3/grid/oui/bin/runInstaller -updateNodeList ORACLE_HOME=/u01/app/11.2.0.3/grid/ "CLUSTER_NODES=rac1" CRS=TRUE -silent




15 节点1
检查是否已经删除
[04:24:28 oracle(db)@rac1 ~]$ cluvfy stage -post nodedel -n rac2 -verbose


Performing post-checks for node removal


Checking CRS integrity...


Clusterware version consistency passed
The Oracle Clusterware is healthy on node "rac1"


CRS integrity check passed
Result:
Node removal check passed


Post-check for node removal was successful.


可以看到节点删除成功