0 说明
接到应用侧需求工单，需要克隆现有的一套测试库出来做业务测试，数据库是11204双节点RAC集群，源端数据库大小5.1T，数据文件总大小7.3T。
检查了本机数据库主机的环境配置，剩余存储和剩余内存均满足复制需求，因此原本的数据库环境主机上，克隆源RAC数据库得到一个新的实例。源库和目标库将共享相同的物理服务器和IP地址，但使用不同的instance_name、db_name和db_unique_name。
1 环境规划
在进行数据库克隆操作之前，需要对源端（主）和目标端（备）的环境进行详细规划，确保各项配置参数清晰明确，以便后续操作顺利进行。
| IP | VIP | 操作系统 | 数据库服务名 | instance_name | db_name | db_unique_name | 数据库版本 | 控制文件路径 | 数据文件路径 | 归档路径 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 源端（主） | 192.168.56.101 192.168.56.102 | 192.168.56.103 192.168.56.104 | rhel 6.8 | rac | rac1 rac2 | rac | rac | 11.2.0.4.0 | +DATA/rac/controlfile/ | +DATA/rac/datafile/ +DATA/rac/tempfile | +FRA/rac/archivelog/ |
| 目标端（备） | 192.168.56.101 192.168.56.102 | 192.168.56.103 192.168.56.104 | rhel 6.8 | rac_lpy | rac_lpy1 rac_lpy2 | rac_lpy | rac_lpy | 11.2.0.4.0 | +DATA/rac_lpy/controlfile/ | +DATA/rac_lpy/datafile/ +DATA/rac_lpy/tempfile | +FRA/rac_lpy/archivelog/ |
2 源端开启归档
rman duplicate命令需要源数据库处于归档模式，以便在克隆过程中持续应用产生的重做日志，保证目标库的数据一致性。
# 主任意节点
[oracle@r11gn1 ~]$ srvctl stop database -d rac
[oracle@r11gn1 ~]$ srvctl start instance -d rac -i rac1 -o mount
# 主1
SQL> alter database archivelog;
SQL> alter system set log_archive_dest_1='location=+FRA' scope=both sid='*';
SQL> archive log list
Database log mode Archive Mode
Automatic archival Enabled
Archive destination +FRA
Oldest online log sequence 48
Next log sequence to archive 49
Current log sequence 49
# 主1
SQL> alter database open;
# 主2
srvctl start instance -d rac -i rac2
3 源端开启强制日志
确保所有数据文件修改均记录到重做日志，防止克隆过程中因未记录（nologging）的直接路径加载（Direct Path Load）导致数据不一致。
# 主1
SQL> alter database force logging;
SQL> Select LOG_MODE,OPEN_MODE,PROTECTION_MODE,DATABASE_ROLE,SWITCHOVER_STATUS,GUARD_STATUS,FORCE_LOGGING from v$database;
LOG_MODE OPEN_MODE PROTECTION_MODE DATABASE_ROLE SWITCHOVER_STATUS GUARD_S FOR
------------ -------------------- -------------------- ---------------- -------------------- ------- ---
ARCHIVELOG READ WRITE MAXIMUM PERFORMANCE PRIMARY NOT ALLOWED NONE YES
4 创建目标端参数文件
基于源库的spfile生成pfile，然后修改关键标识符参数：
SQL> create pfile='/home/oracle/rac_lpy.ora' from spfile;
添加以下参数：
*.db_unique_name='rac_lpy'
定义数据库的唯一名称，必须与源库不同，是区分数据库实例的核心标识。
把rac全部替换为rac_lpy。
修改完成后示例如下：
... *.cluster_database=TRUE *.db_name='rac_lpy' *.db_unique_name='rac_lpy' ... rac_lpy1.instance_number=1 rac_lpy2.instance_number=2 ... rac_lpy1.thread=1 rac_lpy2.thread=2 ... rac_lpy1.undo_tablespace='UNDOTBS1' rac_lpy2.undo_tablespace='UNDOTBS2' ... *.control_files='+DATA/rac_lpy/controlfile/controlfile1.ctl' ...
5 创建备库目录
# audit_file_dest 两个节点都要创建
$ su - oracle
[oracle@r11gn1 ~]$ mkdir -p /u01/app/oracle/admin/rac_lpy/adump
# DATA 主1创建
$ su - grid
[grid@r11gn1 ~]$ asmcmd
ASMCMD> cd data
ASMCMD> mkdir RAC_LPY
ASMCMD> ls
RAC/
RAC_LPY/
adump目录必须创建否则sqlplus / as sysdba登录失败。
6 启动目标端节点1
使用准备好的pfile启动目标库的1节点rac_lpy1，使其运行在 NOMOUNT 状态。这是创建控制文件和进行后续克隆操作的基础。如果能启动成功，则说明参数文件配置没有问题。
# 声明实例名环境变量
[oracle@r11gn1 ~]$ export ORACLE_SID=rac_lpy1
[oracle@r11gn1 ~]$ sqlplus / as sysdba
SQL*Plus: Release 11.2.0.4.0 Production on Tue Jun 3 02:26:34 2025
Copyright (c) 1982, 2013, Oracle. All rights reserved.
Connected to an idle instance.
# 使用pfile启动到nomount状态
SQL> startup nomount pfile='/home/oracle/rac_lpy.ora';
ORACLE instance started.
Total System Global Area 2137886720 bytes
Fixed Size 2254952 bytes
Variable Size 620758936 bytes
Database Buffers 1509949440 bytes
Redo Buffers 4923392 bytes
# 查看参数
SQL> show parameter name
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
cell_offloadgroup_name string
db_file_name_convert string rac, rac_lpy
db_name string rac_lpy
db_unique_name string rac_lpy
global_names boolean FALSE
instance_name string rac_lpy1
lock_name_space string
log_file_name_convert string rac, rac_lpy
processor_group_name string
service_names string rac_lpy
7 创建spfile
将修改好的pfile转换为spfile，并存放在ASM共享存储 (+DATA) 中。这是RAC环境的强制要求，因为所有实例必须能够访问同一个参数文件。
SQL> create spfile='+DATA/rac_lpy/spfilerac_lpy.ora' from pfile='/home/oracle/rac_lpy.ora';
File created.
SQL> shutdown immediate;
ORA-01507: database not mounted
ORACLE instance shut down.
[grid@r11gn1 ~]$ asmcmd
ASMCMD> cd data/rac_lpy
ASMCMD> ls
PARAMETERFILE/
spfilerac_lpy.ora
指定共享spfile启动：
[oracle@r11gn1 dbs]$ echo "SPFILE='+DATA/rac_lpy/spfilerac_lpy.ora'" > initrac_lpy1.ora
节点2也指定：
[oracle@r11gn2 dbs]$ echo "SPFILE='+DATA/rac_lpy/spfilerac_lpy.ora'" > initrac_lpy2.ora
启动节点1，看看是否spfile启动：
SQL> startup nomount;
ORACLE instance started.
Total System Global Area 2137886720 bytes
Fixed Size 2254952 bytes
Variable Size 620758936 bytes
Database Buffers 1509949440 bytes
Redo Buffers 4923392 bytes
SQL> show parameter spfile;
NAME TYPE VALUE
------------------------------------ ----------- -------------------------------
spfile string +DATA/rac_lpy/spfilerac_lpy.ora
8 配置静态监听
因为目标库未启动，监听无法正常注册，为了使rman能够登录，要配置静态监听。
把下面这段监听配置添加到$ORACLE_HOME/network/admin/listener.ora文件中：
节点1配置
SID_LIST_LISTENER = (SID_LIST = (SID_DESC = (SID_NAME = rac1) (ORACLE_HOME = /u01/app/oracle/product/11.2/db_1) ) (SID_DESC = (GLOBAL_DBNAME = rac_lpy) (ORACLE_HOME = /u01/app/oracle/product/11.2/db_1) (SID_NAME = rac_lpy1) ) )
SID_NAME为备库实例名。GLOBAL_DBNAME为服务名，取什么都可以，需要和tnsnames.ora文件的server_name保持一致。
9 配置tns
节点1配置
RAC_LPY = (DESCRIPTION = (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.56.101)(PORT = 1521)) (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.56.102)(PORT = 1521)) (LOAD_BALANCE = off) (CONNECT_DATA = (SERVER = DEDICATED) (SERVICE_NAME = rac_lpy) ) )
tnsping看看服务名通不通：
tnsping RAC_LPY
10 验证目标端连通性
节点1 reload监听，在生产环境要注意，reload会导致数据库短暂无法连接。
[grid@r11gn1 ~]$ lsnrctl reload LSNRCTL for Linux: Version 11.2.0.4.0 - Production on 03-JUN-2025 02:56:47 Copyright (c) 1991, 2013, Oracle. All rights reserved. Starting /u01/app/11.2.0/grid/bin/tnslsnr: please wait... TNSLSNR for Linux: Version 11.2.0.4.0 - Production System parameter file is /u01/app/11.2.0/grid/network/admin/listener.ora Log messages written to /u01/app/grid/diag/tnslsnr/r11gn1/listener/alert/log.xml Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER))) Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=LISTENER))) STATUS of the LISTENER ------------------------ Alias LISTENER Version TNSLSNR for Linux: Version 11.2.0.4.0 - Production Start Date 03-JUN-2025 02:56:47 Uptime 0 days 0 hr. 0 min. 0 sec Trace Level off Security ON: Local OS Authentication SNMP OFF Listener Parameter File /u01/app/11.2.0/grid/network/admin/listener.ora Listener Log File /u01/app/grid/diag/tnslsnr/r11gn1/listener/alert/log.xml Listening Endpoints Summary... (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER))) Services Summary... Service "rac1" has 1 instance(s). Instance "rac1", status UNKNOWN, has 1 handler(s) for this service... Service "rac_lpy" has 1 instance(s). Instance "rac_lpy1", status UNKNOWN, has 1 handler(s) for this service... Service "racdg1" has 1 instance(s). Instance "racdg1", status UNKNOWN, has 1 handler(s) for this service... The command completed successfully
rac_lpy静态监听已生效。
目标端创建密码文件：
[oracle@r11gn1 dbs]$ cp orapwrac1 orapwrac_lpy1 [oracle@r11gn1 dbs]$ scp orapwrac_lpy1 r11gn2:/u01/app/oracle/product/11.2/db_1/dbs/orapwrac_lpy2
密码文件 (orapw
尝试登录：
[oracle@r11gn1 dbs]$ sqlplus sys/oracle@rac_lpy as sysdba
11 创建备库
核心步骤。使用RMAN的 DUPLICATE … FROM ACTIVE DATABASE 命令，通过网络直接从运行的源库 (rac) 克隆数据文件、控制文件等，创建目标库 (rac_lpy)。这是一种高效的“热克隆”方式。
主库rman duplicate到备库：
[oracle@r11gn1 dbs]$ rman target sys/oracle@rac1 auxiliary sys/oracle@rac_lpy
Recovery Manager: Release 11.2.0.4.0 - Production on Tue Jun 3 03:09:50 2025
Copyright (c) 1982, 2011, Oracle and/or its affiliates. All rights reserved.
connected to target database: RAC (DBID=2753265927)
connected to auxiliary database: RAC_LPY (not mounted)
RMAN>
run{
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate auxiliary channel c3 type disk;
allocate auxiliary channel c4 type disk;
duplicate target database to rac_lpy from active database;
release channel c1;
release channel c2;
release channel c3;
release channel c4;
}
备库要先启动到nomount状态。
复制报错：
RMAN-00571: =========================================================== RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS =============== RMAN-00571: =========================================================== RMAN-03002: failure of Duplicate Db command at 06/03/2025 13:11:20 RMAN-05501: aborting duplication of target database RMAN-06136: ORACLE error from auxiliary database: ORA-01503: CREATE CONTROLFILE failed ORA-12720: operation requires database is in EXCLUSIVE mode
duplicate不支持rac to rac，需要在目标端将CLUSTER_DATABASE参数设置为FALSE，然后再做复制。
Rman duplicate fail with RMAN-06136, ORA-01503, ORA-12720, ORA-00494 enqueue [CF]
(Doc ID 1335479.1)
备库修改CLUSTER_DATABASE参数，重启:
SQL> alter system set cluster_database=false scope=spfile;
System altered.
SQL> shutdown immediate;
SQL> startup nomount;
SQL> show parameter cluster_database
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
cluster_database boolean FALSE
在源端重新执行复制：
[oracle@r11gn1 dbs]$ rman target sys/oracle@rac1 auxiliary sys/oracle@rac_lpy
Recovery Manager: Release 11.2.0.4.0 - Production on Tue Jun 3 03:09:50 2025
Copyright (c) 1982, 2011, Oracle and/or its affiliates. All rights reserved.
connected to target database: RAC (DBID=2753265927)
connected to auxiliary database: RAC_LPY (not mounted)
RMAN>
run{
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate auxiliary channel c3 type disk;
allocate auxiliary channel c4 type disk;
duplicate target database to rac_lpy from active database;
release channel c1;
release channel c2;
release channel c3;
release channel c4;
}
日志显示 Finished Duplicate Db at 则说明复制成功。
将cluster_database还原，并重启：
SQL> alter system set cluster_database=true scope=spfile;
System altered.
SQL> shutdown immediate;
SQL> startup；
SQL> show parameter cluster_database
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
cluster_database boolean TRUE
cluster_database_instances integer 2
12 启动节点2
[oracle@r11gn2 ~]$ export ORACLE_SID=rac_lpy2
[oracle@r11gn2 ~]$ se
SQL> startup nomount;
SQL> alter database mount;
Database altered.
SQL> alter database open;
Database altered.
SQL> show parameter name
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
cell_offloadgroup_name string
db_file_name_convert string
db_name string RAC_LPY
db_unique_name string RAC_LPY
global_names boolean FALSE
instance_name string rac_lpy2
lock_name_space string
log_file_name_convert string
processor_group_name string
service_names string RAC_LPY
SQL> select inst_id,db_unique_name,open_mode,switchover_status,database_role from gv$database;
INST_ID DB_UNIQUE_NAME OPEN_MODE SWITCHOVER_STATUS DATABASE_ROLE
---------- ------------------------------ -------------------- -------------------- ----------------
1 RAC_LPY READ WRITE NOT ALLOWED PRIMARY
2 RAC_LPY READ WRITE NOT ALLOWED PRIMARY
SQL> select inst_id,instance_name,host_name,STARTUP_TIME,STATUS,ARCHIVER from gv$instance;
INST_ID INSTANCE_NAME HOST_NAME STARTUP_T STATUS ARCHIVE
---------- ---------------- ---------------------------------------------------------------- --------- ------------ -------
2 rac_lpy2 r11gn2 03-JUN-25 OPEN STARTED
1 rac_lpy1 r11gn1 03-JUN-25 OPEN STARTED
13 数据库注册到集群
使用Oracle集群件 (srvctl) 管理新克隆的RAC数据库 rac_lpy，包括启动、停止、状态查看、资源配置等。
[oracle@r11gn2 ~]$ srvctl add database -d rac_lpy -o $ORACLE_HOME [oracle@r11gn2 ~]$ srvctl add instance -d rac_lpy -i rac_lpy1 -n r11gn1 [oracle@r11gn2 ~]$ srvctl add instance -d rac_lpy -i rac_lpy2 -n r11gn2
先sqlplus关闭所有节点数据库实例，然后用srvctl启动：
[oracle@r11gn1 ~]$ srvctl start database -d rac_lpy [oracle@r11gn1 ~]$ crsctl stat res -t ora.rac_lpy.db 1 ONLINE ONLINE r11gn1 Open 2 ONLINE ONLINE r11gn2 Open
查看数据库配置：
[oracle@r11gn1 dbs]$ srvctl config database -d rac_lpy -a Database unique name: rac_lpy Database name: Oracle home: /u01/app/oracle/product/11.2/db_1 Oracle user: oracle Spfile: Domain: Start options: open Stop options: immediate Database role: PRIMARY Management policy: AUTOMATIC Server pools: rac_lpy Database instances: rac_lpy1,rac_lpy2 Disk Groups: Mount point paths: Services: Type: RAC Database is enabled Database is administrator managed
注册Spfile到集群：
srvctl modify database -d rac_lpy -p '+DATA/rac_lpy/spfilerac_lpy.ora'
注册Disk Groups,：
srvctl modify database -d rac_lpy -a DATA,FRA,OCR_VOTE
注册Services：
[oracle@r11gn1 ~]$ srvctl add service -d rac_lpy -s lpy_srv1 -r rac_lpy1 -a rac_lpy2 -P BASIC -y AUTOMATIC [oracle@r11gn1 ~]$ srvctl add service -d rac_lpy -s lpy_srv2 -r rac_lpy2 -a rac_lpy1 -P BASIC -y AUTOMATIC [oracle@r11gn1 ~]$ srvctl start service -d rac_lpy -s lpy_srv1,lpy_srv2
14 总结
通过以上步骤，可完整实现 Oracle RAC 集群的复制，确保新集群在网络、存储与功能层面与源端隔离且正常运行。该方案适用于数据库迁移、开发测试环境搭建等场景，操作时需严格遵循参数配置与环境隔离原则。
最后修改时间：2025-06-08 23:31:18
「喜欢这篇文章，您的关注和赞赏是给作者最好的鼓励」
关注作者
【版权声明】本文为墨天轮用户原创内容，转载时必须标注文章的来源（墨天轮），文章链接，文章作者等基本信息，否则作者和墨天轮有权追究责任。如果您发现墨天轮中有涉嫌抄袭或者侵权的内容，欢迎发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。
