在知识库中专注阅读，并随时返回相关工具与课程
最近正好需要搭建一套 RAC to RAC 的 ADG，之前有很多朋友希望可以多分享一些 Oracle 一键安装脚本的使用场景，本文演示一下如何使用 Oracle 一键安装脚本快速搭建一套 19C RAC to RAC 的 ADG 环境。 首先，使用 Oracle 一键安装脚本https://ora100.com/shell-install 快速搭建一套 19C RAC 环境： bash r
本章目标：掌握本章核心知识点 前置要求：完成前序章节学习 预计时长：60 分钟
最近正好需要搭建一套 RAC to RAC 的 ADG，之前有很多朋友希望可以多分享一些 Oracle 一键安装脚本的使用场景，本文演示一下如何使用 Oracle 一键安装脚本快速搭建一套 19C RAC to RAC 的 ADG 环境。
首先，使用 Oracle 一键安装脚本 快速搭建一套 19C RAC 环境：
1[root@dg01:/root]# cat /etc/os-release2NAME="CentOS Linux"3VERSION="7 (Core)"4ID="centos"5ID_LIKE="rhel fedora"6VERSION_ID="7"7PRETTY_NAME="CentOS Linux 7 (Core)"8ANSI_COLOR="0;31"9CPE_NAME="cpe:/o:centos:centos:7"10HOME_URL="https://www.centos.org/"11BUG_REPORT_URL="https://bugs.centos.org/"1213CENTOS_MANTISBT_PROJECT="CentOS-7"14CENTOS_MANTISBT_PROJECT_VERSION="7"15REDHAT_SUPPORT_PRODUCT="centos"16REDHAT_SUPPORT_PRODUCT_VERSION="7"1718[root@dg01:/root]# cat /etc/redhat-release19CentOS Linux release 7.9.2009 (Core)2021## 搭建 ADG 不需要创建数据库，只需要安装到 Oracle 软件完成即可22[root@dg01:/soft]# ./OracleShellInstall \23-ud Y `# 安装到Oracle软件结束`\24-lf eth0 `# 公网IP网卡名称`\25-pf eth1 `# 心跳IP网卡名称`\26-n dg`# 主机名前缀`\27-hn dg01,dg02 `# 所有节点主机名`\28-ri 100.99.40.231,100.99.40.232 `# 公网IP地址`\29-vi 100.99.40.233,100.99.40.234 `# 虚拟IP地址`\30-si 100.99.40.235 `# SCAN IP地址`\31-rp oracle `# root用户密码`\32-cn dg-cluster `# 集群名称`\33-sn dg-scan `# SCAN名称`\34-od /dev/sda,/dev/sdb,/dev/sdc `# OCR磁盘组磁盘列表`\35-dd /dev/sde,/dev/sdf,/dev/sdg,/dev/sdh,/dev/sdi,/dev/sdj `# DATA磁盘组磁盘列表`\36-ad /dev/sdd `# ARCH磁盘组磁盘列表`\37-or NORMAL `# OCR磁盘组冗余度`\38-o dgdb `# 数据库名称`\39-tsi 100.98.32.5 `# 时间服务器IP`\40-gpa 36582629 `# Grid PSU/RU补丁编号`\41-jpa 36414915 `# OJVM PSU/RU补丁编号`默认使用专业版脚本。脚本下载：ora100.com/shell-install
19C RAC 软件安装完成（包括补丁安装），共花费 35 分钟。
使用 Duplicate 在线复制的方式搭建 ADG，无需 RMAN 备份，可以通过 RMAN Duplicate 直接在线从主库拷贝文件。
详细的搭建教程可以参考 100天精通Oracle 实战系列 课程中的 （第45天）DataGuard 搭建之 RAC 到 RAC，课程中有完整的步骤以及教程。
搭建 DataGuard 必须要开启归档模式和强制日志模式：
1-- 主库检查是否开启归档和强制日志2SQL> select log_mode,force_logging from v$database;34LOG_MODE FORCE_LOGGING5------------ ---------------------------------------6ARCHIVELOG NO78SQL> alter database force logging;910Database altered.1112SQL> select log_mode,force_logging from v$database;1314LOG_MODE FORCE_LOGGING15------------ ---------------------------------------16ARCHIVELOG YES查看主库的在线日志数量以及大小：
1-- 主库的在线日志组数为 4 组2SQL> set line2223set pagesize10004col member for a605select t2.thread#,t1.group#,t1.member,t2.bytes/1024/1024 from v$logfile t1,v$log t2 where t1.group#=t2.group# order by 1,2;直接在主库增加 10 组 standby redo log，rman duplicate 复制之后，备库也会同步创建：
1-- 每个节点有 4 组在线重做日志，单个大小为 500 M；需要创建 （4+1）*2= 10 组备库重做日志2SQL> ALTER DATABASE ADD STANDBY LOGFILE THREAD 13group 10 ('+DATA', '+FRA') SIZE 500M,4group 11 ('+DATA', '+FRA') SIZE 500M,5group 12 ('+DATA', '+FRA') SIZE 500M,6group 13 ('+DATA', '+FRA') SIZE 500M,7group 14 ('+DATA', '+FRA') SIZE 500M;89SQL> ALTER DATABASE ADD STANDBY LOGFILE THREAD 210group 16 ('+DATA', '+FRA') SIZE 500M,11group 17 ('+DATA', '+FRA') SIZE 500M,12group 18 ('+DATA', '+FRA') SIZE 500M,13group 19 ('+DATA', '+FRA') SIZE 500M,14group 20 ('+DATA', '+FRA') SIZE 500M;rac 的密码文件通常存放于 ASM 盘中，查看对应的密码文件位置：
1[oracle@rptdb1:/home/oracle]$ srvctl config database -d rptdb | grep Password2Password file: +DATA/RPTDB/PASSWORD/pwdrptdb.256.1180304871使用 asmcmd 进入到 ASM 磁盘组中，复制密码文件到文件系统：
1[grid@rptdb1:/home/grid]$ asmcmd -p2ASMCMD [+] > pwcopy +DATA/RPTDB/PASSWORD/pwdrptdb.256.1180304871 /home/grid/3copying +DATA/RPTDB/PASSWORD/pwdrptdb.256.1180304871 -> /home/grid/pwdrptdb.256.1180304871将密码文件拷贝到备库，并且重命名，命名规则为 orapw+sid：
1 ## 拷贝密码文件至备库2[grid@rptdb1:/home/grid]$ scp /home/grid/pwdrptdb.256.1180304871 grid@10.199.40.231:/home/grid/orapwdgdb3grid@10.199.40.231's password: 4pwdrptdb.256.118030487156## 移动密码文件至备库 ASM 盘7[grid@dg01:/home/grid]$ asmcmd -p8ASMCMD [+] > cd data9ASMCMD [+data] > mkdir DGDB10ASMCMD [+data] > cd DGDB11ASMCMD [+data/DGDB] > mkdir PASSWORD12ASMCMD [+data/DGDB] > ls13PASSWORD/14ASMCMD [+data/DGDB] > cd PASSWORD15ASMCMD [+data/DGDB/PASSWORD] > pwcopy /home/grid/orapwdgdb +DATA/DGDB/PASSWORD/orapwdgdb16copying /home/grid/orapwdgdb -> +DATA/DGDB/PASSWORD/orapwdgdb17ASMCMD [+data/DGDB/PASSWORD] > ls18orapwdgdb由于要使用 RMAN 备份在备库恢复，所以需要再主库生成参数文件，修改相关参数后将备库启动到 nomount 状态：
1SQL> create pfile='/home/oracle/pfile_rptdb.ora' from spfile;23File created.将主库生成并修改后的参数文件在备库创建好（这里建议源端目标端均使用 OMF）：
1[oracle@dg01:/home/oracle]$ cat<<-\EOF>/home/oracle/pfile_dgdb.ora2*._gc_policy_time=03*._gc_undo_affinity=FALSE4*._optimizer_use_feedback=FALSE5*._use_adaptive_log_file_sync='FALSE'6*.audit_file_dest='/u01/app/oracle/admin/dgdb/adump'7*.audit_sys_operations=FALSE8*.audit_trail='NONE'9*.cluster_database=true10*.compatible='19.0.0'11*.control_file_record_keep_time=3112*.control_files='+DATA/RPT/CONTROLFILE/control01.ctl','+DATA/RPT/CONTROLFILE/control02.ctl'13*.db_block_size=819214*.db_create_file_dest='+DATA'15*.db_files=819216*.db_name='rptdb'17*.deferred_segment_creation=FALSE18*.diagnostic_dest='/u01/app/oracle'19*.dispatchers='(PROTOCOL=TCP) (SERVICE=dgdbXDB)'20dgdb1.instance_number=121dgdb2.instance_number=222*.local_listener='-oraagent-dummy-'23*.log_archive_format='%t_%s_%r.dbf'24*.nls_language='AMERICAN'25*.nls_territory='AMERICA'26*.open_cursors=30027*.parallel_force_local=TRUE28*.parallel_max_servers=6429*.pga_aggregate_target=14719m30*.processes=256031*.remote_login_passwordfile='exclusive'32*.sga_target=44156m33dgdb2.thread=234dgdb1.thread=135*.undo_retention=1080036dgdb2.undo_tablespace='UNDOTBS2'37dgdb1.undo_tablespace='UNDOTBS1'38*.db_unique_name='dgdb'39*.fal_client='RPTDB_STB'40*.fal_server='RPTDB_PRI'41*.log_archive_config='dg_config=(RPTDB,RPT)'42*.log_archive_dest_1='LOCATION=+ARCH VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=RPT'43*.log_archive_dest_2='SERVICE=RPTDB_PRI VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=RPTDB'44*.standby_file_management='AUTO'45EOF备库根据参数文件在两个节点均创建必要的目录：
1[oracle@dg01:/home/oracle]$ mkdir -p /u01/app/oracle/admin/dbdg/adump2[oracle@dg02:/home/oracle]$ mkdir -p /u01/app/oracle/admin/dbdg/adump备库创建初始化参数文件：
1## 节点一2[oracle@dg01:/home/oracle]$ cat<<-\EOF>$ORACLE_HOME/dbs/initdgdb1.ora3spfile='+DATA/DGDB/spfiledgdb.ora'4EOF56## 节点二7[oracle@dg02:/home/oracle]$ cat<<-\EOF>$ORACLE_HOME/dbs/initdgdb2.ora8spfile='+DATA/DGDB/spfiledgdb.ora'9EOF备库节点一创建 spfile：
1Connected to an idle instance.23SQL> create spfile='+DATA/DGDB/spfiledgdb.ora' from pfile='/home/oracle/pfile_dgdb.ora';45File created.备库集群注册数据库实例：
1[oracle@dg01:/home/oracle]$ srvctl add database -d dgdb -o $ORACLE_HOME -p +DATA/DGDB/spfilerpt.ora -pwfile +DATA/DGDB/PASSWORD/orapwdgdb2[oracle@dg01:/home/oracle]$ srvctl add instance -d dgdb -i dgdb1 -n dg013[oracle@dg01:/home/oracle]$ srvctl add instance -d dgdb -i dgdb2 -n dg024## 打开节点一到 nomount 状态5[oracle@dg01:/home/oracle]$ srvctl start instance -d dgdb -i dgdb1 -o nomount备库检查监听：
1[grid@dg02:/home/grid]$ lsnrctl stat LISTENER_SCAN123LSNRCTL for Linux: Version 19.0.0.0.0 - Production on 16-OCT-2025 11:04:3245Copyright (c) 1991, 2024, Oracle. All rights reserved.67Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=LISTENER_SCAN1)))8STATUS of the LISTENER9------------------------10Alias LISTENER_SCAN111Version TNSLSNR for Linux: Version 19.0.0.0.0 - Production12Start Date 16-OCT-2025 09:47:5613Uptime 0 days 1 hr. 16 min. 36 sec14Trace Level off15Security ON: Local OS Authentication16SNMP OFF17Listener Parameter File /u01/app/19.3.0/grid/network/admin/listener.ora18Listener Log File /u01/app/grid/diag/tnslsnr/dg02/listener_scan1/alert/log.xml19Listening Endpoints Summary...20 (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER_SCAN1)))21 (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=10.199.40.235)(PORT=1521)))22Services Summary...23Service "dgdb" has 1 instance(s).24 Instance "dgdb1", status BLOCKED, has 1 handler(s) for this service...25The command completed successfully1## 所有节点均配置 TNS2cat<<-\EOF>>$ORACLE_HOME/network/admin/tnsnames.ora3RPTDB_PRI =4 (DESCRIPTION =5 (ADDRESS = (PROTOCOL = TCP)(HOST = 10.199.40.204)(PORT = 1521))6 (CONNECT_DATA =7 (SERVER = DEDICATED)8 (SERVICE_NAME = rptdb)9 )10 )11 12RPTDB_STB =13 (DESCRIPTION =14 (ADDRESS = (PROTOCOL = TCP)(HOST = 10.199.40.235)(PORT = 1521))15 (CONNECT_DATA =16 (SERVER = DEDICATED)17 (SERVICE_NAME = dgdb)18 (UR=A)19 )20 )21EOF2223## 测试是否可以连通24tnsping RPTDB_PRI25tnsping RPTDB_STB26sqlplus sys/oracle@RPTDB_PRI as sysdba27sqlplus sys/oracle@RPTDB_STB as sysdba1[oracle@dg01:/home/oracle]$ rman target sys/oracle@RPTDB_PRI AUXILIARY sys/oracle@RPTDB_STB23Recovery Manager: Release 19.0.0.0.0 - Production on Thu Oct 16 11:13:40 20254Version 19.24.0.0.056Copyright (c) 1982, 2019, Oracle and/or its affiliates. All rights reserved.78connected to target database: RPTDB (DBID=170820080)9connected to auxiliary database: RPTDB (not mounted)1011RMAN> run {12allocate channel prmy1 type disk;13allocate channel prmy2 type disk;14allocate channel prmy3 type disk;15allocate channel prmy4 type disk;16allocate channel prmy5 type disk;17allocate channel prmy6 type disk;18allocate auxiliary channel aux1 type disk;19allocate auxiliary channel aux2 type disk;20allocate auxiliary channel aux3 type disk;21allocate auxiliary channel aux4 type disk;22allocate auxiliary channel aux5 type disk;23allocate auxiliary channel aux6 type disk;24duplicate target database for standby from active database dorecover nofilenamecheck;25}等待 RMAN duplicate 复制完成即可。
1-- 重启备库2个节点到只读状态2[oracle@dg01:/home/oracle]$ srvctl stop db -d dgdb3[oracle@dg01:/home/oracle]$ srvctl start db -d dgdb -o "read only"1SQL> alter database recover managed standby database using current logfile disconnect;23Database altered.1alter system set log_archive_config='DG_CONFIG=(RPTDB,DGDB)';2alter system set log_archive_dest_1='LOCATION=+FRA VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=RPTDB';3alter system set log_archive_dest_2='SERVICE=RPTDB_STB VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=DGDB';4alter system set fal_client='RPTDB_PRI';5alter system set fal_server='RPTDB_STB';6alter system set log_archive_dest_state_2=enable;1-- 主库2SQL> set line2222 pages10003col status for a104col type for a105col error for a206col gap_status for a207col synchronization_status for a308col recovery_mode for a609select inst_id,status,DEST_ID,TYPE,ERROR,GAP_STATUS,SYNCHRONIZED,SYNCHRONIZATION_STATUS,RECOVERY_MODE from GV$ARCHIVE_DEST_STATUS where STatus <> 'INACTIVE' and type = 'PHYSICAL';1011-- 备库12SQL> set line222 pages100013select process,thread#,group#,sequence#,status from v$managed_standby;1415SQL> select t2.thread#,t1.group#,t1.member,t2.STATUS,t2.ARCHIVED,t2.bytes/1024/1024 from v$logfile t1,v$standby_log t2 where t1.group#=t2.group# order by 1,2;至此，ADG 搭建完成，备库可用于报表查询。