oracle rac搭建单实例DG步骤（阅读全篇后再做）

环境介绍

主库：

主机名	rac01	rac02
实体IP	10.206.132.232	10.206.132.233
私有IP	192.168.56.12	192.168.56.13
虚拟IP	10.206.132.237	10.206.132.238
SCAN IP	10.206.132.239
实例名	racdb1	racdb2
数据库名	racdb
数据文件目录	DGDATA01,DGDATA02(ASM磁盘）

备库：

主机名	dr-rac
实体IP	10.206.132.245
虚拟IP	10.206.132.246
实例名	racdb
数据库名	racdb
数据文件目录	/oradata01/racdb/,/oradata02/racdb

数据文件目录备库和主库的ASM目录数量和大小要一致，后面会讲原因

1.安装oracle软体（和rac数据库的软体版本一致，本实验使用的是11GR2）

2.主库开启归档并打开force logging

打开force logging

SQL> alter database force logging;

Database altered.

SQL> select force_logging from v$database;

FOR
---
YES

打开归档

关闭数据库（两个节点执行）
$ sqlplus / as sysdba
SQL> shutdown immediate;
将数据库打开至mount状态（两个节点执行）
SQL> startup mount;
修改数据库的归档模式（任一节点即可）
SQL> alter database archivelog;
修改归档路径（两个节点执行）
SQL> alter system set log_archive_dest_1='location=/archlog/racdb';
打开数据库（两个节点）
SQL> alter database open;

3.主库全备（任一节点执行）

$ rman target /

run{
allocate channel c1 device type disk format '/oggdata/backup/full_%d_%T_%s_%p.bak';
allocate channel c2 device type disk format '/oggdata/backup/full_%d_%T_%s_%p.bak';
allocate channel c3 device type disk format '/oggdata/backup/full_%d_%T_%s_%p.bak';
sql 'alter system archive log current';
backup as compressed backupset database plus archivelog;
backup current controlfile;
backup spfile;
}

4.将全备文件传送至备库

$ scp * 10.206.132.245:/data/backup

6.将主库的参数文件传送至备库

SQL> create pfile='/tmp/initracdb.ora' from spfile;

$ scp /tmp/initracdb.ora 10.206.132.245:/oracle/11204/dbs

根据实际的情况，RAC的参数文件和单实例的参数文件有一定的差别，所以我一般会从一个单实例数据库创建一个pfile传送过来并修改

7.修改参数文件并创建参数文件所需目录

其中以双下划线开头的是oracle自动内存管理生成的，可以直接删除，如

testdb.__db_cache_size=1862270976
testdb.__java_pool_size=16777216
testdb.__large_pool_size=33554432
testdb.__oracle_base='/oracle'#ORACLE_BASE set from environment
testdb.__pga_aggregate_target=838860800
testdb.__sga_target=2483027968
testdb.__shared_io_pool_size=0
testdb.__shared_pool_size=503316480
testdb.__streams_pool_size=33554432

其中以单下划线开头的是隐含参数，根据实际情况修改

修改SGA,PGA

修改控制文件位置

*.control_files='/oradata01/racdb/control01.ctl','/oradata01/racdb/control02.ctl'

创建目录

*.audit_file_dest='/oracle/admin/racdb/adump'

*.db_recovery_file_dest='/oracle/fast_recovery_area'

其他参数根据实际情况修改，删除或者添加

8.创建spfile并打开实例

SQL> create spfile from pfile;
SQL> startup nomount;

9.还原standby controlfile

进入主库查看controlfile的备份片

[oracle@rac01 ~]$ rman target /
RMAN> list backup of controlfile;
------------------------------------------------------------------------------------
BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ ---------------
26      Full    17.67M     DISK        00:00:01     15-APR-20      
        BP Key: 26   Status: AVAILABLE  Compressed: NO  Tag: TAG20200415T150050
        Piece Name: /oggdata/backup/full_RACDB_20200415_26_1.bak
  Control File Included: Ckp SCN: 558979       Ckp time: 15-APR-20
------------------------------------------------------------------------------------

备库进入rman进行还原

[oracle@dr-rac dbs]$ rman target /
RMAN> restore standby controlfile from '/data/backup/full_RACDB_20200415_26_1.bak';

10.将数据库启动至mount状态并进行数据库还原（备库）

mount数据库

[oracle@dr-rac dbs]$ sqlplus / as sysdba
SQL> alter database mount;

此时遇到了一个问题，在RAC数据库中数据文件（包括临时表空间，在线联机日志）是在ASM磁盘内的，例如+DGDATA01等，在单实例上面是系统目录，如何对应

查看当前数据文件的状态

SQL> select name from v$datafile;

NAME
--------------------------------------------------------------------------------
+DGDATA01/racdb/datafile/system.260.1037709075
+DGDATA01/racdb/datafile/sysaux.261.1037709079
+DGDATA01/racdb/datafile/undotbs1.262.1037709081
+DGDATA01/racdb/datafile/undotbs2.264.1037709089
+DGDATA01/racdb/datafile/users.265.1037709091
+DGDATA02/racdb/datafile/mytbs01.dbf

SQL> select name from v$tempfile;

NAME
--------------------------------------------------------------------------------
+DGDATA01/racdb/tempfile/temp.263.1037709085

SQL> select member from v$logfile;

MEMBER
--------------------------------------------------------------------------------
+DGDATA01/racdb/redo01.log
+DGDATA01/racdb/redo02.log
+DGDATA01/racdb/redo03.log
+DGDATA01/racdb/redo04.log

因此我们需要在RMAN中将恢复的目录对应起来（下面的脚本根据自己实际情况进行修改）

DECLARE
       dir   VARCHAR2 (100);

       CURSOR dbfs
       IS
          SELECT FILE_NAME FROM DBA_DATA_FILES;
    BEGIN
       FOR dbf IN dbfs
       LOOP
          SELECT SUBSTR (dbf.FILE_NAME, 2, INSTR (dbf.FILE_NAME, '/', 1) - 2) INTO DIR FROM DUAL;
            IF (DIR = 'DGDATA01') THEN
             DBMS_OUTPUT.PUT_LINE('SET NEWNAME FOR DATAFILE '''||dbf.FILE_NAME||''' TO ''/oradata01'||substr(dbf.file_name,instr(dbf.file_name,'/',1),length(dbf.file_name))||''';');
            ELSIF (DIR = 'DGDATA02') THEN
             DBMS_OUTPUT.PUT_LINE('SET NEWNAME FOR DATAFILE '''||dbf.FILE_NAME||''' TO ''/oradata02'||substr(dbf.file_name,instr(dbf.file_name,'/',1),length(dbf.file_name))||''';');
            END IF;
       END LOOP;
    END;


----------------------------------------------------------------------------------------------------------------------
SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/system.260.1037709075' TO '/oradata01/racdb/datafile/system.260.1037709075';
SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/sysaux.261.1037709079' TO '/oradata01/racdb/datafile/sysaux.261.1037709079';
SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/undotbs1.262.1037709081' TO '/oradata01/racdb/datafile/undotbs1.262.1037709081';
SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/undotbs2.264.1037709089' TO '/oradata01/racdb/datafile/undotbs2.264.1037709089';
SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/users.265.1037709091' TO '/oradata01/racdb/datafile/users.265.1037709091';
SET NEWNAME FOR DATAFILE '+DGDATA02/racdb/datafile/mytbs01.dbf' TO '/oradata02/racdb/datafile/mytbs01.dbf';
----------------------------------------------------------------------------------------------------------------------

使用RMAN进行数据库的还原（备库），恢复之前要在对应的目录下建立子文件夹（如racdb/datafile）

[oracle@dr-rac dbs]$ rman target /
RMAN> catalog start with '/data/backup';

run {
    SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/system.260.1037709075' TO '/oradata01/racdb/datafile/system.260.1037709075';
    SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/sysaux.261.1037709079' TO '/oradata01/racdb/datafile/sysaux.261.1037709079';
    SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/undotbs1.262.1037709081' TO '/oradata01/racdb/datafile/undotbs1.262.1037709081';
    SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/undotbs2.264.1037709089' TO '/oradata01/racdb/datafile/undotbs2.264.1037709089';
    SET NEWNAME FOR DATAFILE '+DGDATA01/racdb/datafile/users.265.1037709091' TO '/oradata01/racdb/datafile/users.265.1037709091';
    SET NEWNAME FOR DATAFILE '+DGDATA02/racdb/datafile/mytbs01.dbf' TO '/oradata02/racdb/datafile/mytbs01.dbf';
    allocate channel c1 device type disk;
    allocate channel c2 device type disk;
    allocate channel c3 device type disk;
    restore database;
    release channel c1;
    release channel c2;
    release channel c3;
}

11.修改controlfile中redo和tempfile的位置，以便打开后自动创建（前提还是要创建好对应的目录，否则一会儿数据库打不开）

SQL> alter database rename file '+DGDATA01/racdb/tempfile/temp.263.1037709085' to '/oradata01/racdb/tempfile/temp.263.1037709085';

DECLARE
       dir   VARCHAR2 (100);

       CURSOR LFS
       IS
          SELECT MEMBER FROM v$logfile;
    BEGIN
       FOR LF IN LFS
       LOOP
          SELECT SUBSTR (LF.MEMBER, 2, INSTR (LF.MEMBER, '/', 1) - 2) INTO DIR FROM DUAL;
          IF (DIR = 'DGDATA01') THEN
             DBMS_OUTPUT.PUT_LINE('ALTER DATABASE RENAME FILE '''||LF.MEMBER||''' TO ''/oradata01'||substr(LF.MEMBER,instr(LF.MEMBER,'/',1),length(LF.MEMBER))||''';');
             ELSIF (DIR = 'DGDATA02') THEN
             DBMS_OUTPUT.PUT_LINE('ALTER DATABASE RENAME FILE '''||LF.MEMBER||''' TO ''/oradata02'||substr(LF.MEMBER,instr(LF.MEMBER,'/',1),length(LF.MEMBER))||''';');
          END IF;
       END LOOP;
    END;

-------------------------------------------------------------------------------
ALTER DATABASE RENAME FILE '+DGDATA01/racdb/redo01.log' TO '/oradata01/racdb/redo01.log';
ALTER DATABASE RENAME FILE '+DGDATA01/racdb/redo02.log' TO '/oradata01/racdb/redo02.log';
ALTER DATABASE RENAME FILE '+DGDATA01/racdb/redo03.log' TO '/oradata01/racdb/redo03.log';
ALTER DATABASE RENAME FILE '+DGDATA01/racdb/redo04.log' TO '/oradata01/racdb/redo04.log';
-------------------------------------------------------------------------------

12.打开数据库

这时候报错

SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-10458: standby database requires recovery
ORA-01157: cannot identify/lock data file 1 - see DBWR trace file
ORA-01110: data file 1: '+DGDATA01/racdb/datafile/system.260.1037709075'

这是因为控制文件中的datafile的路径没有改变，两种办法解决

第一种，修改控制文件中数据文件的路径

DECLARE
       dir   VARCHAR2 (100);

       CURSOR dbfs
       IS
          SELECT FILE_NAME FROM DBA_DATA_FILES;
    BEGIN
       FOR dbf IN dbfs
       LOOP
          SELECT SUBSTR (dbf.FILE_NAME, 2, INSTR (dbf.FILE_NAME, '/', 1) - 2) INTO DIR FROM DUAL;
            IF (DIR = 'DGDATA01') THEN
             DBMS_OUTPUT.PUT_LINE('ALTER DATABASE  RENAME FILE '''||dbf.FILE_NAME||''' TO ''/oradata01'||substr(dbf.file_name,instr(dbf.file_name,'/',1),length(dbf.file_name))||''';');
            ELSIF (DIR = 'DGDATA02') THEN
             DBMS_OUTPUT.PUT_LINE('ALTER DATABASE  RENAME FILE  '''||dbf.FILE_NAME||''' TO ''/oradata02'||substr(dbf.file_name,instr(dbf.file_name,'/',1),length(dbf.file_name))||''';');
            END IF;
       END LOOP;
    END;
-----------------------------------------------------------------------------
ALTER DATABASE  RENAME FILE '+DGDATA01/racdb/datafile/system.260.1037709075' TO '/oradata01/racdb/datafile/system.260.1037709075';
ALTER DATABASE  RENAME FILE '+DGDATA01/racdb/datafile/sysaux.261.1037709079' TO '/oradata01/racdb/datafile/sysaux.261.1037709079';
ALTER DATABASE  RENAME FILE '+DGDATA01/racdb/datafile/undotbs1.262.1037709081' TO '/oradata01/racdb/datafile/undotbs1.262.1037709081';
ALTER DATABASE  RENAME FILE '+DGDATA01/racdb/datafile/undotbs2.264.1037709089' TO '/oradata01/racdb/datafile/undotbs2.264.1037709089';
ALTER DATABASE  RENAME FILE '+DGDATA01/racdb/datafile/users.265.1037709091' TO '/oradata01/racdb/datafile/users.265.1037709091';
ALTER DATABASE  RENAME FILE  '+DGDATA02/racdb/datafile/mytbs01.dbf' TO '/oradata02/racdb/datafile/mytbs01.dbf';
-----------------------------------------------------------------------------

第二种，在还原控制文件之前，我们在参数文件中加入如下两个参数

alter system set log_file_name_convert='+DGDATA01/racdb/','/oradata01/racdb/','+DGDATA02/racdb/','/oradata02/racdb/' scope=spfile;
alter system set db_file_name_convert='+DGDATA01/racdb/datafile/','/oradata01/racdb/datafile/','+DGDATA02/racdb/datafile/','/oradata02/racdb/datafile/' scope=spfile; 

第二种还原的控制文件中datafile和logfile自动会定位到正确的目录，同时在第十步中也不需要前面NET NEWNAME，直接还原即可，因此强烈建议这两个参数在还原控制文件之前加入。

参数一定要注意，前面最后加/，后面也要加/，否则替换的时候会出错，例如'+DGDATA01/racdb/','/oradata01/racdb'，会把+DGDATA01/racdb/redo01.log对应成/oradata01/racdbredo01.log

再次打开数据时，又报错了

SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-10458: standby database requires recovery
ORA-01152: file 1 was not restored from a sufficiently old backup
ORA-01110: data file 1: '/oradata01/racdb/datafile/system.260.1037709075'

这是因为我们没有恢复数据库，使用rman恢复数据库，有两种方法

第一种，恢复数据库（这里面的SCN也可以从RMAN的list backup of archivelog all里面看到归档的最高SCN）

恢复之前要加上这两个参数的原因可以看后面的试验
SQL> alter system set db_file_name_convert='+DGDATA01/racdb/datafile','/oradata01/racdb/datafile','+DGDATA02/racdb/datafile','/oradata02/racdb/datafile' scope=spfile;
SQL> alter system set standby_file_management=auto;
查看主库当前的SCN
SQL> select current_scn from v$database;
主库归档当前的REDO
SQL> alter system archive log current;
将主库的归档传送至备库
$ scp /archlog/readb/* 10.206.132.245:/oradata02/archlog
备库恢复
RMAN> catalog start with '/oradata02/archlog';
RMAN> recover database until scn 565400;
打开数据库
SQL> alter database open;

第二种，不用管，继续配置DG库，之后会自动将归档传过来并应用

13.配置备库的监听

从其他库copy过来一个修改即可

$ cd /oracle/11204/network/admin/
$ scp listener.ora 10.206.132.245:`pwd`
$ vi listener.ora
$ lsnrctl star

14.将主库（任一节点）的口令文件传过来

$ cd $ORACLE_HOME/dbs
$ scp orapwracdb1 10.206.132.245:/oracle/11204/dbs/orapwracdb

15.修改主库的TNS（加入自己和备库的TNS）并传送到另一个节点和备库

$ vi tnsnames.ora
-------------------------------------------------------------------------------
RACDB =
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 10.206.132.239)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = racdb)
    )
  )

STANDBY_RACDB =
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 10.206.132.245)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = racdb)
    )
  )
-------------------------------------------------------------------------------
$ scp tnsnames.ora 10.206.132.233:/oracle/home/network/admin/
$ scp tnsnames.ora 10.206.132.245:/oracle/11204/network/admin

16.修改备库的参数

alter system set db_unique_name='standby_racdb' scope=spfile;
alter system set db_file_name_convert='+DGDATA01/racdb/datafile','/oradata01/racdb/datafile','+DGDATA02/racdb/datafile','/oradata02/racdb/datafile' scope=spfile; 
alter system set log_file_name_convert='+DGDATA01/racdb/','/oradata01/racdb','+DGDATA02/racdb','/oradata02/racdb' scope=spfile;
重啟數據庫
alter system set log_archive_config='dg_config=(racdb,standby_racdb)';
alter system set log_archive_dest_1='LOCATION=/oradata02/archlog  valid_for=(all_logfiles,all_roles) db_unique_name=standby_racdb';
alter system set log_archive_dest_2='service=racdb async valid_for=(online_logfiles, primary_role) db_unique_name=racdb';
alter system set log_archive_dest_state_1=enable;
alter system set log_archive_dest_state_2=enable;
alter system set fal_server=racdb;
alter system set fal_client=standby_racdb;
alter system set standby_file_management=auto;    
alter system set log_archive_max_processes=30;

18.修改主库参数（我的参数文件时共享的，因此修改一个节点即可）

alter system set db_file_name_convert='/oradata01/racdb/datafile','+DGDATA01/racdb/datafile','/oradata02/racdb/datafile','+DGDATA02/racdb/datafile' scope=spfile; 
alter system set log_file_name_convert='/oradata01/racdb','+DGDATA01/racdb/','/oradata02/racdb','+DGDATA02/racdb' scope=spfile;
重啟數據庫（根据实际业务，可以先修改，等无业务期间重启实例）
alter system set log_archive_config='dg_config=(racdb,standby_racdb)';
alter system set log_archive_dest_1='LOCATION=/archlog/racdb  valid_for=(all_logfiles,all_roles) db_unique_name=racdb';
alter system set log_archive_dest_2='service=standby_racdb async valid_for=(online_logfiles, primary_role) db_unique_name=standby_racdb';
alter system set log_archive_dest_state_1=enable;
alter system set log_archive_dest_state_2=enable;
alter system set fal_server=standby_racdb;
alter system set fal_client=racdb;
alter system set standby_file_management=auto;    
alter system set log_archive_max_processes=30;

19.打开备库的日志应用

SQL> alter database recover managed standby database disconnect;

20.新建表并插入数据测试

create table testdg(id number);
insert into testdg values (1);
commit;
#因为没有添加standby_logfile因此无法实时应用，需要切换归档
alter system archive log current;

同时发现REDO产生了，应该是修改主库参数文件后传送过来或备库应用日志的时候产生的。

tempfile是在备库打开（open）的时候产生的。

实验

主备目录不相同的情况下，如果全备之后，增加了数据文件或者修改了数据文件的大小，在恢复备库的时候，会发生什么

试验一：主库全备后，增加数据文件的大小

alter database datafile '+DGDATA02/racdb/datafile/mytbs01.dbf' resize 2g;

alter system archive log current;

select current_scn from v$database;
614673

alter system archive log current;

$ scp 2_17_1037709071.dbf 2_18_1037709071.dbf 1_30_1037709071.dbf 1_31_1037709071.dbf 10.206.132.245:/oradata02/archlog

RMAN> catalog start with '/oradata02/archlog';

SQL> shutdown immediate;

SQL> startup mount;

RMAN> recover database until scn 614673;

# cd /oradata02/racdb/datafile
# du -sh mytbs01.dbf
----------------------------------------
2.1G    mytbs01.dbf
----------------------------------------

通过上述实验，发现只要数据文件已经被还原，那么增加数据文件的大小会自动应用到对应的目录上面（感觉应该是根据的数据文件号）

实验二：主库全备后，增加数据文件

alter tablespace mytbs add datafile '+DGDATA02/racdb/datafile/mytbs02.dbf' size 1g;

alter system archive log current;

select current_scn from v$database;
615468

alter system archive log current;

$ scp 2_19_1037709071.dbf 2_20_1037709071.dbf 1_32_1037709071.dbf 1_33_1037709071.dbf 10.206.132.245:/oradata02/archlog

RMAN> catalog start with '/oradata02/archlog';

SQL> shutdown immediate;

SQL> startup mount;

RMAN> recover database until scn 615468;

creating datafile file number=7 name=+DGDATA02/racdb/datafile/mytbs02.dbf
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 04/16/2020 08:24:27
RMAN-20505: create datafile during recovery
ORA-01119: error in creating database file '+DGDATA02/racdb/datafile/mytbs02.dbf'
ORA-17502: ksfdcre:3 Failed to create file +DGDATA02/racdb/datafile/mytbs02.dbf
ORA-15001: diskgroup "DGDATA02" does not exist or is not mounted
ORA-15077: could not locate ASM instance serving a required diskgroup
ORA-29701: unable to connect to Cluster Synchronization Service
RMAN-11003: failure during parse/execution of SQL statement: alter database recover logfile '/oradata02/archlog/2_19_1037709071.dbf'
ORA-00283: recovery session canceled due to errors
ORA-01274: cannot add datafile '+DGDATA02/racdb/datafile/mytbs02.dbf' - file could not be created

此時，我們加入參數
SQL> alter system set db_file_name_convert='+DGDATA01/racdb/datafile','/oradata01/racdb/datafile','+DGDATA02/racdb/datafile','/oradata02/racdb/datafile' scope=spfile; 
SQL> shutdown immediate;
SQL> startup mount;
RMAN> recover database until scn 615468;

Starting recover at 16-APR-20
using target database control file instead of recovery catalog
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=67 device type=DISK
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 04/16/2020 08:29:08
RMAN-06094: datafile 7 must be restored

RMAN> restore datafile 7;

Starting restore at 16-APR-20
using channel ORA_DISK_1

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of restore command at 04/16/2020 08:29:44
RMAN-06085: must use SET NEWNAME command to restore datafile /oracle/11204/dbs/UNNAMED00007

run{
    SET NEWNAME FOR DATAFILE '/oracle/11204/dbs/UNNAMED00007' TO '/oradata02/racdb/datafile/mytbs02.dbf';
    restore datafile 7;
}

datafile 7 is already restored to file /oradata02/racdb/datafile/mytbs02.dbf
restore not done; all files read only, offline, or already restored
Finished restore at 16-APR-20

進入系統，發現/oradata02/racdb/datafile/mytbs02.dbf已經存在，此時，我們只需要修改數據文件即可

SQL> alter database rename file '/oracle/11204/dbs/UNNAMED00007' to '/oradata02/racdb/datafile/mytbs02.dbf';


此時再次測試
alter tablespace mytbs add datafile '+DGDATA02/racdb/datafile/mytbs03.dbf' size 1g;

alter system archive log current;

select current_scn from v$database;
616851

alter system archive log current;

$ scp 2_21_1037709071.dbf 2_22_1037709071.dbf 1_34_1037709071.dbf 1_35_1037709071.dbf 10.206.132.245:/oradata02/archlog

RMAN> catalog start with '/oradata02/archlog';

RMAN> recover database until scn 616851;

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 04/16/2020 08:55:11
RMAN-20505: create datafile during recovery
ORA-01119: error in creating database file '+DGDATA02/racdb/datafile/mytbs03.dbf'
ORA-17502: ksfdcre:3 Failed to create file +DGDATA02/racdb/datafile/mytbs03.dbf
ORA-15001: diskgroup "DGDATA02" does not exist or is not mounted
ORA-15077: could not locate ASM instance serving a required diskgroup
ORA-29701: unable to connect to Cluster Synchronization Service
RMAN-11003: failure during parse/execution of SQL statement: alter database recover logfile '/oradata02/archlog/1_34_1037709071.dbf'
ORA-00283: recovery session canceled due to errors
ORA-01274: cannot add datafile '+DGDATA02/racdb/datafile/mytbs03.dbf' - file could not be created

run{
    SET NEWNAME FOR DATAFILE '/oracle/11204/dbs/UNNAMED00008' TO '/oradata02/racdb/datafile/mytbs03.dbf';
    restore datafile 8;
}

SQL> alter database rename file '/oracle/11204/dbs/UNNAMED00008' to '/oradata02/racdb/datafile/mytbs03.dbf';

RMAN> recover database until scn 616851;

通过实验发现，在主库全备后增加数据文件在恢复的时候由于目录不对应，会报错

此时决定加入另外一个参数alter system set standby_file_management=auto,看恢复的时候是否可以自动创建对应的文件

SQL> alter system set db_file_name_convert='+DGDATA01/racdb/datafile','/oradata01/racdb/datafile','+DGDATA02/racdb/datafile','/oradata02/racdb/datafile' scope=spfile; 
SQL> alter system set standby_file_management=auto;
SQL> shutdown immediate;
SQL> startup mount;
RMAN> catalog start with '/oradata02/archlog';
RMAN> recover database until scn 616851;

全备是在经过试验之前的，也就是没有mytbs02.dbf，mytbs03.dbf,mytbs04.dbf，增加两个db_file_name_convert，standby_file_management参数后，恢复过程没有报错，因此这两个参数应该在恢复之前加上。