Oracle RMAN 恢复数据库到不同主机（一）

一、RMAN 备份的内容

RMAN做数据库全备时包含了 数据文件、归档日志、控制文件和参数文件和备份日志，如下：

arch_20160223_08qukp2t_1_1  arch_20160223_0bqukp92_1_1  ctl_c-3234695588-20160223-01  rmanbak-20160223-0852.log  scp_20160223_09qukp2u_1_1  scp_20160223_0aqukp2u_1_1

二、测试环境

OS：CentOS release 6.4 (Final)

Database：Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production

原平台与当前平台环境一致，但是oracle数据库目录结构不一致。

三、开始恢复

1、  确认原数据库的DBID（通过RMAN的备份日志，或者通过RMAN备份的控制文件名来识别），同时确认一下原数据库的实例名；

2、 将RMAN备份的内容拷贝到目标数据库上；

3、 设置好环境变量：

[oracle@dg1 ~]$ export NLS_DATA_FORMAT='YYYY-MM-DD HH24:MI:SS'
[oracle@dg1 ~]$ export ORACLE_SID=scp
[oracle@dg1 ~]$ rman target /

Recovery Manager: Release 12.1.0.2.0 - Production on Fri May 6 22:18:52 2016

4、装载数据库并进行恢复

    A、装载数据库

RMAN> set dbid=3234695588;

executing command: SET DBID

RMAN> startup nomount;

startup failed: ORA-01078: failure in processing system parameters
LRM-00109: could not open parameter file '/u01/app/oracle/product/12.1.0.2/dbs/initscp.ora'

starting Oracle instance without parameter file for retrieval of spfile
Oracle instance started

  B、先恢复spfile文件（或者是pfile文件，可以修改各项参数），因为在spfile 文件中包含了控制文件的位置

RMAN> restore spfile to pfile '$ORACLE_HOME/dbs/initscp.ora' from '/home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01';

OR：

RMAN> restore spfile to '$ORACLE_HOME/dbs/spfilescp.ora' from '/home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01';

C、根据pfile文件中的相关参数去修改本地系统目录，或者修改这个pfile文件来匹配本地系统中的路径和目录，我们选择后者

需要修改：oracle_base、*.audit_file_dest、*.control_files、*.db_recovery_file_dest、*.db_recovery_file_dest_size、*.diagnostic_dest、*.log_archive_dest_1、

memory_target、undo_tablespace等。并在当前系统中创建好相关的目录和权限。

[oracle@dg1 dbs]$ vim initscp.ora 

scp.__data_transfer_cache_size=0
scp.__db_cache_size=822083584
scp.__java_pool_size=16777216
scp.__large_pool_size=33554432
scp.__oracle_base='/u01/app/oracle'#ORACLE_BASE set from environment
scp.__pga_aggregate_target=654311424
scp.__sga_target=1241513984
scp.__shared_io_pool_size=50331648
scp.__shared_pool_size=301989888
scp.__streams_pool_size=0
*.audit_file_dest='/u01/app/oracle/admin/scp/adump'
*.audit_trail='db'
*.compatible='12.1.0.2.0'
*.control_files='/u01/app/oracle/oradata/scp/control01.ctl','/u01/app/oracle/fast_recovery_area/scp/control02.ctl'
*.db_block_size=8192
*.db_domain=''
*.db_name='scp'
*.db_recovery_file_dest='/u01/app/oracle/fast_recovery_area'
*.db_recovery_file_dest_size=4560m
*.diagnostic_dest='/u01/app/oracle'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=scpXDB)'
*.log_archive_dest_1='LOCATION=/OracleArch'
*.memory_target=1800m
*.open_cursors=300
*.processes=300
*.remote_login_passwordfile='EXCLUSIVE'
*.undo_tablespace='UNDOTBS1'

D、从修改后的pfile文件启动数据库，进行控制文件的恢复

RMAN> shutdown abort;
RMAN> startup nomount pfile='$ORACLE_HOME/dbs/initscp.ora';
RMAN> restore controlfile from '/home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01';

Starting restore at 06-MAY-16
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=243 device type=DISK

channel ORA_DISK_1: restoring control file
channel ORA_DISK_1: restore complete, elapsed time: 00:00:01
output file name=/home/OracleData/scp/control01.ctl
output file name=/home/OracleArch/fast_recovery_area/control02.ctl
Finished restore at 06-MAY-16

E、控制文件恢复后就可以mount数据库了

RMAN> alter database mount;

Statement processed
released channel: ORA_DISK_1

F、至此，控制文件已经恢复，数据库已经mount，所有的RMAN配置参数均已设置，您应该验证路径以确保它们适用于该主机。

RMAN> show all;

RMAN configuration parameters for database with db_unique_name SCP are:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 7 DAYS;
CONFIGURE BACKUP OPTIMIZATION OFF; # default
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # default
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '/orabackup/RmanBackupSet/20160223/ctl_%F';
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET; # default
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
CONFIGURE COMPRESSION ALGORITHM 'BASIC' AS OF RELEASE 'DEFAULT' OPTIMIZE FOR LOAD TRUE ; # default
CONFIGURE RMAN OUTPUT TO KEEP FOR 7 DAYS; # default
CONFIGURE ARCHIVELOG DELETION POLICY TO NONE; # default
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/u01/app/oracle/product/12.1.0.2/dbs/snapcf_scp.f'; # default

G、为了能让RMAN找到恢复文件的位置，我们有两种途径可以实现：一是修改RMAN配置以符合当前备份文件所在位置，其次是将RMAN备份文件拷贝到配置文件中设定的位置  （需要参考RMAN的备份日志）。在第一个方法中，为了让RMAN知道备份文件位置 /home/OracleBack/rmanbak ，我们使用catalog命令:

RMAN> catalog start with '/home/OracleBack/rmanbak';

Starting implicit crosscheck backup at 06-MAY-16
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=243 device type=DISK
Crosschecked 4 objects
Finished implicit crosscheck backup at 06-MAY-16

Starting implicit crosscheck copy at 06-MAY-16
using channel ORA_DISK_1
Finished implicit crosscheck copy at 06-MAY-16

searching for all files in the recovery area
cataloging files...
no files cataloged

searching for all files that match the pattern /home/OracleBack/rmanbak

List of Files Unknown to the Database
=====================================
File Name: /home/OracleBack/rmanbak/scp_20160223_09qukp2u_1_1
File Name: /home/OracleBack/rmanbak/arch_20160223_0bqukp92_1_1
File Name: /home/OracleBack/rmanbak/scp_20160223_0aqukp2u_1_1
File Name: /home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01
File Name: /home/OracleBack/rmanbak/arch_20160223_08qukp2t_1_1

Do you really want to catalog the above files (enter YES or NO)? yes
cataloging files...
cataloging done

List of Cataloged Files
=======================
File Name: /home/OracleBack/rmanbak/scp_20160223_09qukp2u_1_1
File Name: /home/OracleBack/rmanbak/arch_20160223_0bqukp92_1_1
File Name: /home/OracleBack/rmanbak/scp_20160223_0aqukp2u_1_1
File Name: /home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01
File Name: /home/OracleBack/rmanbak/arch_20160223_08qukp2t_1_1

H、对备份集做交叉检查，否则还原数据库时可能会报错

RMAN> crosscheck backup;

using channel ORA_DISK_1
crosschecked backup piece: found to be 'EXPIRED'
backup piece handle=/orabackup/RmanBackupSet/20160223/arch_20160223_08qukp2t_1_1 RECID=8 STAMP=904553565
crosschecked backup piece: found to be 'AVAILABLE'
backup piece handle=/home/OracleBack/rmanbak/arch_20160223_08qukp2t_1_1 RECID=16 STAMP=911172456
crosschecked backup piece: found to be 'EXPIRED'
backup piece handle=/orabackup/RmanBackupSet/20160223/scp_20160223_0aqukp2u_1_1 RECID=9 STAMP=904553567
crosschecked backup piece: found to be 'AVAILABLE'
backup piece handle=/home/OracleBack/rmanbak/scp_20160223_0aqukp2u_1_1 RECID=14 STAMP=911172456
crosschecked backup piece: found to be 'EXPIRED'
backup piece handle=/orabackup/RmanBackupSet/20160223/scp_20160223_09qukp2u_1_1 RECID=10 STAMP=904553567
crosschecked backup piece: found to be 'AVAILABLE'
backup piece handle=/home/OracleBack/rmanbak/scp_20160223_09qukp2u_1_1 RECID=12 STAMP=911172456
crosschecked backup piece: found to be 'EXPIRED'
backup piece handle=/orabackup/RmanBackupSet/20160223/arch_20160223_0bqukp92_1_1 RECID=11 STAMP=904553762
crosschecked backup piece: found to be 'AVAILABLE'
backup piece handle=/home/OracleBack/rmanbak/arch_20160223_0bqukp92_1_1 RECID=13 STAMP=911172456
crosschecked backup piece: found to be 'AVAILABLE'
backup piece handle=/home/OracleBack/rmanbak/ctl_c-3234695588-20160223-01 RECID=15 STAMP=911172456
Crosschecked 9 objects

I、通过控制文件获得表空间及数据文件列表

RMAN> report schema;

RMAN-06139: WARNING: control file is not current for REPORT SCHEMA
Report of database schema for database with db_unique_name SCP

List of Permanent Datafiles
===========================
File Size(MB) Tablespace           RB segs Datafile Name
---- -------- -------------------- ------- ------------------------
1    0        SYSTEM               ***     /u01/app/oracle/oradata/scp/system01.dbf
2    0        ZYTK_AC              ***     /u01/app/oracle/oradata/scp/zytk_ac01.dbf
3    0        SYSAUX               ***     /u01/app/oracle/oradata/scp/sysaux01.dbf
4    0        UNDOTBS1             ***     /u01/app/oracle/oradata/scp/undotbs01.dbf
5    0        EXAMPLE              ***     /u01/app/oracle/oradata/scp/example01.dbf
6    0        USERS                ***     /u01/app/oracle/oradata/scp/users01.dbf
7    0        ZYTK_AC              ***     /u01/app/oracle/oradata/scp/zytk_ac02.dbf
8    0        ZYTK_ID              ***     /u01/app/oracle/oradata/scp/zytk_id01.dbf
9    0        ZYTK_ID              ***     /u01/app/oracle/oradata/scp/zytk_id02.dbf
10   0        ZYTK_OP              ***     /u01/app/oracle/oradata/scp/zytk_op01.dbf
11   0        ZYTK_OP              ***     /u01/app/oracle/oradata/scp/zytk_op02.dbf
12   0        ZYTK_TEST01          ***     /u01/app/oracle/oradata/scp/zytk_test01.dbf

List of Temporary Files
=======================
File Size(MB) Tablespace           Maxsize(MB) Tempfile Name
---- -------- -------------------- ----------- --------------------
1    60       TEMP                 32767       /u01/app/oracle/oradata/scp/temp01.dbf

注意：restore 的时候不会对temp 表空间进行restore。所以等restore 之后，我们需要手工创建temp表空间。

    J、为了将数据文件恢复到不同于原来的目录结构中，必须编写脚本，重新命名数据文件的位置，最后进行全库的恢复。

  （如果数据文件路径和原来的相同则直接还原数据库）

run{
set newname for datafile 1 to '/home/OracleData/scp/system01.dbf';
set newname for datafile 2 to '/home/OracleData/scp/zytk_ac01.dbf';
set newname for datafile 3 to '/home/OracleData/scp/sysaux01.dbf';
set newname for datafile 4 to '/home/OracleData/scp/undotbs01.dbf';
set newname for datafile 5 to '/home/OracleData/scp/example01.dbf';
set newname for datafile 6 to '/home/OracleData/scp/users01.dbf';
set newname for datafile 7 to '/home/OracleData/scp/zytk_ac02.dbf';
set newname for datafile 8 to '/home/OracleData/scp/zytk_id01.dbf';
set newname for datafile 9 to '/home/OracleData/scp/zytk_id02.dbf';
set newname for datafile 10 to '/home/OracleData/scp/zytk_op01.dbf';
set newname for datafile 11 to '/home/OracleData/scp/zytk_op02.dbf';
set newname for datafile 12 to '/home/OracleData/scp/zytk_test01.dbf';
restore database;
switch datafile all;  
}

对switch datafile all命令的说明：

--对于nocatalog 模式下，rman备份的信息是保存在控制文件里的，包括数据文件的路径信息。 这里的  switch datafile all  的作用，就是更新控制文件里的信息。如果不更新控制文件的话，则进行recover操作时还在原来的位置找文件。

K、还原数据库以后，进行 recover 操作

RMAN> recover database;

Starting recover at 06-MAY-16
using channel ORA_DISK_1

starting media recovery

channel ORA_DISK_1: starting archived log restore to default destination
channel ORA_DISK_1: restoring archived log
archived log thread=1 sequence=148
channel ORA_DISK_1: reading from backup piece /home/OracleBack/rmanbak/arch_20160223_0bqukp92_1_1
channel ORA_DISK_1: piece handle=/home/OracleBack/rmanbak/arch_20160223_0bqukp92_1_1 tag=ZYTK_DB_FULL
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:01
archived log file name=/home/OracleArch/ArchiveLog/1_148_899483175.dbf thread=1 sequence=148
unable to find archived log
archived log thread=1 sequence=149
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 05/06/2016 23:53:58
RMAN-06054: media recovery requesting unknown archived log for thread 1 with sequence 149 and starting SCN of 3507749