单实例数据库迁移到RAC

一.准备工作

因为要完全恢复，所以在拿到除数据库的备份片外，还要拿备份后时间跑出的归档文件，当然，还要拿到最后一个curren日志组的归档。当然如果做不完全恢复的话，不需要current日志组的归档也是可以的。

注：因为我的本次实验是在rac下，已经有数据库后，删除后模拟做的迁移，所以有些小细节可能不同，不过原理都是一样的。

1.环境：数据库 11.2.0.4 Linux：centos 6.5

2.RAC配置

--------------------------------------------------------------------------------
NAME           TARGET  STATE        SERVER                   STATE_DETAILS      
--------------------------------------------------------------------------------
Local Resources
--------------------------------------------------------------------------------
ora.DATA.dg
               ONLINE  ONLINE       hua1                                        
               ONLINE  ONLINE       hua2                                        
ora.FRA.dg
               ONLINE  ONLINE       hua1                                        
               ONLINE  ONLINE       hua2                                        
ora.LISTENER.lsnr
               ONLINE  ONLINE       hua1                                        
               ONLINE  ONLINE       hua2                                        
ora.asm
               ONLINE  ONLINE       hua1                     Started            
               ONLINE  ONLINE       hua2                     Started            
ora.gsd
               OFFLINE OFFLINE      hua1                                        
               OFFLINE OFFLINE      hua2                                        
ora.net1.network
               ONLINE  ONLINE       hua1                                        
               ONLINE  ONLINE       hua2                                        
ora.ons
               ONLINE  ONLINE       hua1                                        
               ONLINE  ONLINE       hua2                                        
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.LISTENER_SCAN1.lsnr
      1        ONLINE  ONLINE       hua1                                        
ora.LISTENER_SCAN2.lsnr
      1        ONLINE  ONLINE       hua2                                        
ora.LISTENER_SCAN3.lsnr
      1        ONLINE  ONLINE       hua2                                        
ora.cvu
      1        ONLINE  ONLINE       hua2                                        
ora.hua1.vip
      1        ONLINE  ONLINE       hua1                                        
ora.hua2.vip
      1        ONLINE  ONLINE       hua2                                        
ora.oc4j
      1        ONLINE  ONLINE       hua2                                        
ora.scan1.vip
      1        ONLINE  ONLINE       hua1                                        
ora.scan2.vip
      1        ONLINE  ONLINE       hua2                                        
ora.scan3.vip
      1        ONLINE  ONLINE       hua2  

二.目标端数据恢复

2.1 在节点1上创建所需目录

[oracle@hua1 ~]$ mkdir -p /u01/app/oracle/admin/PROD1/adump

[oracle@hua1 ~]$ mkdir -p /u01/app/oracle/archivelog

[oracle@hua1 ~]$ mkdir -p /u01/app/oracle/oradata/PROD1

[oracle@hua1 ~]$ mkdir -p /u01/app/oracle/backup

2.2 拷贝源数据库的备份片、归档文件和最后一个redo.log

2.3 新建参数文件 vim initPROD1.ora

PROD2.__db_cache_size=272629760
PROD1.__db_cache_size=310378496
PROD2.__java_pool_size=4194304
PROD1.__java_pool_size=4194304
PROD2.__large_pool_size=8388608
PROD1.__large_pool_size=71303168
PROD1.__oracle_base='/u01/app/oracle'#ORACLE_BASE set from environment
PROD2.__oracle_base='/u01/app/oracle'#ORACLE_BASE set from environment
PROD2.__pga_aggregate_target=335544320
PROD1.__pga_aggregate_target=310378496
PROD2.__sga_target=503316480
PROD1.__sga_target=528482304
PROD2.__shared_io_pool_size=0
PROD1.__shared_io_pool_size=0
PROD2.__shared_pool_size=209715200
PROD1.__shared_pool_size=134217728
PROD2.__streams_pool_size=0
PROD1.__streams_pool_size=0
*.audit_file_dest='/u01/app/oracle/admin/PROD1/adump'
*.audit_trail='db'
*.cluster_database=false
*.compatible='11.2.0.4.0'
*.control_files='+DATA/prod1/controlfile/current.258.991842525'#Restore Controlfile
*.db_block_size=8192
*.db_create_file_dest='+DATA'
*.db_domain=''
*.db_name='PROD1'
*.db_recovery_file_dest='+FRA'
*.db_recovery_file_dest_size=4621074432
*.diagnostic_dest='/u01/app/oracle'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=PRODXDB)'
PROD2.instance_number=2
PROD1.instance_number=1
*.log_archive_format='%t_%s_%r.dbf'
*.memory_target=838860800
*.open_cursors=300

注：等数据库恢复完毕后，再把luster_database设置为true。

2.4 用pfile文件创建spfile文件

2.4.1启动数据库

[oracle@hua1 ~]$ sqlplus / as sys