2.datagurd搭建(最大性能模式)

思路：

1.需要相同版本(最好环境也一样)

2.主库有实例，从库上要求没有实例

3.主库需要归档

　　3.1查看主库是否是force loging 模式

　　select force_logging from v$database;  若为no，则执行alter database force logging

　　3.2是否开启归档

　　archive log list 查看：

　　若没有归档，将主库启动到mount状态

　　然后执行:alter database archivelog;

4.主库和备库都配置tns监听　

orcl_p =
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 主库ip)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = orcl)
    )
  )

orcl_s =
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 备库ip)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = orcl)
    )
  )

5.主库修改初始化参数(重启才能生效)

sql>alter system set DB_UNIQUE_NAME=orcl_p scope=spfile;
sql>alter system set LOG_ARCHIVE_CONFIG='DG_CONFIG=(orcl_s,orcl_p)' scope=both;
sql>alter system set LOG_ARCHIVE_DEST_1='LOCATION=/oradata/orcl/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) db_unique_name=orcl_p' scope=spfile;
sql>alter ssytem set LOG_ARCHIVE_DEST_2='SERVICE=orcl_s  ASYNC VALID_FOR=(ONLINE_LOGFILE,PRIMARY_ROLE) db_unique_name=orcl_s' scope=spfile;
sql>alter system set LOG_ARCHIVE_DEST_STATE_2=DEFER scope=both;
sql>alter system set LOG_ARCHIVE_MAX_PROCESSES=4 scope=4;
sql>alter system set FAL_SERVER=orcl_s scope=both;
sql>alter system set FAL_CLIENT=orcl_p scope=both;
sql>alter system set STANDBY_FILE_MANAGEMENT=AUTO scope=both;
sql>alter system set log_file_name_convert='/oradata/orcl/redo','/oradat/orcl/redo/'scope=spfile;

6.主库全备份(这里注意一定要将主库启动到mount状态而不是open状态才能开始全备)

run {
backup database format '/oracle/backup/db_%t_%s_%U.bak' tag='db_full01';'

backup archivelog all format '/oracle/backup/arch_%t_%s_%U.bak' tag='arch01';'
}

7.在主库上中创建standby controlfile(这个文件最后是要传到备库中的相应位置上的，这里主库也是mount状态)

sql>alter dabase create standby controlfile as '/oracle/backup/control01.ctl'

8.创建备库初始化参数(这里根据实际路径进行修改) 然后再根据主库的参数文件进行修改

这里主要说一下要注意的几个参数：

*.control_files='/xxxx/xxxx/control01.ctl','/xxxx/xxxx/control02.ctl'
*.DB_UNIQUE_NAME=orcl_s
*.LOG_ARCHIVE_CONFIG='DG_CONFIG=(orcl_s,orcl_p)' 
LOG_ARCHIVE_DEST_1='LOCATION=/oradata/orcl/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) db_unique_name=orcl_p'
LOG_ARCHIVE_DEST_2='SERVICE=orcl_s  ASYNC VALID_FOR=(ONLINE_LOGFILE,PRIMARY_ROLE) db_unique_name=orcl_s'
LOG_ARCHIVE_DEST_STATE_2=DEFER
LOG_ARCHIVE_MAX_PROCESSES=4
FAL_SERVER=orcl_s 
FAL_CLIENT=orcl_p 
STANDBY_FILE_MANAGEMENT=AUTO 
set log_file_name_convert='/oradata/orcl/redo','/oradat/orcl/redo/'

9.将主库的密码文件到备库中

10.在备库中创建参数文件中相应的路径(这里指的是目录，用mkdir进行创建，且注意权限)

11.将在主库中生成的standby controlfile传输到备库的相应目录中

12.这里将第6步中的数据文件和归档日志文件的备份集拷贝到从库中(这里注意目录和权限)

13.在主备库中分别添加standby redo log(该日志一般会比online redolog多一个)

alter database add standby logfile thread 1 group 8('/oradata/orcl/redo/redo10.log') size 200M;

....

14.将备库启动到mount状态，然后执行如下命令

rman>restore database from tag='db_full01';
rman>restore archivelog all;
rman>recove database;

15.然后将备库开启open状态

alter database open read only

16.开启日志日志传输(主库上操作)

sql>alter system set log_archive_dest_state_2=enable scope=both;

17.开启日志应用进程(备库上操作)

sql>alter database recover managed Standby database disconnect from session no timeout;

验证同步是否正常：

18.在主库和备库分别查询v$ARCHIVED_LOG的最大序列号，将所得结果相比，相同表名同步正常

sql>select max(sequence#),thread# from v$archived_log group by thread#

　对于使用redo apply同步方式，可以在备库通过MRPn状态进一步确认同步情况

sql>select m.PROCESS,m.STATUS  from v$managed_standby m where m.PROCESS like 'MRP%';

     如果使用redo apply 方式同步，同步过程中应有MRPn进程存在，如果redo apply正常，MRPn状态为applying_log

总结：

　　好了，以上备份需要注意一定要将主库启动到mount状态才能进行全备，生成standby controlfile 文件也一定要在mount状态才行，否则，在后续的备份恢复中，备库不能进行open状态(由于datafile的scn和controlfile的scn不一致导致的)，这一点很重要！！！