oracle 学习之DG的搭建

1.配置过程

• 确认为归档模式

SQL> select log_mode from v$database;
 
LOG_MODE
------------
ARCHIVELOG

• 配置归档日志存放路径（/home/oracle/arch提前创建）

主库：
SQL>alter system set log_archive_dest_1='location=/home/oracle/arch
db_unique_name=ora11g';

备库：
SQL> alter system set log_archive_dest_2='service=s74
db_unique_name=bei';

• 将主库、备库加载入DG中

SQL> alter system set log_archive_config='dg_config=(ora11g,bei)';
System altered.
SQL> alter system set standby_file_management=auto;
System altered.

• 开启强制记录日志功能

• 拷贝全库至备库

查看文件存放位置（日志文件、数据文件、控制文件都在此目录下）：
SQL> select * from v$logfile;
 
拷贝重做日志文件、数据文件至从库：
SQL>scp   *  172.16.150.66:/home/oracle/standby
 
生成备库的控制文件，并拷贝至备库：
SQL> alter database create standby controlfile as '/home/oracle/bei.ctl';
[oracle@localhost ora11g]$scp /home/oracle/1.ctl   172.16.150.74:/home/oracle/standby
 
拷贝参数文件、口令文件至备库：
SQL> show parameter spfile;
[oracle@oracle dbs]$ scp /u01/oracle/product/11.2.0/dbs /spfileora11g.ora 172.16.150.74:/home/oracle/standby
[oracle@oracle dbs]$ scp /u01/oracle/product/11.2.0/dbs / orapwora11g/172.16.150.74:/home/oracle/standby
 
将参数文件、口令文件拷贝至$ORACLE_HOME/dbs

• 主库配置tnsnames.ora

[oracle@oracle network]$ cd $ORACLE_HOME/network/admin
 
ORACLE =
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = oracle.com)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = oracle.com)
    )
  )
 
s74=
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.150.74)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = bei)
    )
  )

• 修改备库中的参数文件

ora11g.__db_cache_size=281018368
ora11g.__java_pool_size=4194304
ora11g.__large_pool_size=4194304
ora11g.__oracle_base='/u01/oracle'#ORACLE_BASE set from environment
ora11g.__pga_aggregate_target=301989888
ora11g.__sga_target=452984832
ora11g.__shared_io_pool_size=0
ora11g.__shared_pool_size=150994944
ora11g.__streams_pool_size=0
*.audit_file_dest='/home/oracle/standby/adump' #修改路径并创建目录
*.audit_trail='db'
*.compatible='11.2.0.0.0'
*.control_files='/home/oracle/standby/bei.ctl'#修改路径并创建目录
*.db_block_size=8192
*.db_domain='com'
*.db_name='oracle'
*.db_recovery_file_dest='/home/oracle/standby/fast_recovery_area' #修改路径并创建目录
*.db_recovery_file_dest_size=4322230272
*.diagnostic_dest='/u01/app/oracle'  #修改路径并创建目录
*.dispatchers='(PROTOCOL=TCP) (SERVICE=oracleXDB)'
*.local_listener='LISTENER_ORACLE'
*.log_archive_config='dg_config=(oracle,bei)'
*.log_archive_dest_1='location=/home/oracle/standby/arch db_unique_name=bei'
*.memory_target=754652928
*.open_cursors=300
*.processes=150
*.remote_login_passwordfile='EXCLUSIVE'
*.standby_file_management='AUTO'
*.undo_tablespace='UNDOTBS1'
db_unique_name=bei  #添加
db_file_name_convert='/u01/oracle/oradata/oracle','/home/oracle/standby' #添加
log_file_name_convert='/u02/oracle/oradata/oracle/','/home/oracle/standby' #添加

• 启动备库

重新生成spfile：
SQL> create spfile from pfile;
 
启动库：
SQL> alter database mount standby database;
Database altered.
SQL> select database_role from v$database;
DATABASE_ROLE
----------------
PHYSICAL STANDBY

• 启动监听

• 测试联通

自连：sqlplus  sys/oracle@172.16.150.74:1521/bei as sysdba
连接对方库：sqlplus  sys/123456@172.16.150.40:1521/ora11g as sysdba

• 备库开启实时归档日志功能

alter database recover managed standby database disconnect from session;

• 主库切换到备库

SQL> select database_role,switchover_status  from v$database;
 
DATABASE_ROLE     SWITCHOVER_STATUS
---------------- --------------------
PRIMARY      TO STANDBY
alter database commit to switchover to standby with session
shutdown

• 备库切换到主库

SQL> select database_role,switchover_status  from v$database;
DATABASE_ROLE     SWITCHOVER_STATUS
---------------- --------------------
PHYSICAL STANDBY TO PRIMARY（SESSIONS ACTIVE）

alter database commit to switchover to primary with session
shutdown

2.遇到的问题

• 数据库完成主库、备库之间的切换，再次启动主库时，出现：

SQL> archive log list
Database log mode           Archive Mode
Automatic archival           Enabled
Archive destination           /home/oracle/standby/arch
Oldest online log sequence     58
Next log sequence to archive   58
Current log sequence           58

SQL> select sequence# from v$log;

SEQUENCE#
  ----------
  58
  0
  0

原因：没有进行日志切换

• 备库变成主库后，切换日志，从库的sequence号并没有发生改变

　　　　主库如何相互切换

　　　　如何实现redo_log的同步

　　　　

3.其它问题

• db_name service_name unique_name之间的关系

SQL> show parameter name;
NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
db_file_name_convert                 string
db_name                              string     oracle
db_unique_name                       string     oracle
global_names                         boolean     FALSE
instance_name                        string     ora11g
lock_name_space                      string
log_file_name_convert                string
processor_group_name                 string
service_names                        string     oracle.com

db_name:
1.在DG中db_name一样而db_unique_name不一致
2.在RAC中各节点db_name一样而instance_name不一样
db_unique_name：
1.在DataGuard中，主备库拥有相同的DB_NAME，为了区别，就必须有不同的DB_UNIQUE_NAME
instance_name:
1.可以和db_name不一致
service_name
1.客户端连接数据库时采用服务名

• tnsnames.ora

s74=       #
  (DESCRIPTION =
    (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.150.74)#服务器IP
    (PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = bei) #数据库服务名
    )
  )

• db_domain的作用