搭建 Oracle Data Guard
用 rman duplicate 进行创建 Standby 库
ip 192.168.0.191 192.168.0.192
db_name moxi moxi
instance_name moxi moxi
db_unique_name mth atlas
service_names moxi moxi
tns alias first second
-------------------------
1 启用 Force Logging
-------------------------
将 Primary 数据库置为 Force Logging 模式。通过下列语句，看状态：
select database_role, force_logging from v$database;
DATABASE_ROLE FOR
---------------- ---
PRIMARY NO
修改模式
alter database force logging;
取消 Force logging 模式：
alter database no force logging;
-----------------------------
2 创建 standby redo logs
-----------------------------
在最大保护和最高可用性模式，还有 real-time apply时，必须为 Standby 数据库配置 Standby Redo logs。
有 2 点注意的地方：
每个 standby redo log 文件的大小必须大于等于 Primary 库的 online redo 的大小。最好是一样大。standby redo log group 要比 primary 的 redo group 至少多一组。
这里要注意 Thread，每个 Thread 都有对应的所有的组。一般情况下，单实例只有 1 个 thread，那么单实例到 standby 只需要比现在的组多一个就可以了。
如果是 RAC 到 DG，RAC 的每个节点都可以看成是一个 thread，这种情况下，standby 端的 redo log group 就是：（每线程的日志组数 + 1）* 最大线程数。
比如有 2 个线程，每个线程有 3 组，standby redo 就需要（3+1）* 2 = 8 组。
线程查看命令：
select group#, thread#, bytes, members from v$log;
GROUP# THREAD# BYTES MEMBERS
---------- ---------- ---------- ----------
1 1 52428800 1
2 1 52428800 1
3 1 52428800 1
select thread#, groups from v$thread;
THREAD# GROUPS
---------- ----------
1 3
set line 200
col member for a80
set pagesize 9999
select group#, status, type, member from v$logfile order by 1,4;
GROUP# STATUS TYPE MEMBER
---------- ------- ------- --------------------------------------------------
1 ONLINE /data1/app/oracle/oradata/moxi/redo01.log
2 ONLINE /data1/app/oracle/oradata/moxi/redo02.log
3 ONLINE /data1/app/oracle/oradata/moxi/redo03.log
Standby Redo Log 对于单线程：
alter database add standby logfile thread 1 group 4 ('/data1/app/oracle/oradata/moxi/st_1_4.log') size 50m;
alter database add standby logfile thread 1 group 5 ('/data1/app/oracle/oradata/moxi/st_1_5.log') size 50m;
alter database add standby logfile thread 1 group 6 ('/data1/app/oracle/oradata/moxi/st_1_6.log') size 50m;
alter database add standby logfile thread 1 group 7 ('/data1/app/oracle/oradata/moxi/st_1_7.log') size 50m;
查看 redo log：
set line 200
col member for a80
set pagesize 9999
select group#, status, type, member from v$logfile order by 1,4;
----------------------------
3 设置主数据库初始化参数
----------------------------
首先查看下列参数：
col name for a50
col value for a50
set line 150
set pagesize 9999
select name, value from v$parameter
where name in
('db_name','db_unique_name','service_names','log_archive_config','log_archive_dest_1',
'log_archive_dest_2','log_archive_dest_state_1','log_archive_dest_state_2',
'standby_file_management','fal_client','fal_server','db_file_name_convert',
'log_file_name_convert','remote_login_passwordfile');
修改如下：
alter system set log_archive_config='dg_config=(mth,atlas)';
alter system set log_archive_dest_1='location=/data2/app/arch/valid_for=(all_logfiles,all_roles) db_unique_name=mth';
alter system set log_archive_dest_2='service=second valid_for=(online_logfile,primary_role) db_unique_name=atlas';
alter system set log_archive_dest_state_1=enable;
alter system set log_archive_dest_state_2=defer;
alter system set standby_file_management=auto;
alter system set fal_client=first;
alter system set fal_server=second;
alter system set db_file_name_convert='/data1/app/oracle/oradata/moxi/','/data1/app/oracle/oradata/moxi/' scope=spfile;
alter system set log_file_name_convert='/data1/app/oracle/oradata/moxi/','/data1/app/oracle/oradata/moxi/' scope=spfile;
------------------
4 设置归档模式
------------------
shutdown immediate;
startup mount;
alter database archivelog;
alter database open;
archive log list;
-----------------------------------------------------
5 Create a Control File for the Standby Database
-----------------------------------------------------
注意：如果用了 rman duplicate，那么步骤 5 创建 standby 控制文件，可以省略掉了。
------------------------------------------------------------------------------
查看 primary 数据库 control 文件信息：
col name for a60
select name from v$controlfile;
在 Primary 库上执行如下语句，为 Standby 数据库创建控制文件：
shutdown immediate;
startup mount;
alter database create standby controlfile as '/data2/app/control01.ctl';
alter database open;
-------------------------------------------------------
6 Create a Parameter File for the Standby Database
-------------------------------------------------------
在 Primary 上，创建 pfile 文件：
create pfile from spfile;
修改参数文件 pfile
*.control_files='/data1/app/oracle/oradata/moxi/control01.ctl','/data1/app/oracle/flash_recovery_area/moxi/control02.ctl'
*.db_name='moxi'
*.db_unique_name='ATLAS'
*.fal_client='SECOND'
*.fal_server='FIRST'
*.log_archive_config='dg_config=(mth,atlas)'
*.log_archive_dest_1='location=/data2/app/arch/valid_for=(all_logfiles,all_roles) db_unique_name=atlas'
*.log_archive_dest_2='service=first valid_for=(online_logfile,primary_role) db_unique_name=mth'
*.log_archive_dest_state_1='ENABLE'
*.log_archive_dest_state_2='ENABLE'
*.service_names='MOXI'
*.standby_file_management='AUTO'
备库
用 rman duplicate 创建 standby 库是一定要在备库的 pfile 文件添加下面内容。
*.log_file_name_convert='/data1/app/oracle/oradata/moxi/','/data1/app/oracle/oradata/moxi/'
*.db_file_name_convert='/data1/app/oracle/oradata/moxi/','/data1/app/oracle/oradata/moxi/'
---------------------------------------------------------------
7 Copy Files from the Primary System to the Standby System
---------------------------------------------------------------
将 rman 备份、密码文件、参数文件、控制文件都拷贝到备数据库中。
在备库中运行如下命令：
mkdir -p /data1/app/oracle/admin/moxi/adump
mkdir -p /data1/app/oracle/oradata/moxi
mkdir -p /data1/app/oracle/flash_recovery_area
mkdir -p /data1/app/oracle/flash_recovery_area/moxi
RMAN 备份 Primary 数据库，Standby 库用 duplicate 进行恢复，下面是在 Primary 主库操作：
run {
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate channel c3 type disk;
allocate channel c4 type disk;
allocate channel c5 type disk;
backup as compressed backupset format '/data2/app/backup/db_%U' database plus archivelog;
backup current controlfile for standby format='/data2/app/backup/control_%U';
release channel c1;
release channel c2;
release channel c3;
release channel c4;
release channel c5;
}
将 rman 备份全部拷贝到 standby 备库相同的目录下
$ cd /data1/app/backup
$ scp -r * oracle@node2:/data1/app/backup
在 Standby 数据库执行 rman 恢复
sqlplus / as sysdba
startup nomount pfile=/data1/app/oracle/product/11.2.0/dbhome_1/dbs/initmoxi.ora
开始 rman duplicate
注意说明：rman 后面的 target sys 用户是主数据库的，后面 auxiliary 和 / 指的是 Standby 数据库
$ rman target sys/123@first auxiliary /
duplicate target database for standby nofilenamecheck dorecover;
创建 Standby spfile 文件
sqlplus / as sysdba
create spfile from pfile='/data1/app/oracle/product/11.2.0/dbhome_1/dbs/initmoxi.ora';
------------------------------------------------------
8 Configure listeners for the primary and standby
------------------------------------------------------
配置监听器
$ netmgr
在主库和备库上分别执行：netmgr 命令配置 listener 和 tnsname.ora。
$ cat listener.ora
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(GLOBAL_DBNAME = moxi)
(ORACLE_HOME = /data1/app/oracle/product/11.2.0/dbhome_1)
(SID_NAME = moxi)
)
)
LISTENER =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.0.191)(PORT = 1521))
)
ADR_BASE_LISTENER = /data1/app/oracle
$ cat tnsnames.ora
FIRST =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.0.191)(PORT = 1521))
)
(CONNECT_DATA =
(SID = moxi)
)
)
SECOND =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.0.192)(PORT = 1521))
)
(CONNECT_DATA =
(SID = moxi)
)
)
----------------
9 DG 主备库启动
----------------
standby 监听：lsnrctl start
第一次开启，因为第一次只有 pfile，所以要指定 pfile 开启 standby 备库，然后就可以在 standby 生成 spfile 了：
startup pfile='/data1/app/oracle/product/11.2.0/dbhome_1/dbs/initmoxi.ora' nomount
create spfile from pfile;
alter database mount standby database;
以后直接执行如下，启动备援资料库，语法如下：----------->“等同于 startup mount”
SQL> connect sys/oracle as sysdba
SQL> startup nomount
SQL> alter database mount standby database
下面查看 standby 备库的状态
select status from v$instance;
set line 200
select open_mode, protection_mode, protection_level, database_role, switchover_status, force_logging from v$database;
执行下面就可以 applied redo 日志，如果是物理恢复就是 MRP。
alter database recover managed standby database disconnect from session;
alter database recover managed standby database disconnect;
主库监听器：lsnrctl start
主库： startup
----------------
10 开始同步
----------------
下面是在 Primary 数据库上执行如下。
alter system set log_archive_dest_state_2=enable scope=both;
在 Primary 库执行如下，进行日志的切换。
alter system switch logfile;
select status,error from v$archive_dest where dest_id=2;
在 standby 备库 执行如下：
alter database open;
alter database recover managed standby database using current logfile disconnect;
SQL> select open_mode from v$database;
OPEN_MODE
--------------------
READ ONLY WITH APPLY
------------
11 验证
------------
下面查看 standby 备库的状态
select status from v$instance;
set line 200
select open_mode, protection_mode, protection_level, database_role, switchover_status, force_logging from v$database;
下面是在 primary 主库执行如下：
col dest_name for a20
col destination for a20
select dest_name, destination, status, error from v$archive_dest;
上面 error 列没有显示，说明 redo 传输成功。
下面在 primary 主库执行日志切换：
SQL> alter system switch logfile;
下面在 primary 和 standby 分别查看日志是否相等，如果相等说明传输成功：
SQL> select max(sequence#) from v$archived_log;
下面查看 redo 传送状态，在 standby 中执行：
set line 200
col name for a50
select name, sequence#, applied, first_time, next_time from v$archived_log order by sequence#;
---------------------
最大性能转换到最大可用
---------------------
shutdown immediate
startup mount
alter systemset log_archive_dest_2='service=second LGWR SYNC AFFIRM valid_for=(online_logfile,primary_role) db_unique_name=atlas';
alter database set standby database to maximize availability;
alter database open;
select open_mode, protection_mode from v$database;
------------
主、备库切换
------------
1. 在主库
alter database commit to switchover to physical standby with session shutdown;
shutdown immediate
2. 在备库
alter database commit to switchover to primary with session shutdown;
shutdown immediate
startup
3. 在主库
startup mount
4. 检查数据库角色
set line 200
select name, open_mode, protection_mode, protection_level, database_role, switchover_status, force_logging from v$database;
-------------------
failover 非正常切换
-------------------
failover 切换一般是 Primary 数据库发生故障后的切换，这种情况是 Standby 数据库发挥其作用的情况。这种切换发生后，可能会造成数据的丢失。而且这个过程不是可逆的，Data Guard 环境会被破坏。
由于 Primary 数据库已经无法启动，所以 Failover 切换所需的条件并不多，只要检查 standby 是否运行在最大保护模式下，如果是的话，需要将其置为最大性能模式，否则切换到 primary 角色也无法启动。
1. 查看是否有日志gap,没有应用的日志：
sql> select unique thread#, max(sequence#) over(partition by thread#) last from v$archived_log;
sql> select thread#, low_sequence#, high_sequence# from v$archive_gap;
如果有，则拷贝过来并且注册
sql> alter database register physical logfile ' 路径';
重复查看直到没有应用的日志：
2. 然后停止应用归档：
sql> alter database recover managed standby database cancel;
3. 下面将 standby 数据库切换为 priamry 数据库：
sql> alter database recover managed standby database finish;
或
sql> alter database recover managed standby database finish force;
sql> select database_role from v$database;
database_role
----------------
physical standby
sql> alter database commit to switchover to primary;
sql> alter database open; 或者 shutdown immediate + startup
检查数据库是否已经切换成功：
sql> select database_role from v$database;
database_role
----------------
primary
至此，Failover 切换完成。这个时候应该马上对新的 Primary 数据库进行备份。