物理standby部署

环境：

主机ip	dg角色	db_name	db_unique_name
192.168.56.110	primary	SHPD	SHPD
192.168.56.111	standby	SHPD	SHST

主机必要条件check

select NAME,LOG_MODE,FORCE_LOGGING from v$database;
alter database force logging;

前提

1. SHPD和SHST都已安装好oracle软件，配置好环境变量，listener.ora，tnsnames.ora

2. SHPD数据库已创建好，启动到open，SHST启动到nomount

3. 目录结构同clone db

4. init.ora也使用clonedb的，db_unique_name需要修改，其余的下面在具体配置

tnsname.ora内容：

shpd =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.56.110)(PORT = 1521))
)
(CONNECT_DATA =
(SERVICE_NAME = SHPD)
)
)

shst =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.56.111)(PORT = 1521))
)
(CONNECT_DATA =
(SERVICE_NAME = SHST)
)
)

配置主库参数

--参数check
col name format a30;
col value format a200;
set linesize 1000;
set pagesize 1000;
select name,value from v$parameter
where name in('fal_server',
'fal_client',
'standby_file_management',
'db_file_name_convert',
'log_file_name_convert',
'db_name',
'db_unique_name',
'log_archive_dest_1',
'log_archive_dest_2',
'log_archive_config')
order by name;
--主备db和log路径转换
select 'datafile',name from v$datafile
union all
select 'tempfile',name from v$tempfile
union all
select 'logifle',member from V$LOGFILE where TYPE='ONLINE';

ALTER SYSTEM SET db_unique_name='SHPD' SCOPE=SPFILE;
ALTER SYSTEM SET db_file_name_convert='/u01/oradata/SHST/datafile','/u01/oradata/SHPD/datafile','/u01/oradata/SHST/tempfile','/u01/oradata/SHPD/tempfile' scope=spfile;
ALTER SYSTEM SET log_file_name_convert='/u01/oradata/SHST/onlinefile','/u01/oradata/SHPD/onlinefile' scope=spfile;
ALTER SYSTEM SET log_archive_config='DG_CONFIG=(SHPD,SHST)' SCOPE=BOTH;
ALTER SYSTEM SET log_archive_dest_1='location=/u01/oradata/SHPD/arch valid_for=(all_logfiles,all_roles) db_unique_name=SHPD' SCOPE=BOTH;
ALTER SYSTEM SET log_archive_dest_2='SERVICE=SHST LGWR ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=SHST' SCOPE=BOTH;
alter system set LOG_ARCHIVE_DEST_STATE_1=enable scope=both;
alter system set LOG_ARCHIVE_DEST_STATE_2=enable scope=both;
ALTER SYSTEM SET fal_server='SHST' SCOPE=BOTH;
ALTER SYSTEM SET fal_client='SHPD' SCOPE=BOTH;
ALTER SYSTEM SET standby_file_management=AUTO SCOPE=BOTH;

配置从库参数

ALTER SYSTEM SET db_unique_name='SHST' scope=spfile;
ALTER SYSTEM SET db_name='SHPD' scope=spfile;
ALTER SYSTEM SET db_file_name_convert='/u01/oradata/SHPD/datafile','/u01/oradata/SHST/datafile','/u01/oradata/SHPD/tempfile','/u01/oradata/SHST/tempfile' scope=spfile;
ALTER SYSTEM SET log_file_name_convert='/u01/oradata/SHPD/onlinefile','/u01/oradata/SHST/onlinefile' scope=spfile;
ALTER SYSTEM SET log_archive_config='DG_CONFIG=(SHPD,SHST)'  SCOPE=BOTH;
ALTER SYSTEM SET log_archive_dest_1='location=/u01/oradata/SHST/arch valid_for=(all_logfiles,all_roles) db_unique_name=SHST' SCOPE=BOTH;
ALTER SYSTEM SET log_archive_dest_2='SERVICE=SHPD LGWR ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=SHPD' SCOPE=BOTH;
alter system set LOG_ARCHIVE_DEST_STATE_1=enable scope=both;
alter system set LOG_ARCHIVE_DEST_STATE_2=enable scope=both;
ALTER SYSTEM SET fal_server='SHPD' SCOPE=BOTH;
ALTER SYSTEM SET fal_client='SHST' SCOPE=BOTH;
ALTER SYSTEM SET standby_file_management=AUTO SCOPE=BOTH;

添加日志文件

根据primary的online redo log大小和组数，按规则添加Standby online log，添加的group的编号最好有预留，方便扩展

select * from V$LOGFILE;
select * from v$log;
select * from V$STANDBY_LOG;
/*主库添加Standby online log*/
alter database add standby logfile
group 11 '/u01/oradata/SHPD/onlinefile/standbyredo04.log' size 50M,
group 12 '/u01/oradata/SHPD/onlinefile/standbyredo05.log' size 50M,
group 13 '/u01/oradata/SHPD/onlinefile/standbyredo06.log' size 50M,
group 14 '/u01/oradata/SHPD/onlinefile/standbyredo07.log' size 50M;
/*备库添加Standby online log*/
alter database add standby logfile group 10('/u01/app/oracle/oradata/intedb/standby_redo10.log') size 50M;
alter database add standby logfile group 11('/u01/app/oracle/oradata/intedb/standby_redo11.log') size 50M;
alter database add standby logfile group 12('/u01/app/oracle/oradata/intedb/standby_redo12.log') size 50M;
alter database add standby logfile group 13('/u01/app/oracle/oradata/intedb/standby_redo13.log') size 50M;
alter database add standby logfile group 14('/u01/app/oracle/oradata/intedb/standby_redo14.log') size 50M;

备库同步主库数据

生成standby controlfile，并复制到standby目录（忽略）

--ALTER DATABASE CREATE STANDBY CONTROLFILE AS '/tmp/control01.ctl';
--cp /tmp/control01.ctl /tmp/control02.ctl
--scp /tmp/control01.ctl /tmp/control02.ctl oracle@192.168.56.111:/u01/oradata/SHST/controlfile/

Standby数据恢复，Standby上执行

##Method1 在线还原数据，需要配置配置静态监听，STANDBY上执行

rman target sys@SHPD auxiliary sys@SHST
duplicate database for standby from active database nofilenamecheck;

 ##Method2利用备份集，有target连接

先备份，duplicate复制库

rman target sys@SHPD auxiliary /
duplicate target database for standby;

 ##Method3利用备份集，无target连接

先备份，duplicate复制库

rman auxiliary /
duplicate database for standby backup location '/data/backup' nofilenamecheck;

开启日志应用

--打开Standby数据库
alter database open read only;
--开启实时日志应用，启用Real-Time Apply；开启MRP进程，停止redo应用
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT FROM SESSION;
--关闭MRP进程（停止应用日志），开启redo应用
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
--使用arch同步模式
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;

---

Monitoring DG

DG状态监控

--查看DB状态
select instance_name,status from v$instance;
--验证测试data guard
ALTER SYSTEM ARCHIVE LOG  CURRENT;
select name,controlfile_type,open_mode,protection_mode,
protection_level,database_role,SWITCHOVER_STATUS,db_unique_name,
guard_status,PRIMARY_DB_UNIQUE_NAME
from v$database;
SELECT * FROM v$dataguard_stats;
SELECT NAME,applied FROM v$archived_log order by COMPLETION_TIME desc;
select * from V$ARCHIVE_DEST;
select * from v$archive_dest_status;
select * from v$archive_gap;
select * from v$archived_log;
SELECT * FROM v$standby_log;
select * from v$managed_standby;
select * from V$LOG;
select * from v$logfile;
select * from V$LOG_HISTORY; 

监控redo传输服务

Step 1-查看主库最新的归档日志，如果dest2的最新归档日志和dest1的相同，表示已传输到dest2

SELECT dest_id,MAX(SEQUENCE#), THREAD#
FROM V$ARCHIVED_LOG
WHERE RESETLOGS_CHANGE# = (SELECT MAX(RESETLOGS_CHANGE#) FROM V$ARCHIVED_LOG)
GROUP BY dest_id,THREAD#;

Step 2-查看备库最新的归档日志，如果dest1的最新归档日志和主库上的相同，表示接受成功

SELECT dest_id,dest_name,DESTINATION, STATUS, ARCHIVED_THREAD#, ARCHIVED_SEQ#
FROM V$ARCHIVE_DEST_STATUS
WHERE STATUS <> 'DEFERRED' 
AND STATUS <> 'INACTIVE';

Step 3-确认已在目标上接收到归档重做日志文件

SELECT LOCAL.THREAD#, LOCAL.SEQUENCE# 
FROM (SELECT THREAD#, SEQUENCE# 
FROM V$ARCHIVED_LOG WHERE DEST_ID=1)LOCAL 
WHERE LOCAL.SEQUENCE# NOT IN(SELECT SEQUENCE# FROM V$ARCHIVED_LOG WHERE DEST_ID=2 
AND THREAD# = LOCAL.THREAD#);

Step 4-Trace the progression of redo transmitted to a redo transport destination.LOG_ARCHIVE_TRACE

  

---

switchover步骤

switchover包含两个步骤，首先主库被转换成STANDBY，然后STANDBY被转换成主库，确认没有用户连接到数据库

/*check1-主库确认没有日志缺口，应该返回VALID和NO GAP*/
select STATUS, GAP_STATUS from V$ARCHIVE_DEST_STATUS; 
/*check2-确认备库的临时文件和主库一样*/
SELECT * FROM v$tempfile;
/*check3-确认所有重做日志都已在备库应用，查询备库*/
select NAME, VALUE, DATUM_TIME from V$DATAGUARD_STATS;
/*check4-主数据库的状态检查，分别确认主库和备库都可以进行角色切换，查询主库-返回TO STANDBY,SESSIONS ACTIVE，那么主库就可以进行切换；备库返回TO PRIMARY,SESSIONS ACTIVE,NOT ALLOWED，就可以切换，备库如果返回SWITCHOVER LATENT或SWITCHOVER PENDING，就要去检查告警日志，看有什么问题，一般是需要应用一些日志，如果是需要应用日志的话，在备库执行如下命令：*/
select name,open_mode,database_role,protection_mode,SWITCHOVER_STATUS From v$database;
--recover standby database using backup controlfile;
/*先将primary切换为standby
如果SWITCHOVER_STATUS为TO STANDBY，直接switchover；
如果SWITCHOVER_STATUS为SESSIONS ACTIVE，，需要加with session shutdown，
实例会关闭
说明：PRIMARY进行转换完毕后，查看状态会变成RECOVERY NEEDED
*/
alter database commit to switchover to physical standby;
alter database commit to switchover to physical standby with session shutdown;
startup mount;
/*备库切为主库
如果SWITCHOVER_STATUS为TO_PRIMARY，直接switchover；
如果SWITCHOVER_STATUS为SESSIONS ACTIVE，先停MRP，switchover加上with session shutdown*/
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL; 
alter database commit to switchover to primary;
alter database commit to switchover to primary with session shutdown; 
alter database open; 
/*在切换后的standby上开启实时日志应用*/ 
alter database open read only; 
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT FROM SESSION; 

---

 failover步骤

一般情况下failover都是表示primary数据库瘫痪，最起码也是起不来了，因此这种类型的切换基本上不需要primary数据库做什么操作

/*standby上先查看有没有日志间隙*/
SELECT * FROM v$archive_gap;
/*如果存在间隙（online redolog），还能将主库启动到挂载（mount）状态，可以试着将刷新未传输的online日志到备库，如能刷新，故障转移可能不会丢失任何数据*/
alter system flush redo to 'JED'; 
/*如果存在间隙，为了尽可能少的丢失数据，应该尝试将所有归档日志应用，应该将主库上的所有归档日志复制到备库，可能有一些归档日志已经在备库存在，但这样能有尽可能多的归档日志，然后要解决备库中的任何日志缺口，首先，复制所有主库的归档到备库，并在数据库中注册它们*/
alter database register physical logfile '&logfile_path_name'; 
/*在备库停止日志应用*/
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL; 
/*结束应用任何日志，一旦执行了这个命令，备库就必须转换为主库，要不就得重建*/
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH;
/*转换备库为主库*/
alter database commit to switchover to primary with session shutdown;
alter database open;
/*检查*/
select SWITCHOVER_STATUS from V$DATABASE;
/*获得原备库转换为主库时的SCN*/
SELECT to_char(STANDBY_BECAME_PRIMARY_SCN) from V$DATABASE;

切换为快照standby

步骤：

1. 停止Redo Apply
2. 数据库处于mount，不是open
3. Ensure that a fast recovery area has been configured. It is not necessary for flashback database to be enabled.
4. 执行sql，完成转换:

• • ALTER DATABASE CONVERT TO SNAPSHOT STANDBY;

--查看当前备用数据库的角色，设置FRA,停止mrp
select open_mode,database_role,flashback_on from v$database;
alter database recover managed standby database cancel;
alter system set db_recovery_file_dest_size=500m scope=both;
alter system set db_recovery_file_dest='/u01/oradata/SHPD/fra' scope=both;
--关闭并启动到mount
shutdown immediate;
startup mount;
--转换物理standby到快照standby
alter database convert to snapshot standby;
select status from v$instance;
alter database open;
select open_mode,database_role from v$database;
--oracle创建了还原点
select name,storage_size from v$restore_point;

切换回物理standby

--查看当前备用数据库的角色
select open_mode,database_role from v$database;
--关闭并启动到mount
shutdown immediate;
startup mount;
--转换快照standby到物理standby
alter database convert to physical standby;
shutdown immediate;
startup;

保护模式切换 

/*主备保护模式切换，要注意的是log_archive_dest参数的切换，由于最大保护模式与最大可用模式的重做归档进程、网络传输与磁盘写选项是一样的--所以可以直接更改模式就可以了，不需要先修改参数信息*/
--物理standby先停用mrp
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
--主库启动到mount
shutdown immediate;
startup mount;
--切换到高可用或高保护
ALTER SYSTEM SET log_archive_dest_2='SERVICE=SHPD SYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=SHPD' SCOPE=BOTH;
alter database set standby database to maximize protection;
alter database set standby database to maximize availability;
alter database open;
--切换到高性能模式
ALTER SYSTEM SET log_archive_dest_2='SERVICE=SHPD ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=SHPD' SCOPE=BOTH;
alter database set standby database to maximize performance;