RAC11.2.0.4+单节点物理standby dataguard-rman恢复方式

一、rac主库配置
1.1、db_name、db_unique_name、instance_name、service_name
db_name：数据库名，多节点的数据库名相同，实例名不相同
db_unique_name：HA的应用，它不同于db_name，dg中要求主从库db_name相同；db_unique_name直接决定了service_name，即服务名
创建数据库时会指定全局数据库名和sid，sid直接决定了instance_name；监听中global_name和db_name可以不同

rac主库：db_name=amondbs，db_unique_name=amondbs_primary，ORACLE_SID=amondbs1,amondbs2，service=amondbs_pri_ser
单实例备库：db_name=amondbs，db_unique_name=amondbs_standby，ORACLE_SID=amondbs_phydb，service=amondbs_sta_ser

rac的数据库名相同，但是实例名不同，实例名（instance_name）来源于ORACLE_SID；默认本地服务名也称为动态注册监听默认服务名（service_names，也可以通过lsnrctl status查看）来源于db_unique_name、db_name；配置连接字符串时connect_data配置的service_name，service_name其实就是service_names的值

（0）修改db_unique_name和ORACLE_SID后，会自动创建$ORACLE_BASE/diag/rdbms/db_unique_name/ORACLE_SID/trace
（1）密码文件是以实例名为标识，即ORACLE_SID，与db_unique_name无关
（2）执行backup archivelog all将当前重做日志切换并归档，然后再做备份归档
（3）静态注册中的sid_name：数据库运行的实例名，必须和instance_name相同；global_dbname：可以是任意值，配置tns时service_name跟它相同
（4）一个instance_name可以对应多个service_name
（5）oracle用户安装oracle数据库，指定Global Database Name不会自动填充，它的值默认就是db_name、db_unique_name的值

1.2、rac主库必须置为归档模式
[email protected][+ASM1]/root$crsctl start crs
[email protected][amondbs1]/home/oracle$sqlplus / as sysdba
SQL> select name,open_mode,log_mode,force_logging from gv$database;

NAME      OPEN_MODE            LOG_MODE     FOR
--------- -------------------- ------------ ---
AMONDBS   READ WRITE           ARCHIVELOG   YES
AMONDBS   READ WRITE           ARCHIVELOG   YES

alter database archivelog;
alter database force logging;

1.3、rac主库必须置为force logging模式
SQL> select inst_id,name,log_mode,force_logging from gv$database;

   INST_ID NAME      LOG_MODE     FOR
---------- --------- ------------ ---
         1 AMONDBS   ARCHIVELOG   YES
         2 AMONDBS   ARCHIVELOG   YES

SQL> alter database force logging;

1.4、rac主库执行rman全备尤其是控制文件
[email protected][amondbs1]/oracle$mkdir /oracle/rman_backup
RMAN> 
run
{delete noprompt obsolete;
allocate channel c1 type disk MAXPIECESIZE 1G;
allocate channel c2 type disk MAXPIECESIZE 1G;
backup AS COMPRESSED BACKUPSET filesperset 2 format '/oracle/rman_backup/db_full_%T_%d_%t_%s_%p' database;
backup AS COMPRESSED BACKUPSET format '/oracle/rman_backup/arch_full_%T_%d_%t_%s_%p' archivelog all delete input;
backup current controlfile for standby format '/oracle/rman_backup/ctl_standby_%T_%d_%t_%s_%p';
release channel c1;
release channel c2;
}

1.5、rac主库创建从库的初始化参数文件到全备目录init+sid
SQL> show parameter pfile
SQL> create pfile='/oracle/rman_backup/initamondbs_phydb.ora' from spfile='+DATADG/amondbs/spfileamondbs.ora';  ##spfile应该指定磁盘组的spfile初始化参数文件

1.6、rac主库修改双节点共享的初始化参数文件，spfile是共享的，每个节点的pfile有指向spfile的路径
SQL> select name from v$datafile;

NAME
--------------------------------------------------------------------------------
+DATADG/system01.dbf
+DATADG/sysaux01.dbf
+DATADG/undotbs101.dbf
+DATADG/undotbs201.dbf
+DATADG/users.dbf
+DATADG/testtbs01

SQL> shutdown immediate;

##通过共享的spfile产生pfile，修改rac所有节点的参数，dg配置说明了*/oradata/db_unique_name创建目录
*.db_name='amondbs'
*.db_unique_name='amondbs_primary'
*.log_archive_config='dg_config=(amondbs_primary,amondbs_standby)'	
*.log_archive_dest_2='service=amondbs_sta_ser lgwr async affirm valid_for=(online_logfiles,primary_role) db_unique_name=amondbs_standby'
*.log_archive_dest_state_1=enable
*.log_archive_dest_state_2=defer
*.remote_login_passwordfile='EXCLUSIVE'
*.db_file_name_convert='/home/oracle/oracle/oradata/amondbs_standby','+DATADG/amondbs/datafile'  ##standby库oracle用户创建$ORACLE_BASE/oradata
*.log_file_name_convert='/home/oracle/oracle/oradata/amondbs_standby','+DATADG/amondbs'
*.fal_server=amondbs_standby
*.standby_file_managent=auto
还有角色切换后新角色的配置

##rac单个节点生成spfile即可
##关闭双节点数据库shutdown，启动后竟然使用修改的单节点pfile启动数据库，因为create pfile改变了SPFILE=，所以必须create spfile创建双节点共享的初始化参数文件、修改pfile
SQL> create spfile='+DATADG/amondbs/spfileamondbs.ora' from pfile='/oracle/rman_backup/initamondbs_phydb.ora';
[email protected][amondbs1]/oracle/db11g_11.2.0.4/dbs$vi initamondbs1.ora
SPFILE='+DATADG/amondbs/spfileamondbs.ora'
SQL> startup  
SQL> alter system set log_archive_dest_state_2=defer sid='*' scope=both;   

##此时服务名不再是默认的服务名，@amondbs_primary时报错，如果配置tnsnames，service_name根据db_unique_name
[email protected][amondbs1]/oracle/db11g_11.2.0.4/dbs$lsnrctl status
Service "+ASM" has 1 instance(s).
  Instance "+ASM1", status READY, has 1 handler(s) for this service...
Service "amondbs_primary" has 1 instance(s).
  Instance "amondbs1", status READY, has 1 handler(s) for this service...

[email protected][+ASM1]/root$sqlplus sys/a@amondbs_primary as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on Fri Aug 15 10:58:25 2014

Copyright (c) 1982, 2013, Oracle.  All rights reserved.

ERROR:
ORA-12154: TNS:could not resolve the connect identifier specified


注意：
（1）rac主库的*.db_unique_name='amondbs_primary'最好不要设置，保持默认避免scan listener起不来，因为tnsnames.ora中配置的服务名是amondbs；db_unique_name和tns服务名必须保持相同
（2）SQL> alter system set log_archive_dest_2='service=amondbs_sta_ser lgwr sync affirm valid_for=(online_logfiles,primary_role) db_unique_name=amondbs_standby' sid='*' scope=both;

1.6、rac主库修改口令文件，使双节点sys用户口令相同，dg从库的sys口令与双节点sys口令相同
sqlplus sys/[email protected]:1521/amondbs_primary as sysdba

1.7、rac主库配置tnsname.ora资料库，即配置amondbs_pri_ser、amondbs_sta_ser服务名，需要改双节点
##rac主库配置tnsname.ora资料库后，记得使用tnsping测试网络是否可用
[email protected][amondbs1]/home/oracle$vi $ORACLE_HOME/network/admin/tnsnames.ora
# tnsnames.ora Network Configuration File: /oracle/db11g_11.2.0.4/network/admin/tnsnames.ora
# Generated by Oracle configuration tools.
amondbs_pri_ser =
        (DESCRIPTION =
                (ADDRESS_LIST =
                     (LOAD_BALANCE = on)
                   (FAILOVER = on)
                   (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.4)(PORT=1521))
                   (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.6)(PORT=1521))
                )
                (CONNECT_DATA =
                        (SERVICE_NAME = amondbs_primary)
                        (FAILOVER_MODE = (METHOD = basic)(TYPE = select)(DELAY = 5)(RETRIES = 3))
                )
        )

amondbs_sta_ser =
        (DESCRIPTION =
                (ADDRESS_LIST =
                        (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.11)(PORT = 1521))
                )
                (CONNECT_DATA =
                        (SERVER = dedicated)
                        (SERVICE_NAME = amondbs_standby)
                )
        )

###测试一下
sqlplus system/123456@amondbs_pri_ser

1.8、为主库添加standby redo，以便主库切换为从库后使用；跟从库创建的standby redo 线程、日志组相同；路径取log_file_name_convert
alter database add standby logfile thread 1 group 5 ('+DATADG/amondbs/standbyredo01.log') size 50m,
group 6 ('+DATADG/amondbs/standbyredo02.log') size 50m,
group 7 ('+DATADG/amondbs/standbyredo03.log') size 50m;
alter database add standby logfile thread 2 group 8 ('+DATADG/amondbs/standbyredo04.log') size 50m,
group 9 ('+DATADG/amondbs/standbyredo05.log') size 50m,
group 10 ('+DATADG/amondbs/standbyredo06.log') size 50m;

SQL> select * from v$logfile;

    GROUP# STATUS  TYPE    MEMBER                                             IS_
---------- ------- ------- -------------------------------------------------- ---
         1         ONLINE  +DATADG/amondbs/redo01.log                         NO
         2         ONLINE  +DATADG/amondbs/redo02.log                         NO
         3         ONLINE  +DATADG/amondbs/redo03.log                         NO
         4         ONLINE  +DATADG/amondbs/redo04.log                         NO
         5         STANDBY +DATADG/amondbs/standbyredo01.log                  NO
         6         STANDBY +DATADG/amondbs/standbyredo02.log                  NO
         7         STANDBY +DATADG/amondbs/standbyredo03.log                  NO
         8         STANDBY +DATADG/amondbs/standbyredo04.log                  NO
         9         STANDBY +DATADG/amondbs/standbyredo05.log                  NO
        10         STANDBY +DATADG/amondbs/standbyredo06.log                  NO




注意：
（1）如果指定了db_unique_name，创建$ORACLE_BASE/diag/rdbms/db_unique_name目录/sid；$ORACLE_BASE/diag/rdbms/amondbs_standby/amondbs_phydb


二、standby从库配置
注意：主库和从库的SID都是amondbs，与db_unique_name不同

1.1、oracle用户创建目录
$ mkdir -p $ORACLE_BASE/oradata/amondbs_standby  ##db_unique_name
$ mkdir -p $ORACLE_BASE/fast_recovery_area
$ mkdir /home/oracle/oracle/admin/amondbs_standby/adump
$ mkdir /home/oracle/oracle/archivelog

1.2、将rac主库的备份拷贝到从库

1.3、从库修改初始化参数文件->启动oracle
在$ORACLE_HOME/dbs创建一个临时pfile，然后create spfile会自动生成带有sid的spfile
*.audit_file_dest='/home/oracle/oracle/admin/amondbs_standby/adump'
*.audit_trail='DB'
*.cluster_database=false
*.compatible='11.2.0.4.0'
*.control_files='/home/oracle/oracle/oradata/amondbs_standby/control01.ctl','/home/oracle/oracle/oradata/amondbs_standby/control02.ctl'
*.db_block_size=8192
*.db_domain=''
*.db_recovery_file_dest='/home/oracle/oracle/fast_recovery_area'
*.db_recovery_file_dest_size=6005194752
*.diagnostic_dest='/home/oracle/oracle/'  ##oracle基础目录
*.event=''
*.log_archive_format='%t_%s_%r.dbf'
*.memory_target=400000000
*.open_cursors=300
*.processes=150
*.undo_management='auto'
*.undo_tablespace='undotbs1'
*.db_name='amondbs'
*.db_unique_name='amondbs_standby'
*.log_archive_config='dg_config=(amondbs_primary,amondbs_standby)'
*.log_archive_dest_1='LOCATION=/home/oracle/oracle/archivelog'
*.log_archive_dest_2='service=amondbs_pri_ser lgwr async affirm valid_for=(online_logfiles,primary_role) db_unique_name=amondbs_primary'
*.log_archive_dest_state_1=enable
*.log_archive_dest_state_2=enable
*.remote_login_passwordfile='EXCLUSIVE'
*.db_file_name_convert='+DATADG/amondbs/datafile','/home/oracle/oracle/oradata/amondbs_standby'
*.log_file_name_convert='+DATADG/amondbs','/home/oracle/oracle/oradata/amondbs_standby'
*.fal_server=amondbs_primary
*.standby_file_managent=auto

SQL> create spfile from pfile
1.4、创建监听
$ lsnrctl status  ##没有监听
动态监听
$ lsnrctl status

LSNRCTL for Linux: Version 11.2.0.4.0 - Production on 15-JAN-2016 11:10:58

Copyright (c) 1991, 2013, Oracle.  All rights reserved.

Connecting to (ADDRESS=(PROTOCOL=tcp)(HOST=)(PORT=1521))
STATUS of the LISTENER
------------------------
Alias                     LISTENER
Version                   TNSLSNR for Linux: Version 11.2.0.4.0 - Production
Start Date                15-JAN-2016 11:10:15
Uptime                    0 days 0 hr. 0 min. 42 sec
Trace Level               off
Security                  ON: Local OS Authentication
SNMP                      OFF
Listener Log File         /home/oracle/oracle/diag/tnslsnr/dataguard/listener/alert/log.xml
Listening Endpoints Summary...
  (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=dataguard)(PORT=1521)))
Services Summary...
Service "amondbs_standby" has 1 instance(s).
  Instance "amondbs", status READY, has 1 handler(s) for this service...
The command completed successfully

本机tnsping成功，其他机器tnsping报错：TNS-12543: TNS:destination host unreachable，本机的防火墙在开着

配置listener.ora
LISTENER =
  (DESCRIPTION_LIST =
    (DESCRIPTION =
      ##(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1521))
      (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.0.51)(PORT = 1521))
    )
  )

ADR_BASE_LISTENER = /home/oracle/oracle

ENABLE_GLOBAL_DYNAMIC_ENDPOINT_LISTENER=ON 
#####仍然存在动态监听，可以远程使用sys登录
$ lsnrctl status

LSNRCTL for Linux: Version 11.2.0.4.0 - Production on 15-JAN-2016 13:09:44

Copyright (c) 1991, 2013, Oracle.  All rights reserved.

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.0.51)(PORT=1521)))
STATUS of the LISTENER
------------------------
Alias                     LISTENER
Version                   TNSLSNR for Linux: Version 11.2.0.4.0 - Production
Start Date                15-JAN-2016 13:07:56
Uptime                    0 days 0 hr. 1 min. 48 sec
Trace Level               off
Security                  ON: Local OS Authentication
SNMP                      OFF
Listener Parameter File   /home/oracle/oracle/product/11g/network/admin/listener.ora
Listener Log File         /home/oracle/oracle/diag/tnslsnr/dataguard/listener/alert/log.xml
Listening Endpoints Summary...
  (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=192.168.0.51)(PORT=1521)))
Services Summary...
Service "amondbs_standby" has 1 instance(s).
  Instance "amondbs", status READY, has 1 handler(s) for this service...
The command completed successfully

配置listener.ora，加SID_LIST，动态监听和静态监听同时存在  

$ lsnrctl start
#######使用tnsping进行测试
$ vi $ORACLE_HOME/network/admin/tnsnames.ora
amondbs_pri_ser =
        (DESCRIPTION =
                (ADDRESS_LIST =
                   (LOAD_BALANCE = on)
                   (FAILOVER = on)
                   (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.31)(PORT=1521))
                   (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.41)(PORT=1521))
                )
                (CONNECT_DATA =
                        (SERVICE_NAME = amondbs_primary)
                        (FAILOVER_MODE = (METHOD = basic)(TYPE = select)(DELAY = 5)(RETRIES = 3))
                )
        )

amondbs_sta_ser =
        (DESCRIPTION =
                (ADDRESS_LIST =
                        (ADDRESS = (PROTOCOL = tcp)(HOST = 192.168.0.51)(PORT = 1521))
                )
                (CONNECT_DATA =
                        (SERVER = dedicated)
                        (SERVICE_NAME = amondbs_standby)
                )
        )

1.5、备库创建密码文件和主库相同
orapwd file=$ORACLE_HOME/dbs/orapw$ORACLE_SID password=123456 entries=10 force=yes

最好把主库的密码文件拷贝到从库，重命名为orapw
sqlplus sys/123456@amondbs_sta_ser as sysdba从主库可以连接，但是system不能连接，原因：sys使用控制文件认证，普通用户使用数据文件认证，这不是使用
静态注册的一个原因

1.6、从库启动到nomount，sqlplus后根据ORACLE_SID找对应pfile，然后启动，根据ORACLE_SID切换到不同的实例
注意将主库rman的备份拷贝到从库的同一目录，不然备份是过期的，执行set newname报错ORA-01180: can not create datafile 1
$sqlplus / as sysdba
SQL> startup nomount

rman target /
RMAN> restore standby controlfile from '/oracle/rman_backup/ctl_standby_20160115_AMONDBS_901207070_50_1'; 
output file name=/home/oracle/oracle/oradata/amondbs_standby/control01.ctl
output file name=/home/oracle/oracle/oradata/amondbs_standby/control02.ctl
RMAN> alter database mount;
RMAN> list backup;
List of Backup Sets
===================
BS Key  Type LV Size       Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ ---------------
42      Full    15.28M     DISK        00:00:14     15-JAN-16      
        BP Key: 42   Status: AVAILABLE  Compressed: YES  Tag: TAG20160115T151616
        Piece Name: /oracle/rman_backup/db_full_20160115_AMONDBS_901206977_43_1

RMAN> restore database;

Starting restore at 15-JAN-16
Starting implicit crosscheck backup at 15-JAN-16
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=19 device type=DISK
Crosschecked 8 objects
Finished implicit crosscheck backup at 15-JAN-16

Starting implicit crosscheck copy at 15-JAN-16
using channel ORA_DISK_1
Finished implicit crosscheck copy at 15-JAN-16

searching for all files in the recovery area
cataloging files...
no files cataloged

using channel ORA_DISK_1

channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_DISK_1: restoring datafile 00002 to /home/oracle/oracle/oradata/amondbs_standby/sysaux.261.900422621
channel ORA_DISK_1: restoring datafile 00005 to /home/oracle/oracle/oradata/amondbs_standby/users.265.900422741
channel ORA_DISK_1: reading from backup piece /oracle/rman_backup/db_full_20160115_AMONDBS_901206977_43_1
channel ORA_DISK_1: piece handle=/oracle/rman_backup/db_full_20160115_AMONDBS_901206977_43_1 tag=TAG20160115T151616
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:25
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_DISK_1: restoring datafile 00003 to /home/oracle/oracle/oradata/amondbs_standby/undotbs1.262.900422677
channel ORA_DISK_1: restoring datafile 00004 to /home/oracle/oracle/oradata/amondbs_standby/undotbs2.264.900422721
channel ORA_DISK_1: reading from backup piece /oracle/rman_backup/db_full_20160115_AMONDBS_901206994_44_1
channel ORA_DISK_1: piece handle=/oracle/rman_backup/db_full_20160115_AMONDBS_901206994_44_1 tag=TAG20160115T151616
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:15
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
channel ORA_DISK_1: restoring datafile 00001 to /home/oracle/oracle/oradata/amondbs_standby/system.260.900422559
channel ORA_DISK_1: reading from backup piece /oracle/rman_backup/db_full_20160115_AMONDBS_901206977_42_1
channel ORA_DISK_1: piece handle=/oracle/rman_backup/db_full_20160115_AMONDBS_901206977_42_1 tag=TAG20160115T151616
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:35

1.9、接受日志信息，这里没有创建standby logfile，所以接受的是归档日志
SQL> ALTER SYSTEM SET log_archive_dest_state_2=enable SCOPE=BOTH SID='*';
此时主节点数据库alert日志如下：
******************************************************************
LGWR: Setting 'active' archival for destination LOG_ARCHIVE_DEST_2
******************************************************************
ALTER SYSTEM SET log_archive_dest_state_2='ENABLE' SCOPE=BOTH SID='*';
Fri Jan 15 16:43:10 2016
Archived Log entry 65 added for thread 1 sequence 42 ID 0x235d93c2 dest 1:

此时从库数据库alert日志如下：
Fri Jan 15 16:42:53 2016
Using STANDBY_ARCHIVE_DEST parameter default value as /home/oracle/oracle/archivelog
Fri Jan 15 16:42:55 2016
Primary database is in MAXIMUM PERFORMANCE mode
RFS[1]: Assigned to RFS process 4966
RFS[1]: No standby redo logfiles created
RFS[1]: Opened log for thread 1 sequence 43 dbid 593350850 branch 900422530
Fri Jan 15 16:42:57 2016
RFS[2]: Assigned to RFS process 4970
RFS[2]: Opened log for thread 2 sequence 24 dbid 593350850 branch 900422530
Fri Jan 15 16:42:58 2016
RFS[3]: Assigned to RFS process 4972
RFS[3]: Opened log for thread 2 sequence 22 dbid 593350850 branch 900422530
Fri Jan 15 16:42:59 2016
RFS[4]: Assigned to RFS process 4968
RFS[4]: Opened log for thread 2 sequence 23 dbid 593350850 branch 900422530
Archived Log entry 1 added for thread 2 sequence 24 rlc 900422530 ID 0x235d93c2 dest 2:
Archived Log entry 2 added for thread 2 sequence 22 rlc 900422530 ID 0x235d93c2 dest 2:
Archived Log entry 3 added for thread 2 sequence 23 rlc 900422530 ID 0x235d93c2 dest 2:
Fri Jan 15 16:43:00 2016
RFS[5]: Assigned to RFS process 4974
RFS[5]: Opened log for thread 1 sequence 41 dbid 593350850 branch 900422530
Archived Log entry 4 added for thread 1 sequence 41 rlc 900422530 ID 0x235d93c2 dest 2:
Fri Jan 15 16:43:08 2016
Primary database is in MAXIMUM PERFORMANCE mode
RFS[6]: Assigned to RFS process 4978
RFS[6]: No standby redo logfiles created
RFS[6]: Opened log for thread 2 sequence 26 dbid 593350850 branch 900422530
Fri Jan 15 16:43:11 2016
RFS[7]: Assigned to RFS process 4982
RFS[7]: Opened log for thread 1 sequence 42 dbid 593350850 branch 900422530
Fri Jan 15 16:43:11 2016
RFS[8]: Assigned to RFS process 4984
RFS[8]: Opened log for thread 2 sequence 25 dbid 593350850 branch 900422530
Archived Log entry 5 added for thread 1 sequence 42 rlc 900422530 ID 0x235d93c2 dest 2:
Archived Log entry 6 added for thread 2 sequence 25 rlc 900422530 ID 0x235d93c2 dest 2:


1.9、启用redo应用
SQL> alter database recover managed standby database disconnect from session;

#####正常信息：如果删除归档目录的日志，执行redo应用时，FAL会从主库fetch日志
##Media Recovery Log /home/oracle/oracle/archivelog/2_22_900422530.dbf
##Error opening /home/oracle/oracle/archivelog/2_22_900422530.dbf
##Attempting refetch
##Media Recovery Waiting for thread 2 sequence 22
##Fetching gap sequence in thread 2, gap sequence 22-22
##FS[15]: Allowing overwrite of partial archivelog for thread 2 sequence 22

SQL> alter database recover managed standby database cancel;
SQL> alter database open;  ##可以打开数据库，之前打不开，ORA-10458: standby database requires recovery
SQL> select open_mode,log_mode from v$database;  ###控制文件是standby

OPEN_MODE            LOG_MODE
-------------------- ------------
READ ONLY            ARCHIVELOG

1.10、打开日志应用后，从库MRP启动，创建了online redo 
#MRP0 started with pid=30, OS id=5258 
#MRP0: Background Managed Standby Recovery process started (amondbs)
#Serial Media Recovery started
#Managed Standby Recovery not using Real Time Apply
#Waiting for all non-current ORLs to be archived...
#All non-current ORLs have been archived.
#Errors in file /home/oracle/oracle/diag/rdbms/amondbs_standby/amondbs/trace/amondbs_mrp0_5258.trc:
#ORA-00313: open failed for members of log group 1 of thread 1
#ORA-00312: online log 1 thread 1: '/home/oracle/oracle/oradata/amondbs_standby/redo01.log'
#ORA-27037: unable to obtain file status
#Linux-x86_64 Error: 2: No such file or directory
#Additional information: 3
#Clearing online redo logfile 1 /home/oracle/oracle/oradata/amondbs_standby/redo01.log
#Clearing online log 1 of thread 1 sequence number 45

SQL> select group#,members,thread#,sequence#,bytes/1024/1024,status,archived from v$log;

    GROUP#    MEMBERS    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS           ARC
---------- ---------- ---------- ---------- --------------- ---------------- ---
         1          1          1         45              50 CLEARING         YES
         2          1          1         46              50 CURRENT          YES
         3          1          2         29              50 CURRENT          YES
         4          1          2         28              50 CLEARING         YES

1.11、创建standby日志，创建的大小要和主库online redo保持一致
创建的原则：
① Standby Redo Log 的文件大小与主库 online redo log 文件大小相同
② Standby Redo Log 日志文件组的个数依照下面的原则进行计算：
Standby redo log组数公式 >= (每个instance日志组个数+1)*instance个数
例如在我的环境中，只有一个节点，这个节点有三组redo，所以Standby redo log组数公式>=(3+1)*1  == 4所以需要创建4组Standby redo log
③ 每一日志组为了安全起见，可以做多路镜像

从库创建standby redo日志--------
alter database add standby logfile thread 1 group 5 size 50m,group 6 size 50m,group 7 size 50m;  ##默认创建到db_recovery_file_dest
alter database drop standby logfile group 5,group 6,group 7;

SQL> alter database add standby logfile thread 1 group 5 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo01.log') size 50m,group 6 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo02.log') size 50m,group 7 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo03.log') size 50m;
SQL> alter database add standby logfile thread 2 group 8 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo04.log') size 50m,group 9 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo05.log') size 50m,group 10 ('/home/oracle/oracle/oradata/amondbs_standby/standbyredo06.log') size 50m;

SQL> select group#,thread#,sequence#,bytes/1024/1024,status,archived from v$standby_log;

    GROUP#    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS     ARC
---------- ---------- ---------- --------------- ---------- ---
         5          1          0              50 UNASSIGNED YES
         6          1          0              50 UNASSIGNED YES
         7          1          0              50 UNASSIGNED YES
         8          2          0              50 UNASSIGNED YES
         9          2          0              50 UNASSIGNED YES
        10          2          0              50 UNASSIGNED YES

主库，主库的日志信息-------------
SQL> select group#,members,thread#,sequence#,bytes/1024/1024,status,archived from v$log;

SQL> select group#,members,thread#,sequence#,bytes/1024/1024,status,archived from v$log;

    GROUP#    MEMBERS    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS           ARC
---------- ---------- ---------- ---------- --------------- ---------------- ---
         1          1          1         45              50 INACTIVE         YES
         2          1          1         46              50 CURRENT          NO
         3          1          2         29              50 CURRENT          NO
         4          1          2         28              50 INACTIVE         YES

SQL> alter system archive log current;
SQL> select group#,members,thread#,sequence#,bytes/1024/1024,status,archived from v$log;

    GROUP#    MEMBERS    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS           ARC
---------- ---------- ---------- ---------- --------------- ---------------- ---
         1          1          1         47              50 CURRENT          NO
         2          1          1         46              50 ACTIVE           YES
         3          1          2         29              50 ACTIVE           YES
         4          1          2         30              50 CURRENT          NO
从库####
SQL> select group#,thread#,sequence#,bytes/1024/1024,status,archived from v$standby_log;

    GROUP#    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS     ARC
---------- ---------- ---------- --------------- ---------- ---
         5          1         47              50 ACTIVE     YES
         6          1          0              50 UNASSIGNED YES
         7          1          0              50 UNASSIGNED YES
         8          2         30              50 ACTIVE     YES
         9          2          0              50 UNASSIGNED YES
        10          2          0              50 UNASSIGNED YES

SQL> alter system archive log current;
SQL> select group#,members,thread#,sequence#,bytes/1024/1024,status,archived from v$log;

    GROUP#    MEMBERS    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS           ARC
---------- ---------- ---------- ---------- --------------- ---------------- ---
         1          1          1         47              50 ACTIVE           YES
         2          1          1         48              50 CURRENT          NO
         3          1          2         31              50 CURRENT          NO
         4          1          2         30              50 ACTIVE           YES

从库####
SQL> select group#,thread#,sequence#,bytes/1024/1024,status,archived from v$standby_log;

    GROUP#    THREAD#  SEQUENCE# BYTES/1024/1024 STATUS     ARC
---------- ---------- ---------- --------------- ---------- ---
         5          1         48              50 ACTIVE     YES
         6          1          0              50 UNASSIGNED YES
         7          1          0              50 UNASSIGNED YES
         8          2         31              50 ACTIVE     YES
         9          2          0              50 UNASSIGNED YES
        10          2          0              50 UNASSIGNED YES
总结：主库的current online redo和standby库的active状态的standby redo 日志序列号相同，证明redo传输服务启用了实时传输。

1.12、备库实时应用redo
11gR2开始，物理standby数据库在open状态也能应用日志，称为active dataguard
SQL> alter database recover managed standby database using current logfile disconnect from session; ##启动MRPn进程，进行real time apply，通过读取standby redo 进行恢复

SQL> select dbid,switchover_status,db_unique_name,database_role,open_mode,current_scn from v$database;

      DBID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 DATABASE_ROLE    OPEN_MODE            CURRENT_SCN
---------- -------------------- ------------------------------ ---------------- -------------------- -----------
 593350850 NOT ALLOWED          amondbs_standby                PHYSICAL STANDBY READ ONLY WITH APPLY      637067

1.13、创建表空间测试文件同步
主库操作创建表空间，从库验证-----
set line 200
col FILE_NAME for a50
select tablespace_name,file_name,bytes/1024/1024 as total_space from dba_data_files;

TABLESPACE_NAME                FILE_NAME                                          TOTAL_SPACE
------------------------------ -------------------------------------------------- -----------
SYSTEM                         +DATADG/amondbs/datafile/system.260.900422559              700
SYSAUX                         +DATADG/amondbs/datafile/sysaux.261.900422621              600
UNDOTBS1                       +DATADG/amondbs/datafile/undotbs1.262.900422677            200
UNDOTBS2                       +DATADG/amondbs/datafile/undotbs2.264.900422721            200
USERS                          +DATADG/amondbs/datafile/users.265.900422741                 5

create tablespace cuishoutest  
  datafile '+DATADG/amondbs/datafile/cuishoutest01.dbf' size 100m
  autoextend on next 1m maxsize unlimited
  logging
  extent management local autoallocate 
  blocksize 8k
  segment space management auto
  flashback on;

主库操作删除表空间，从库验证----
drop tablespace cuishoutest including contents and datafiles;

1.14、测试dataguard switchover功能
####检查从库是否归档模式、force logging模式----
select name,open_mode,log_mode,force_logging from gv$database;

NAME                                                                                                 OPEN_MODE            LOG_MODE     FOR
---------------------------------------------------------------------------------------------------- -------------------- ------------ ---
AMONDBS                                                                                              READ ONLY WITH APPLY ARCHIVELOG   YES


####关闭日志传输
SQL> alter system set log_archive_dest_state_2=defer scope=both;

（1）停止rac节点2
SQL> select thread#,inst_id,instance_name,status from gv$instance;

   THREAD#    INST_ID INSTANCE_NAME    STATUS
---------- ---------- ---------------- ------------
         2          2 amondbs2         OPEN
         1          1 amondbs1         OPEN

SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          2 SESSIONS ACTIVE      amondbs_primary                MAXIMUM PERFORMANCE  PRIMARY          YES READ WRITE                649547
 593350850          1 SESSIONS ACTIVE      amondbs_primary                MAXIMUM PERFORMANCE  PRIMARY          YES READ WRITE                649547
 
（2）rac节点1切换为从库
SQL> show parameter instance_name 
SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 TO STANDBY           amondbs_primary                MAXIMUM PERFORMANCE  PRIMARY          YES READ WRITE                650442

###主库切为从库，同时关闭实例，需要启动，open后数据库处于read only
SQL> alter database commit to switchover to physical standby with session shutdown;

SQL> startup
Physical standby database opened for read only access.

SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 TO PRIMARY           amondbs_primary                MAXIMUM PERFORMANCE  PHYSICAL STANDBY YES READ ONLY                 670661

（3）从库切换为主库，此时单节点主库只使用线程1的日志组
SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 TO PRIMARY           amondbs_standby                MAXIMUM PERFORMANCE  PHYSICAL STANDBY YES READ ONLY WITH APPLY      670661


SQL> alter database commit to switchover to primary;
######数据库和切换后的主库不一致，打开日志应用
*
ERROR at line 1:
ORA-16139: media recovery required
Database not available for switchover
  Incomplete recovery SCN:0:670662 archive SCN:0:690868
Database not available for switchover
  Incomplete recovery SCN:0:670662 archive SCN:0:690868

SQL> alter database recover managed standby database using current logfile disconnect from session; 

#####不cancel日志应用，直接切换
SQL> alter database commit to switchover to primary;
Role Change: Canceling MRP - no more redo to apply
MRP0: Background Media Recovery process shutdown (amondbs)
Backup controlfile written to trace file /home/oracle/oracle/diag/rdbms/amondbs_standby/amondbs/trace/amondbs_ora_6312.trc
SwitchOver after complete recovery through change 690868
Online log /home/oracle/oracle/oradata/amondbs_standby/redo01.log: Thread 1 Group 1 was previously cleared
Online log /home/oracle/oracle/oradata/amondbs_standby/redo02.log: Thread 1 Group 2 was previously cleared
Online log /home/oracle/oracle/oradata/amondbs_standby/redo03.log: Thread 2 Group 3 was previously cleared
Online log /home/oracle/oracle/oradata/amondbs_standby/redo04.log: Thread 2 Group 4 was previously cleared
Standby became primary SCN: 690866

SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 NOT ALLOWED          amondbs_standby                MAXIMUM PERFORMANCE  PRIMARY          YES MOUNTED                        0

SQL> alter database open;

SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 NOT ALLOWED          amondbs_standby                MAXIMUM PERFORMANCE  PRIMARY          YES READ WRITE                691176

（4）open rac节点2
SQL> startup

SQL> select dbid,inst_id,switchover_status,db_unique_name,protection_mode,database_role,force_logging,open_mode,current_scn from gv$database;

      DBID    INST_ID SWITCHOVER_STATUS    DB_UNIQUE_NAME                 PROTECTION_MODE      DATABASE_ROLE    FOR OPEN_MODE            CURRENT_SCN
---------- ---------- -------------------- ------------------------------ -------------------- ---------------- --- -------------------- -----------
 593350850          1 SESSIONS ACTIVE      amondbs_primary                MAXIMUM PERFORMANCE  PHYSICAL STANDBY YES READ ONLY                 690867
 593350850          2 SESSIONS ACTIVE      amondbs_primary                MAXIMUM PERFORMANCE  PHYSICAL STANDBY YES READ ONLY                 690867

（5）打开主库日志传输服务
 alter system set log_archive_dest_state_2=enable scope=both;


总结：
（1）从库的online redo和standby redo，根据主库的对应的thread1或n online redo切换而切换
（2）主库是rac，从库是单实例，物理standby概念-物理结构相同决定了从库数据文件和redo 文件的个数及其对应的thread个数与主库相同
（3）备库是rac时，节点1和节点2对redo传输和应用有序、协调地工作