1、linux平台
Data Guard 环境:
操作系统: redhat 5.5
Primary数据库:
IP地址:211.87.147.69
数据库SID:mynew1
DB_UNIQUE_NAME:mynew1_pd
Standby数据库:
IP地址:211.87.147.68
数据库SID:mynew1
DB_UNIQUE_NAME:mynew1_st
一. Primary 端的配置
1. 主库设置为force logging 模式
SQL> alter database force logging;
2. 主库设为归档模式
SQL> archive log list;
SQL> shutdown immediate
SQL> startup mount
SQL> alter database archivelog;
SQL> archive log list;
3. 添加redo log file
添加一个新的Standby Redologs组(注意组号不要与当前存在的Online Redologs组重复),并为该组指定一个成员:
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 4 ('/u01/app/oracle/oradata/mynew1/redo04.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 5 ('/u01/app/oracle/oradata/mynew1/redo05.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 6 ('/u01/app/oracle/oradata/mynew1/redo06.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 7 ('/u01/app/oracle/oradata/mynew1/redo07.log') size 50M;
4. 修改listener.ora 和tnsnames.ora 文件
Listener.ora 文件:
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = /u01/app/oracle/product/10.2.0/db_1)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DBNAME = mynew1)
(ORACLE_HOME = /u01/app/oracle/product/10.2.0/db_1)
(SID_NAME = mynew1)
)
)
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = localhost.localdomain)(PORT = 1521))
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC0))
)
)
注意:SID_LIST_LISTENER 配置的是静态注册,如果没有该参数,而且Data Guard 启动顺序又不正确,那么在主库可能会报 PING[ARC1]: Heartbeat failed to connect to standby 'orcl_st'. Error is 12514. 错误,导致归档无法完成。
Tnsnames.ora 文件:
ORCL_ST =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 211.87.147.68)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = mynew1)
)
)
ORCL_PD =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 211.87.147.69)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = mynew1)
)
)
5. 创建备库的密码文件和控制文件
SQL> alter database create standby controlfile as '/u01/app/oracle/control01.ctl';
-- 说明: 判断一个数据库是Primary还是Standby,就是通过控制文件来判断的。
[oracle@localhost dbs]$ orapwd file=/u01/app/oracle/product/10.2.0/db_1/dbs/orapwmynew1 password=liu123
如果已经存在,就不用创建了。 缺省情况下,win下口令文件的格式是pwdsid.ora,unix下的格式是orapwSID(大小写敏感)
6. 修改初始化参数文件
SQL> create pfile='/u01/app/oracle/initmynew1.ora' from spfile;
在initmynew1.ora 添加如下内容:
####主库参数######
*.DB_NAME ='mynew1'; ——一般原来参数里有,不需要添加
#--监听中配置的
*.DB_UNIQUE_NAME='mynew1_pd'
#--列出DG中所有DB_UNIQUE_NAME
*.LOG_ARCHIVE_CONFIG='DG_CONFIG=(mynew1_pd,mynew1_st)'
*.LOG_ARCHIVE_DEST_1='LOCATION=/u01/app/oracle/archive VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=mynew1_pd'
*.LOG_ARCHIVE_DEST_2='service=mynew1_st DB_UNIQUE_NAME=mynew1_st'
*.LOG_ARCHIVE_DEST_STATE_1=ENABLE
*.LOG_ARCHIVE_DEST_STATE_2=ENABLE
*.REMOTE_LOGIN_PASSWORDFILE=EXCLUSIVE
#####备库参数#######
*.FAL_SERVER=mynew1_st
*.FAL_CLIENT=mynew1_pd
*.standby_file_management='AUTO'
*.standby_archive_dest='/u01/app/oracle/archive'
#如果主备库目录不同,还需要添加:
#*.log_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
#*.db_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
-- 注意:mynew1_st , mynew1_pd 是在tnsnames文件中配置的
用'/u01/app/oracle/initmynew1.ora' 这个pfile 启动数据库,并生成spfile。
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup pfile='/u01/app/oracle/initmynew1.ora';
ORACLE instance started.
Total System Global Area 167772160 bytes
Fixed Size 1218316 bytes
Variable Size 79694068 bytes
Database Buffers 83886080 bytes
Redo Buffers 2973696 bytes
Database mounted.
Database opened.
SQL> create spfile from pfile='/u01/app/oracle/initmynew1.ora';
File created.
二. Standby 端配置
1. 创建备库存放数据文件和后台跟踪目录, 这个目录可以和主库相同, 如果不同,就需要在主库的初始化文件中进行转换。
如:
*.log_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
*.db_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
在standby创建如下目录:
$ORACLE_BASE/ORADATA/mynew1
$ORACLE_BASE/admin/mynew1
$ORACLE_BASE/admin/mynew1/adump
$ORACLE_BASE/admin/mynew1/bdump
$ORACLE_BASE/admin/mynew1/cdump
$ORACLE_BASE/admin/mynew1/dpdump
$ORACLE_BASE/admin/mynew1/pfile
$ORACLE_BASE/admin/mynew1/udump
2. 将主库的密码文件,控制文件,数据文件,参数文件,日志文件copy到备库。
说明一点,这个控制文件是我们自己创建的standby 控制文件。将copy过来的控制文件再复制三份就可以了。 主备的控制文件是不一样的。 这里除了采用直接copy 文件之外,还可以采用Rman 恢复来做。 直接copy 需要停数据库,如果采用RMAN 的话,就不需要停机了。
现在进入主库(primary database)执行如下操作:
[oracle@localhost orcl]$ pwd
/u01/app/oracle/oradata/orcl
[oracle@localhost orcl]$ ls
control01.ctl redo01.log redo04.log redo07.log temp01.dbf
control02.ctl redo02.log redo05.log sysaux01.dbf undotbs01.dbf
control03.ctl redo03.log redo06.log system01.dbf users01.dbf
[oracle@localhost mynew1]$ scp *.dbf 211.87.147.68://u01/app/oracle/oradata/mynew1
[oracle@localhost u01]$ scp *.ctl 211.87.147.68://u01/app/oracle/oradata/mynew1
[oracle@localhost mynew1]$ scp *.log 211.87.147.68://u01/app/oracle/oradata/mynew1
[oracle@localhost u01]$ scp initorcl.ora 211.87.147.68://u01/app/oracle/product/10.2.0/db_1/dbs
[oracle@localhost dbs]$ scp orapworcl 211.87.147.68://u01/app/oracle/product/10.2.0/db_1/dbs
可以直接通过scp命令将主库(primary database)的文件复制到备库(standby database) 前提是主库 备库的ip配置正确,对应的目录都存在。
3. 修改初始化参数文件
进入备库的dbs目录,把刚刚拷贝过来的initmynew1.ora用vi编辑器打开:
修改之后如下:
####主库参数######
*.DB_NAME ='mynew1';
#--监听中配置的
*.DB_UNIQUE_NAME='mynew1_st'
#--列出DG中所有DB_UNIQUE_NAME
*.LOG_ARCHIVE_CONFIG='DG_CONFIG=(mynew1_pd,mynew1_st)'
*.LOG_ARCHIVE_DEST_1='LOCATION=/u01/archive VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=mynew1_st'
*.LOG_ARCHIVE_DEST_2='service=mynew1_pd DB_UNIQUE_NAME=mynew1_pd'
*.LOG_ARCHIVE_DEST_STATE_1=ENABLE
*.LOG_ARCHIVE_DEST_STATE_2=ENABLE
*.REMOTE_LOGIN_PASSWORDFILE=EXCLUSIVE
#####备库参数#######
*.FAL_SERVER=mynew1_pd
*.FAL_CLIENT=mynew1_st
*.standby_file_management='AUTO'
*.standby_archive_dest='/u01/archive'
#如果主备库目录不同,还需要添加:
#*.log_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
#*.db_file_name_convert='/u02/oradata/orcl/','/u03/oradata/orcl/'
4. 修改listener.ora 和 tnsnames.ora 文件,如果不存在,就从主库上copy 过去,不要修改tnsnames.ora。记得吧listener.ora的host地址改成备库的IP地址。
5. 在备库添加redo log file
如果主库没有添加redo log file,可以先用copy 过来的初始化文件将数据库启动到mount 状态。在创建个spfile,最后添加redo log。
SQL> startup mount pfile='/u01/app/oracle/product/10.2.0/db_1/dbs/initmynew1.ora'
ORACLE instance started.
SQL> create spfile from pfile='/u01/app/oracle/product/10.2.0/db_1/dbs/initmynew1.ora';
添加一个新的Standby Redologs组(注意组号不要与当前存在的Online Redologs组重复),并为该组指定一个成员:
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 4 ('/u01/app/oracle/oradata/orcl/redo04.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 5 ('/u01/app/oracle/oradata/orcl/redo05.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 6 ('/u01/app/oracle/oradata/orcl/redo06.log') size 50M;
SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 7 ('/u01/app/oracle/oradata/orcl/redo07.log') size 50M;
提示,由于从Primary数据库复制文件时并没有复制Online Redologs,因此物理Standby数据库在第一次启动REDO应用时,会在Alert文件中报Online Redo Logfile文件不存在,没有关系,物理Standby会自动重建这批文件,同时你也不用担心会丢失数据,Online Redologs中的数据会以归档文件的形式从Primary端接收。
至此,Data Guard 的操作已经完成,下面来开始验证。
注意Data Guard 启动顺序:
启动顺序:先standby ,后primary;
关闭顺序:先primary 后standby;
在备库将实例启动到mount 状态:
SQL> startup nomount;
SQL>alter database mount standby database ;
SQL>ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
SQL>alter database recover managed standby database disconnect from session;
在备库启动监听:
$lsnrctl start
在主库启动实例:
SQL> startup;
在主库启动监听:
$lsnrctl start
在主库验证归档目录是否有效:
SQL> SELECT STATUS,DESTINATION, ERROR FROM V$ARCHIVE_DEST;
如果有错误,要排查原因。
SQL> alter system switch logfile;
SQL> select max(sequence#) from v$archived_log;
MAX(SEQUENCE#)
--------------
70
主备查询结果一致,Data Guard 搭建结束。
注意:如果在主库执行 alter database clear unarchived logfile或alter database open resetlogs , 则dataguard要重建。
在以上操作中,遇到了各种错误,包括tnsping不通,备库无法归档等错误详情处理方法参照博客:
三. 一些其他操作
1. 首先查看当前的保护模式 ---primary数据库操作
SQL> select protection_mode,protection_level from v$database;
PROTECTION_MODE PROTECTION_LEVEL
-------------------- --------------------
MAXIMUM PERFORMANCE MAXIMUM PERFORMANCE
2. 设置新的数据保护模式并重启数据库 --primary数据库操作
当保护模式更改顺序:
maximize protection ---> maximize availability ----> maximize performance
当在把dataguard的保护级别按这上面的顺序减低的时候, 不需要primary库在mount状态,否则primary 必须在mount 状态。
如:
SQL> alter database set standby database to maximize availability;
alter database set standby database to maximize availability
*
ERROR at line 1:
ORA-01126: database must be mounted in this instance and not open in any
instance
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup mount;
ORACLE instance started.
Total System Global Area 167772160 bytes
Fixed Size 1218316 bytes
Variable Size 79694068 bytes
Database Buffers 83886080 bytes
Redo Buffers 2973696 bytes
Database mounted.
SQL> alter database set standby database to maximize availability;
Database altered.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-03113: end-of-file on communication channel
报错了,这是因为最大可用性需要先修改日志传送方式为lgwr同步方式,否则,数据库是无法open.
Maximum protection/AVAILABILITY模式必须满足以下条件
Redo Archival Process: LGWR
Network Tranmission mode: SYNC
Disk Write Option: AFFIRM
Standby Redo Logs: Yes
standby database type: Physical Only
SQL> alter system set log_archive_dest_2='service=mynew1_st lgwr sync AFFIRM';
System altered.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup mount;
ORACLE instance started.
Total System Global Area 167772160 bytes
Fixed Size 1218316 bytes
Variable Size 79694068 bytes
Database Buffers 83886080 bytes
Redo Buffers 2973696 bytes
Database mounted.
SQL> alter database set standby database to maximize availability;
Database altered.
SQL> alter database open;
Database altered.
SQL> select protection_mode,protection_level from v$database;
PROTECTION_MODE PROTECTION_LEVEL
-------------------- --------------------
MAXIMUM AVAILABILITY MAXIMUM AVAILABILITY
alter database set standby database to maximize performance;
提示:maximize后可跟{PROTECTION | AVAILABILITY | PERFORMANCE},分别对应最大保护,最高可用性及最高性能。
在最大保护模式下,直接关闭备库是不行的,如果在备库上关闭数据库,会有如下提示:
SQL> shutdown immediate
ORA-01154: database busy. Open, close, mount, and dismount not allowed now
SQL>
在最大保护模式下,备库是不允许关闭的,此时首先关闭主库,然后备库就可以顺利关闭了。
注意: 主库的保护模式修改之后,备库的模式也会改变,和主库保持一致。
3. 查看日志归档情况
主库进行日志切换:
SQL>Alter system switch logfile;
select max(sequence#) from v$archived_log;
select max(sequence#) from v$log_history;
select group#,sequence#,archived,status from v$log;
select name,sequence#,applied from v$archived_log;
select sequence#,applied from v$archived_log;
若不同步,
1) 看log日志, archive是否有丢失
2)可以在备库坐如下操作:
alter database recover managed standby database cancel;
alter database recover managed standby database disconnect from session;
四. 主备库切换
4.1 Switchover
一般SWITCHOVER切换都是计划中的切换,特点是在切换后,不会丢失任何的数据,而且这个过程是可逆的,整个DATA GUARD环境不会被破坏,原来DATA GUARD环境中的所有物理和逻辑STANDBY都可以继续工作。
在进行DATA GUARD的物理STANDBY切换前需要注意:
1)确认主库和从库间网络连接通畅;
2)确认没有活动的会话连接在数据库中;
3)PRIMARY数据库处于打开的状态,STANDBY数据库处于MOUNT状态;
4)确保STANDBY数据库处于ARCHIVELOG模式;
5)如果设置了REDO应用的延迟,那么将这个设置去掉;
6)确保配置了主库和从库的初始化参数,使得切换完成后,DATA GUARD机制可以顺利的运行。
主库:
1. 查看switchover 状态
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
SWITCHOVER_STATUS
--------------------
TO STANDBY
附: A:switchover_status出现session active/not allowed
当出现session active的时候表示还有活动的session,则运行
Alter database commit to switchover to physical standby with session shutdown;
当出现not allowed时,在官方文档说转换会不成功,但是我测试的时候成功了。
B.ora- 01153: an incompatible media recovery is active
运行下面代码
Alter database recover managed standby database finish;
或者Alter database recover managed standby database finish force;
Alter database recover managed standby database disconnect from session;
2 切换成备库
SQL>Alter database commit to switchover to physical standby with session shutdown;
或者
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY;
Database altered.
3 启动到mount和应用日志状态
SQL> SHUTDOWN IMMEDIATE
SQL> startup nomount;
SQL> alter database mount standby database;
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
4. 查看数据库模式
SQL>select dest_name,status,database_mode,recovery_mode,protection_mode from v$archive_dest_status;
SQL>select status,database_mode from v$archive_dest_status;
备库:
1.查看switchover状态
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
TO PRIMARY
附:若不是用此语句切换:ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY with session shutdown
补充:若出现:ORA-16139: media recovery required
是因为没有执行:alter database recover managed standby database disconnect from session;
2. 切换成主库
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
Database altered.
SQL> shutdown immediate;
SQL> startup;
SQL> alter system switch logfile;
3. 查看数据库模式
SQL>select dest_name,status,database_mode,recovery_mode,protection_mode from v$archive_dest_status;
SQL>select status,database_mode from v$archive_dest_status;
验证同步:
SQL> select max(sequence#) from v$archived_log;
MAX(SEQUENCE#)
--------------
78
4.2. Failovers:
FAILOVER切换一般是PRIMARY数据库发生故障后的切换,这种情况是STANDBY数据库发挥其作用的情况。这种切换发生后,可能会造成数据的丢失。而且这个过程不是可逆的,DATA GUARD环境会被破坏。
由于PRIMARY数据库已经无法启动,所以FAILOVER切换所需的条件并不多,只要检查STANDBY是否运行在最大保护模式下,如果是的话,需要将其置为最大性能模式,否则切换到PRIMARY角色也无法启动。
1. 查看是否有日志GAP,没有应用的日志:
SQL> SELECT UNIQUE THREAD#, MAX(SEQUENCE#) OVER(PARTITION BY THREAD#) LAST FROM V$ARCHIVED_LOG;
SQL> SELECT THREAD#, LOW_SEQUENCE#, HIGH_SEQUENCE# FROM V$ARCHIVE_GAP;
如果有,则拷贝过来并且注册
SQL> ALTER DATABASE REGISTER PHYSICAL LOGFILE '路径';
重复查看直到没有应用的日志:
2. 然后停止应用归档:
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
Database altered.
3. 下面将STANDBY数据库切换为PRIMARY数据库:
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH;
或 SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH FORCE;
Database altered.
SQL> SELECT DATABASE_ROLE FROM V$DATABASE;
DATABASE_ROLE
----------------
PHYSICAL STANDBY
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
Database altered.
SQL> ALTER DATABASE OPEN; 或者 shutdown immediate+startup
Database altered.
检查数据库是否已经切换成功:
SQL> SELECT DATABASE_ROLE FROM V$DATABASE;
DATABASE_ROLE
----------------
PRIMARY
至此,FAILOVER切换完成。这个时候应该马上对新的PRIMARY数据库进行备份。
2、windows平台
server A: 10.85.10.44 SID=orcl
target server B: 10.85.10.43
oracle version:10.2.0.1
OS platform : windows XP
Primary设置:
1. 设置主数据库为force logging 模式
SQL>sqlplus "/as sysdba"
SQL>alter database force logging;
2. 设置主数据库为归档模式
SQL> archive log list
SQL> shutdown immediate
SQL> startup mount
SQL> alter database archivelog;
SQL> archive log list
3. 添加"备用联机日志文件"
先查看日志文件位置:
SQL>select * from v$logfile;
在添加:
SQL> alter database add standby logfile group 4 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo04.log') size 50m;
SQL> alter database add standby logfile group 5 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo05.log') size 50m;
SQL> alter database add standby logfile group 6 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo06.log') size 50m;
SQL> alter database add standby logfile group 7 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo07.log') size 50m;
4. 创建主库的初始化参数给备库用
SQL>Create pfile from spfile;
产生的文件名为initorcl.ora 存放目录默认放在$ORACLE_HOME/database下
5. 在主库创建备库的控制文件 和密码文件
SQL>Alter database create standby controlfile as 'D:/oracle/product/10.2.0/control01.ctl';
D:>orapwd file=D:/PWDorcl.ora password=admin entries=5;
缺省情况下,win下口令文件的格式是pwdsid.ora,unix下的格式是orapwSID(大小写敏感)
6. 在主库创建监听和配置tnsnams.ora
listener.ora配置如下:
# listener.ora Network Configuration File: D:/oracle/product/10.2.0/db_1/network/admin/listener.ora
# Generated by Oracle configuration tools.
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = D:/oracle/product/10.2.0/db_1)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DBNAME = orcl)
(ORACLE_HOME = D:/oracle/product/10.2.0/db_1)
(SID_NAME = orcl)
)
)
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1))
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.44)(PORT = 1521))
)
)
tnsnames.ora配置如下:
# tnsnames.ora Network Configuration File: D:/oracle/product/10.2.0/db_1/network/admin/tnsnames.ora
# Generated by Oracle configuration tools.
primary =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.44)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = orcl)
)
)
standby =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.43)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = orcl)
)
)
EXTPROC_CONNECTION_DATA =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1))
)
(CONNECT_DATA =
(SID = PLSExtProc)
(PRESENTATION = RO)
)
)
7. 在initorcl.ora添加以下内容:
*.log_archive_format='%T%S%r.ARC'
*.DB_UNIQUE_NAME='primary'
*.log_archive_config='DG_CONFIG=(primary,standby)'
*.log_archive_dest_1='location=D:/oracle/product/10.2.0/oradata/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=primary'
*.log_archive_dest_2='SERVICE=standby arch ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=standby'
*.STANDBY_FILE_MANAGEMENT=AUTO
*.LOG_ARCHIVE_DEST_STATE_1=ENABLE
*.LOG_ARCHIVE_DEST_STATE_2=ENABLE
*.FAL_SERVER='standby'
*.FAL_CLIENT='primary'
关闭数据库,在用initorcl.ora重启,并创建spfile文件
SQL>startup pfile='$ORACLE_HOME/database/initorcl.ora'
SQL>create spfile from pfile='$ORACLE_HOME/database/initorcl.ora'
8.A. 拷贝数据文件,参数文件,控制文件,密码文件到备库上
initorcl.ora参数文件,PWDorcl.ora密码文件考到$ORACLE_HOME/database下的,
controlbak.ctl 考到$ORACLE_base/oradata/orcl/下,并分别重命名为control01.ctl,control01.ctl,control01.ctl
数据文件考到$ORACLE_base/oradata/orcl/下
B. 用Rman拷贝,不用停机
$ rman target /
RMAN> backup full format 'D:/FULL_%d_%T_%s.bak' database include current controlfile for standby;
RMAN> sql 'alter system archive log current';
RMAN> Backup ArchiveLog all format='D:/arch_%d_%T_%s.bak';
备份完后将备份文件拷到standby上同样的目录,强调:同样的目录,在standby进行rman 恢复即可
9. 启动主数据库
SQL>startup
---------------------------------------------------------------------
Standy操作:
1. 用oradim工具创建备库orcl实例
oradim.exe -new -sid orcl -startmode m
oradim.exe -edit -sid orcl -startmode a
2. 创建备库存放数据文件和后台跟踪目录
$ORACLE_BASE/ORADATA/ORCL
$ORACLE_BASE/admin/orcl
$ORACLE_BASE/admin/orcl/adump
$ORACLE_BASE/admin/orcl/bdump
$ORACLE_BASE/admin/orcl/cdump
$ORACLE_BASE/admin/orcl/dpdump
$ORACLE_BASE/admin/orcl/pfile
$ORACLE_BASE/admin/orcl/udump
$ORACLE_BASE/admin/orcl/
3. 添加"备用联机日志文件"
SQL>startup mount
先查看日志文件位置:
SQL>select * from v$logfile;
在添加:
SQL> alter database add standby logfile group 4 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo04.log') size 50m;
SQL> alter database add standby logfile group 5 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo05.log') size 50m;
SQL> alter database add standby logfile group 6 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo06.log') size 50m;
SQL> alter database add standby logfile group 7 ('D:/ORACLE/PRODUCT/10.2.0/ORADATA/ORCL/redo07.log') size 50m;
4. 在备库创建监听和配置tnsnams.ora
listener.ora配置如下:
# listener.ora Network Configuration File: D:/oracle/product/10.2.0/db_1/network/admin/listener.ora
# Generated by Oracle configuration tools.
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = D:/oracle/product/10.2.0/db_1)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DBNAME = orcl)
(ORACLE_HOME = D:/oracle/product/10.2.0/db_1)
(SID_NAME = orcl)
)
)
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1))
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.44)(PORT = 1521))
)
)
tnsnames.ora配置如下:
# tnsnames.ora Network Configuration File: D:/oracle/product/10.2.0/db_1/network/admin/tnsnames.ora
# Generated by Oracle configuration tools.
primary =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.44)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = orcl)
)
)
standby =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.85.3.43)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = orcl)
)
)
EXTPROC_CONNECTION_DATA =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1))
)
(CONNECT_DATA =
(SID = PLSExtProc)
(PRESENTATION = RO)
)
)
5. 测试主备之间网络连通
primary:
C:>lsnrctl start
C:>tnsping standby
standby
C:>lsnrctl start
C:>tnsping primary
6. 配置备库初始化参数
编辑$ORACLE_HOME/database目录下的initorcl.ora添加以下内容
*.log_archive_format='%T%S%r.ARC'
*.DB_UNIQUE_NAME='standby'
*.log_archive_config='DG_CONFIG=(primary,standby)'
*.log_archive_dest_1='location=D:/oracle/product/10.2.0/oradata/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=standby'
*.log_archive_dest_2='SERVICE=primary arch ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=primary'
*.STANDBY_FILE_MANAGEMENT=AUTO
*.LOG_ARCHIVE_DEST_STATE_1=ENABLE
*.LOG_ARCHIVE_DEST_STATE_2=ENABLE
*.FAL_SERVER='primary'
*.FAL_CLIENT='standby'
7. 启动备用数据库
C:>set ORACLE_SID=orcl
SQL>sqlplus "/as sysdba"
SQL>startup nomount pfile='D:/oracle/product/10.2.0/db_1/database/initorcl.ora';
SQL>create spfile from pfile='D:/oracle/product/10.2.0/db_1/database/initorcl.ora';
若采用Rman备份的, 则在此standby 端进行 Rman还原数据库:
$rman target sys/admin@primary auxiliary /
RMAN> duplicate target database for standby dorecover nofilenamecheck;
SQL>alter database mount standby database ;
SQL>alter database recover managed standby database disconnect from session;
---------------------------------------------------------------------------
测试
注意Data Guard 启动顺序:
启动顺序:先standby ,后primary;
关闭顺序:先primary 后standby;
在备库将实例启动到mount 状态:
SQL> startup nomount;
SQL>alter database mount standby database ;
SQL>ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
SQL>alter database recover managed standby database disconnect from session;
在备库启动监听:
$lsnrctl start
在主库启动实例:
SQL> startup;
在主库启动监听:
$lsnrctl start
在主库验证归档目录是否有效:
SQL> SELECT STATUS,DESTINATION, ERROR FROM V$ARCHIVE_DEST;
如果有错误,要排查原因。
SQL> alter system switch logfile;
SQL> select max(sequence#) from v$archived_log;
MAX(SEQUENCE#)
--------------
70
主备查询结果一致,Data Guard 搭建结束。
1. 测试主库产生的归档日志是否能正常传送到归档日志
主库进行日志切换:
SQL>Alter system switch logfile;
然后分别查看主库和备库的D:/arch目录下是否产生了同样的归档日志
文件。
select max(sequence#) from v$archived_log;
select max(sequence#) from v$log_history;
select group#,sequence#,archived,status from v$log;
select name,sequence#,applied from v$archived_log;
select sequence#,applied from v$archived_log;
若不同步,
1. 看log日志, archive是否有丢失
2. 可以在备库坐如下操作:
alter database recover managed standby database cancel;
alter database recover managed standby database disconnect from session;
-----------------------------------------------------------------------
主备库切换
1. switchover
一般SWITCHOVER切换都是计划中的切换,特点是在切换后,不会丢失任何的数据,而且这个过程是可逆的,整个DATA GUARD环境不会被破坏,原来DATA GUARD环境中的所有物理和逻辑STANDBY都可以继续工作。 在进行DATA GUARD的物理STANDBY切换前需要注意:
确认主库和从库间网络连接通畅;
确认没有活动的会话连接在数据库中;
PRIMARY数据库处于打开的状态,STANDBY数据库处于MOUNT状态;
确保STANDBY数据库处于ARCHIVELOG模式;
如果设置了REDO应用的延迟,那么将这个设置去掉;
确保配置了主库和从库的初始化参数,使得切换完成后,DATA GUARD机制可以顺利的运行。
主库:
1. 查看switchover 状态
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
to standby
附: A:switchover_status出现session active/not allowed
当出现session active的时候表示还有活动的session,则运行
Alter database commit to switchover to physical standby with session shutdown;
当出现not allowed时,在官方文档说转换会不成功,但是我测试的时候成功了,如果大家在测试不成功的时候再和我说,让我看看在什么情况下会不成功。
B.ora- 01153: an incompatible media recovery is active
运行下面代码
Alter database recover managed standby database finish;
或者Alter database recover managed standby database finish force;
Alter database recover managed standby database disconnect from session;
2 切换成备库
SQL>Alter database commit to switchover to physical standby with session shutdown;
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY;
Database altered.
3 启动到mount和应用日志状态
SQL> SHUTDOWN IMMEDIATE
SQL> startup nomount;
SQL> alter database mount standby database;
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
4. 查看数据库模式
SQL>select dest_name,status,database_mode,recovery_mode,protection_mode from v$archive_dest_status;
SQL>select status,database_mode from v$archive_dest_status;
备库:
1.查看switchover状态
SQL> SELECT SWITCHOVER_STATUS FROM V$DATABASE;
TO PRIMARY
附:若不是用此语句切换:ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY with session shutdown
2. 切换成主库
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
Database altered.
SQL> shutdown immediate;
SQL> startup;
SQL> alter system switch logfile;
3. 查看数据库模式
SQL>select dest_name,status,database_mode,recovery_mode,protection_mode from v$archive_dest_status;
SQL>select status,database_mode from v$archive_dest_status;
注意地方:
如果做了switchover,主库参数设置成以下方式,会触发ora-16009错误
Alert system set log_archive_dest_2=’service=primary
DB_UNIQUE_NAME=orcl’ scope=spfile;
然后再alert_orcl.log 日志中会出现以下错误内容
Thu Nov 27 10:19:12 2008
Redo Shipping Client Connect
-- Connected User is Valid
RFS[2]: Assigned to RFS process 1292
RFS[2]: Database mount ID mismatch [0x4781d95f:0x47823be1]
RFS[2]: Client instance is standby database i
RFS[2]: Not using real app
Thu Nov 27 10:19Errors in file
d:/oracle/product/10.2.0/admin/orcl/udump/orc
ORA-16009: 远程归档日
从metalink上查到:
* fact: Oracle Server - Enterprise Edition 9
* symptom: Errors appears in alert.log on primary database
* symptom: RFS: client instance is standby database instead
* symptom: RFS: Not using real application clusters
* symptom: Errors appear in alert.log on standby database
* symptom:
database
standby database
primary database
* symptom: Standby redo log files are defined on the standby database
* cause: The standby redo log files are synchronously filled with redo
from the primary database. When a logswitch occur on the primary database,
those files are archived on the standby database before being applyed on
it. The archiving process on the standby database should only archive to
the local disks on tprimarfix:
Disable the remote archiving on the standby databasExample: alter system set log_archive_dest_2 = ''
是因为没有把standby 上的log_archive_dest_2 清空导致的。
另外也有可
bug 4676659
Standby may not be recognised (ORA-16009)
When the log transport is LGWR ASYNC and logical standby has
LOG_ARCHIVE_DEST setting VALID_FOR=(ONLINE_LOGFILE, PRIMARY_ROLE)
ORA-16009 is reportedregular interWorkaround:
There is no workaround to prevent ORA-16009 from appearing in alert logs.
2. Failovers:
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;
Database altered.
3. 下面将STANDBY数据库切换为PRIMARY数据库:
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH;
或 SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH FORCE;
Database altered.
SQL> SELECT DATABASE_ROLE FROM V$DATABASE;
DATABASE_ROLE
----------------
PHYSICAL STANDBY
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
Database altered.
SQL> ALTER DATABASE OPEN; 或者 shutdown immediate+startup
Database altered.
检查数据库是否已经切换成功:
SQL> SELECT DATABASE_ROLE FROM V$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。
数据库模式查看:
1. 首先查看当前的保护模式 ---primary数据库操作
SQL> select protection_mode,protection_level from v$database;
PROTECTION_MODE PROTECTION_LEVEL
-------------------- --------------------
MAXIMUM PERFORMANCE MAXIMUM PERFORMANCE
2. 设置新的数据保护模式并重启数据库 --primary数据库操作
当保护模式更改顺序:
maximize protection ---> maximize availability ----> maximize performance
当在把dataguard的保护级别按这上面的顺序减低的时候, 不需要primary库在mount状态,否则primary 必须在mount 状态。
如:
SQL> alter database set standby database to maximize availability;
alter database set standby database to maximize availability
*
ERROR at line 1:
ORA-01126: database must be mounted in this instance and not open in any
instance
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup mount;
ORACLE instance started.
Total System Global Area 167772160 bytes
Fixed Size 1218316 bytes
Variable Size 79694068 bytes
Database Buffers 83886080 bytes
Redo Buffers 2973696 bytes
Database mounted.
SQL> alter database set standby database to maximize availability;
Database altered.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-03113: end-of-file on communication channel
报错了,这是因为最大可用性需要先修改日志传送方式为lgwr同步方式,否则,数据库是无法open.
Maximum protection/AVAILABILITY模式必须满足以下条件
Redo Archival Process: LGWR
Network Tranmission mode: SYNC
Disk Write Option: AFFIRM
Standby Redo Logs: Yes
standby database type: Physical Only
SQL> alter system set log_archive_dest_2='service=orcl_st lgwr sync AFFIRM';
System altered.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup mount;
ORACLE instance started.
Total System Global Area 167772160 bytes
Fixed Size 1218316 bytes
Variable Size 79694068 bytes
Database Buffers 83886080 bytes
Redo Buffers 2973696 bytes
Database mounted.
SQL> alter database set standby database to maximize availability;
Database altered.
SQL> alter database open;
Database altered.
SQL> select protection_mode,protection_level from v$database;
PROTECTION_MODE PROTECTION_LEVEL
-------------------- --------------------
MAXIMUM AVAILABILITY MAXIMUM AVAILABILITY
alter database set standby database to maximize performance;
提示:maximize后可跟{PROTECTION | AVAILABILITY | PERFORMANCE},分别对应最大保护,最高可用性及最高性能。
在最大保护模式下,直接关闭备库是不行的,如果在备库上关闭数据库,会有如下提示:
SQL> shutdown immediate
ORA-01154: database busy. Open, close, mount, and dismount not allowed now
SQL>
在最大保护模式下,备库是不允许关闭的,此时首先关闭主库,然后备库就可以顺利关闭了。
注意: 主库的保护模式修改之后,备库的模式也会改变,和主库保持一致。