Oracle 10gR2 Dataguard搭建(非duplicate方式)
我的实验环境：
源生产库(主库)：
IP地址：192.168.1.30
Oracle 10.2.0.5 单实例
新DG库（备库）：
IP地址：192.168.1.31
Oracle 10.2.0.5 单实例
- 1.源生产库开启归档
- 2.rman备份源生产库
- 3.修改源生产库参数
- 4.配置tnsnames.ora
- 5.同步密码文件
- 6.配置pfile文件
- 7.创建备库控制文件
- 8.还原备库
- 9.开启日志应用
- 10.switchover测试
- 11.创建还原点，激活备库测试
- 12.failover测试
1.源生产库开启归档
部署Dataguard环境，要求主库必须开启归档模式；如果没有开启，需要先申请停机开启归档。
shutdown immediate
startup mount
alter database archivelog;
alter database open;
archive log list;
注意：归档日志存放位置，并制定归档日志删除策略；
归档日志删除举例(删除7天前归档)：
vi delarch.sql
crosscheck archivelog all;
delete noprompt archivelog all completed before "sysdate-7";
vi delarch.sh
#!/bin/bash
#Oracle ENV (you may need to change it.)
export ORACLE_BASE=/u01/app/oracle
export ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_1
export ORACLE_SID=ora10
export PATH=$ORACLE_HOME/bin:$PATH
rman target / @/home/oracle/rman/delarch.sql log=/home/oracle/rman/delarch.log
如果之前主库开启了归档，且有合理的备份策略；则实际上很可能不需要此步骤单独清除归档。
2.rman备份源生产库
将主库rman全备，视具体情况选择是否压缩备份集，我这里空间有限选择压缩备份集。
备份脚本：
[oracle@oradb30 rman]$ cat backup.sql
run {
allocate channel d1 type disk;
allocate channel d2 type disk;
backup as compressed backupset database format '/orabak/rman/data_%d_%T_%s.bak' plus archivelog format '/orabak/rman/log_%d_%T_%s.bak';
release channel d1;
release channel d2;
}
[oracle@oradb30 rman]$ cat backup.sh
#!/bin/bash
export ORACLE_BASE=/u01/app/oracle
export ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_1
export ORACLE_SID=ora10
export PATH=$ORACLE_HOME/bin:$PATH
rman target / @backup.sql log backup.log
执行备份：
[oracle@oradb30 rman]$ nohup sh backup.sh &
[oracle@oradb30 rman]$ tail -200f backup.log
Recovery Manager: Release 10.2.0.5.0 - Production on Wed Jan 4 21:47:23 2017
Copyright (c) 1982, 2007, Oracle. All rights reserved.
connected to target database: ORA10 (DBID=914008358)
RMAN> run {
2> allocate channel d1 type disk;
3> allocate channel d2 type disk;
4> backup as compressed backupset database format '/orabak/rman/data_%d_%T_%s.bak' plus archivelog format '/orabak/rman/log_%d_%T_%s.bak';
5> release channel d1;
6> release channel d2;
7> }
8>
using target database control file instead of recovery catalog
allocated channel: d1
channel d1: sid=143 devtype=DISK
allocated channel: d2
channel d2: sid=142 devtype=DISK
Starting backup at 04-JAN-17
current log archived
channel d1: starting compressed archive log backupset
channel d1: specifying archive log(s) in backup set
input archive log thread=1 sequence=46 recid=45 stamp=932415090
input archive log thread=1 sequence=47 recid=46 stamp=932420845
channel d1: starting piece 1 at 04-JAN-17
channel d2: starting compressed archive log backupset
channel d2: specifying archive log(s) in backup set
input archive log thread=1 sequence=43 recid=42 stamp=932414314
input archive log thread=1 sequence=44 recid=43 stamp=932414362
input archive log thread=1 sequence=45 recid=44 stamp=932415036
channel d2: starting piece 1 at 04-JAN-17
channel d1: finished piece 1 at 04-JAN-17
piece handle=/orabak/rman/log_ORA10_20170104_102.bak tag=TAG20170104T214725 comment=NONE
channel d1: backup set complete, elapsed time: 00:00:02
channel d2: finished piece 1 at 04-JAN-17
piece handle=/orabak/rman/log_ORA10_20170104_103.bak tag=TAG20170104T214725 comment=NONE
channel d2: backup set complete, elapsed time: 00:00:02
Finished backup at 04-JAN-17
Starting backup at 04-JAN-17
channel d1: starting compressed full datafile backupset
channel d1: specifying datafile(s) in backupset
input datafile fno=00001 name=/oradata/ora10/ORA10/datafile/o1_mf_system_d5tw48bw_.dbf
input datafile fno=00002 name=/oradata/ora10/ORA10/datafile/o1_mf_undotbs1_d5tw48d8_.dbf
input datafile fno=00004 name=/oradata/ora10/ORA10/datafile/o1_mf_users_d5tw48dg_.dbf
channel d1: starting piece 1 at 04-JAN-17
channel d2: starting compressed full datafile backupset
channel d2: specifying datafile(s) in backupset
input datafile fno=00003 name=/oradata/ora10/ORA10/datafile/o1_mf_sysaux_d5tw48c3_.dbf
input datafile fno=00007 name=/oradata/ora10/ORA10/datafile/o1_mf_forhapoc_d5x5bm2b_.dbf
input datafile fno=00005 name=/oradata/ora10/ORA10/datafile/o1_mf_dbs_d_ji_d5x45hbj_.dbf
input datafile fno=00006 name=/oradata/ora10/ORA10/datafile/o1_mf_dbs_i_ji_d5x45jd9_.dbf
channel d2: starting piece 1 at 04-JAN-17
channel d2: finished piece 1 at 04-JAN-17
piece handle=/orabak/rman/data_ORA10_20170104_105.bak tag=TAG20170104T214728 comment=NONE
channel d2: backup set complete, elapsed time: 00:00:35
channel d1: finished piece 1 at 04-JAN-17
piece handle=/orabak/rman/data_ORA10_20170104_104.bak tag=TAG20170104T214728 comment=NONE
channel d1: backup set complete, elapsed time: 00:00:42
Finished backup at 04-JAN-17
Starting backup at 04-JAN-17
current log archived
channel d1: starting compressed archive log backupset
channel d1: specifying archive log(s) in backup set
input archive log thread=1 sequence=48 recid=47 stamp=932420890
channel d1: starting piece 1 at 04-JAN-17
channel d1: finished piece 1 at 04-JAN-17
piece handle=/orabak/rman/log_ORA10_20170104_106.bak tag=TAG20170104T214810 comment=NONE
channel d1: backup set complete, elapsed time: 00:00:02
Finished backup at 04-JAN-17
Starting Control File and SPFILE Autobackup at 04-JAN-17
piece handle=/orabak/rman/20170104/controlfilec-914008358-20170104-06 comment=NONE
Finished Control File and SPFILE Autobackup at 04-JAN-17
released channel: d1
released channel: d2
Recovery Manager complete.
[1]+ Done nohup sh backup.sh
[oracle@oradb30 rman]$
备份完成后,将备份集拷贝到备机。
3.修改源生产库参数
查看主库文件(数据文件、临时文件、重做日志文件)存放目录：
select name from v$datafile union all
select name from v$tempfile union all
select member from v$logfile;
修改源生产库参数：
--设置convert参数（我这里源端和目标端都设置了db_create_file_dest，没有设置这两个参数）
alter system set log_file_name_convert='','' scope=spfile;
alter system set db_file_name_convert='','' scope=spfile;
--设置数据库为force logging
alter database force logging;
--设置db_unique_name, log_archive_config(这里主库的db_unique_name，出于对现有的生产环境最小影响考虑没有改)
alter system set db_unique_name='ora10' scope=spfile;
alter system set log_archive_config='DG_CONFIG=(ora10,ora10dg)';
--归档日志目录
alter system set log_archive_dest_1='LOCATION=/orabak/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=ora10';
alter system set log_archive_dest_2='SERVICE=ora10dg ASYNC LGWR VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=ora10dg';
--归档日志文件命名规则
alter system set log_archive_format='arch_%r_%t_%s.arc' scope=spfile;
--DG的fal_server和fal_client
alter system set fal_server='ora10dg';
alter system set fal_client='ora10';
--设置standby_file_management为自动
alter system set standby_file_management=AUTO;
--设置备库日志文件组，数量一般为目标库日志文件组+1
alter database add standby logfile group 11 size 52428800;
alter database add standby logfile group 12 size 52428800;
alter database add standby logfile group 13 size 52428800;
alter database add standby logfile group 14 size 52428800;
4.配置tnsnames.ora
Dataguard环境，log_archive_config的配置，就是用到tnsnames.ora配置文件中的别名。
--配置tnsnames.ora
在主库所有节点上的tnsnames.ora添加相应的连接串，并传给备库。
---主库tnsnames.ora添加
vi $ORACLE_HOME/network/admin/tnsnames.ora
ORA10 =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.1.30)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = ora10)
)
)
ORA10DG =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.1.31)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = ora10dg)
)
)
将tnsnames.ora上传到备库，如果之前tnsnames有其他内容，可以选择添加内容，只要最终保证主备库可以互相访问即可。
5.同步密码文件
--密码文件从主库拷贝到备库
[oracle@oradb30 dbs]$ scp orapwora10 192.168.1.31:/u01/app/oracle/product/10.2.0/db_1/dbs/
oracle@192.168.1.31's password:
orapwora10 100% 1536 1.5KB/s 00:00
[oracle@oradb30 dbs]$
6.配置pfile文件
Dataguard环境下，需要将主库的参数文件传到备库，进行修改，尤其注意路径在备库都存在且有相应权限。
--创建pfile文件
从主库的spfile中导出pfile文件，上传到备库，并做适当的修改调整。
--在主库上创建pfile文件，并上传到备库
create pfile='/tmp/init.ora' from spfile;
SQL> create pfile='/tmp/init.ora' from spfile;
File created.
SQL> exit
Disconnected from Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
[oracle@oradb30 dbs]$ scp /tmp/init.ora 192.168.1.31:/tmp/
reverse mapping checking getaddrinfo for bogon failed - POSSIBLE BREAK-IN ATTEMPT!
oracle@192.168.1.31's password:
init.ora 100% 1430 1.4KB/s 00:00
[oracle@oradb30 dbs]$
--备库最终pfile参数文件
[oracle@oradb31 admin]$ vi /tmp/init.ora
*.audit_file_dest='/u01/app/oracle/admin/ora10/adump'
*.background_dump_dest='/u01/app/oracle/admin/ora10/bdump'
*.compatible='10.2.0.5.0'
*.core_dump_dest='/u01/app/oracle/admin/ora10/cdump'
*.db_block_size=8192
*.db_create_file_dest='/oradata'
*.db_domain=''
*.db_file_multiblock_read_count=16
*.db_name='ora10'
*.db_recovery_file_dest='/orabak/flash_recovery_area'
*.db_recovery_file_dest_size=2147483648
*.db_unique_name='ora10dg'
*.dispatchers='(PROTOCOL=TCP) (SERVICE=ora10XDB)'
*.fal_client='ora10dg'
*.fal_server='ora10'
*.job_queue_processes=10
*.log_archive_config='DG_CONFIG=(ora10dg,ora10)'
*.log_archive_dest_1='LOCATION=/orabak/arch VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=ora10dg'
*.log_archive_dest_2='SERVICE=ora10 ASYNC LGWR VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=ora10'
*.log_archive_format='arch_%r_%t_%s.arc'
*.open_cursors=300
*.pga_aggregate_target=96468992
*.processes=150
*.remote_login_passwordfile='EXCLUSIVE'
*.sga_target=290455552
*.standby_file_management='AUTO'
*.undo_management='AUTO'
*.undo_tablespace='UNDOTBS1'
*.user_dump_dest='/u01/app/oracle/admin/ora10/udump'
启动备库到nomount状态：
SQL> startup nomount pfile='/tmp/init.ora';
ORA-01261: Parameter db_recovery_file_dest destination string cannot be translated
ORA-01262: Stat failed on a file destination directory
Linux-x86_64 Error: 2: No such file or directory
SQL> exit
Disconnected
[oracle@oradb31 admin]$ mkdir -p /orabak/flash_recovery_area
[oracle@oradb31 admin]$ sqlplus / as sysdba
SQL*Plus: Release 10.2.0.5.0 - Production on Wed Jan 4 23:24:25 2017
Copyright (c) 1982, 2010, Oracle. All Rights Reserved.
Connected to an idle instance.
SQL> startup nomount pfile='/tmp/init.ora';
ORACLE instance started.
Total System Global Area 293601280 bytes
Fixed Size 2095768 bytes
Variable Size 146802024 bytes
Database Buffers 138412032 bytes
Redo Buffers 6291456 bytes
SQL> exit
7.创建备库控制文件
Dataguard环境下，需要在主库创建备库的控制文件并传输到备库。
--创建控制文件
主库执行，创建备库控制文件
alter database create standby controlfile as '/tmp/control01.ctlbak';
[oracle@oradb30 dbs]$ scp /tmp/control01.ctlbak 192.168.1.31:/tmp/
reverse mapping checking getaddrinfo for bogon failed - POSSIBLE BREAK-IN ATTEMPT!
oracle@192.168.1.31's password:
control01.ctlbak 100% 6928KB 6.8MB/s 00:00
8.还原备库
8.1 在备库上启动数据库到nomount状态
#创建SPFILE
SQL>create SPFILE from pfile='/tmp/init.ora';
#启动到nomount状态
startup nomount
8.2 启动数据库到mount状态
restore standby controlfile from '/tmp/control01.ctlbak';
alter database mount;
可能需要手工注册备份集；
crosscheck backupset;
catalog start with '/orabak/rman/';
8.3 还原备份
vi /home/oracle/scripts/restore.sh
rman target / <<EOF! > ora10_restore.log
run {
allocate channel d1 type disk;
allocate channel d2 type disk;
restore database;
release channel d1;
release channel d2;
}
exit;
EOF!
nohup sh restore.sh &
vi /home/oracle/scripts/recover.sh
rman target / <<EOF! > ora10_recover.log
run {
allocate channel d1 type disk;
allocate channel d2 type disk;
recover database;
release channel d1;
release channel d2;
}
exit;
EOF!
nohup sh recover.sh &
9.开启日志应用
备库开启日志应用
SQL> alter database recover managed standby database using current logfile disconnect from session;
10. switchover测试
switchover,主备角色互换。
注意，主机的crontab 定时任务确认也正确迁移。
--源生产库：
SELECT SWITCHOVER_STATUS FROM V$DATABASE;
alter database commit to switchover to physical standby with session shutdown;
--备库（switchover为新生产库）：
SELECT SWITCHOVER_STATUS FROM V$DATABASE;
ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY WITH SESSION SHUTDOWN;
ALTER DATABASE OPEN;
--返回源生产库执行（switchover为新备库）：
shutdown immediate
startup mount
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT;
如果发现备库没有实时恢复，重置链路的状态再次切换日志试试：
SQL> alter system set log_archive_dest_state_2=defer;
SQL> alter system set log_archive_dest_state_2=enable;
SQL> alter system switch logfile;
11. 创建还原点，激活备库测试
11.1 创建restore point
**1) 在备库上设置flashback区域及大小**
SQL> alter system set db_recovery_file_dest='/orabak/flash_recovery_area';
SQL> alter system set db_recovery_file_dest_size=1000g;
SQL> select current_scn||'' from v$database;
2) Standby库取消redo日志应用并创建一个数据库闪回点
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
--这里注意下，如果你的备库环境是RAC，需要关闭其他节点，然后才可以创建restore point：
SQL> drop restore point before_application_test;
SQL> CREATE RESTORE POINT before_application_test GUARANTEE FLASHBACK DATABASE;
Restore point created.
11.2 激活备库
**1) 停止主库的日志传到备库**
SQL> ALTER SYSTEM SET LOG_ARCHIVE_DEST_STATE_2=defer;
2) 激活备库
SQL> alter system set job_queue_processes = 0 scope=both sid='*';
SQL> ALTER DATABASE ACTIVATE STANDBY DATABASE;
SQL> shutdown immediate
SQL> startup
3) 检查dblinks
select * from dba_db_links;
如果有dblink需要删除或其他方式禁用；
4）通知新的应用IP，开始测试应用工作
注意：为了预防测试过程中有job或者dblink修改了其他数据库信息，一般要修改job_queue_processes参数，删除dblink。
--参数
alter system set job_queue_processes = 0 scope=both sid='*';
--dblinks
select * from dba_db_links;
11.3 闪回恢复备库
**1) 恢复备库**
SQL> shutdown immediate
SQL> startup mount;
SQL> FLASHBACK DATABASE TO RESTORE POINT before_application_test;
SQL> ALTER DATABASE CONVERT TO PHYSICAL STANDBY;
SQL> shutdown immediate
SQL> startup mount
SQL> alter database recover managed standby database using current logfile disconnect;
恢复参数job_queue_processes设置值：
alter system set job_queue_processes = 10 scope=both sid='*';
恢复参数db_recovery_file_dest_size设置值：
alter system set db_recovery_file_dest_size=2g;
2) 恢复日志传输
恢复主库的日志传到备库
ALTER SYSTEM SET LOG_ARCHIVE_DEST_STATE_2=enable;
12. failover测试
failover，源生产库不再可用，备库强制激活为主库；
#取消DG应用
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
#重启下数据库（建议）
shutdown immediate;
startup
#操作不可逆，确定实际情况需要failover
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE FINISH force;
SELECT OPEN_MODE, DATABASE_ROLE, SWITCHOVER_STATUS, FORCE_LOGGING, DATAGUARD_BROKER, GUARD_STATUS FROM V$DATABASE;
#尝试常规切换为主库
ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY WITH SESSION SHUTDOWN;
如果这一步的常规切换失败，提示需要介质恢复，那么：
1）恢复备库 recover standby database until cancel;
2）激活备库 alter database activate standby database;
#最后重新启动数据库
shutdown immediate;
startup
failover之后，即便原主库恢复也要重新搭建DG了。