这两天研究了下oracle 11g dataguard 物理standby 功能,总体来说这个功能满足公司需求,好了,不多说了,以下是详细的配置过程。
数据库的安装可以参考之前写的六步搞定Linux Oracle 11gR2 配置安装
注意:分别在主库和备库都安装上oracle软件,不装数据库。
主库:
IP:192.168.77.5 主机名:nod1 ORACLE_SID=test ORACLE_BASE=/oracle/app/oracle ORACLE_HOME=/oracle/app/oracle/product/11.2.0/dbhome_1
备库
IP:192.168.77.10 主机名:nod2 ORACLE_SID=test ORACLE_BASE=/oracle/app/oracle ORACLE_HOME=/oracle/app/oracle/product/11.2.0/dbhome_1
安装配置步骤:
1、主库运行dbca创建数据库(安装过程中需要注意:先关掉flash recovery),数据库名为test
2、登陆到数据库上
[oracle@nod1 ~]$ sqlplus /nolog SQL*Plus: Release 10.2.0.1.0 - Production on 星期四 3月 11 20:31:09 2010 Copyright (c) 1982, 2005, Oracle. All rights reserved. SQL> conn / as sysdba; Connected. SQL>
3、将主库设置为 FORCE LOGGING 模式
SQL> ALTER DATABASE FORCE LOGGING; Database altered.
4、创建一个密码文件,如果数据库是用dbca创建的则会在$ORACLE_HOME/dbs/下自动创建一个叫orapwdSID的一个密码文件。(否则可以用orapwd命令创建一个)。
5、配置Standby Redo Log。创建组数至少要比主库的online redo log组数多一个。
SQL> select group#,bytes from v$log; GROUP# BYTES ---------- ---------- 1 52428800 2 52428800 3 52428800 这里创建4组 SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 4 ('/oracle/app/oracle/oradata/test/stdbyredo01.log') SIZE 52428800; Database altered. SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 5 ('/oracle/app/oracle/oradata/test/stdbyredo02.log') SIZE 52428800; Database altered. SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 6 ('/oracle/app/oracle/oradata/test/stdbyredo03.log') SIZE 52428800; Database altered. SQL> ALTER DATABASE ADD STANDBY LOGFILE GROUP 7 ('/oracle/app/oracle/oradata/test/stdbyredo04.log') SIZE 52428800; 验证一下创建是否成功 SQL> SELECT GROUP#,THREAD#,SEQUENCE#,ARCHIVED,STATUS FROM V$STANDBY_LOG; GROUP# THREAD# SEQUENCE# ARC STATUS ---------- ---------- ---------- --- ---------- 4 0 0 YES UNASSIGNED 5 0 0 YES UNASSIGNED 6 0 0 YES UNASSIGNED 7 0 0 YES UNASSIGNED Database altered. 如果想删除某个组可以用下面的命令 SQL> alter database drop standby logfile group 4; 如果给某个组添加成员可以用下面命令 SQL> ALTER DATABASE ADD STANDBY LOGFILE MEMBER '/oracle/app/oracle/oradata/test/stdbyredo04_1.log' TO GROUP 7; 删除组中的成员用以下命令 SQL> ALTER DATABASE DROP STANDBY LOGFILE MEMBER '/oracle/app/oracle/oradata/test/stdbyredo04_1.log'; Database altered. SQL> host rm -f /oracle/app/oracle/oradata/test/stdbyredo04_1.log
6、检查数据库是否处于归档状态
SQL> archive log list; Database log mode Archive Mode Automatic archival Enabled Archive destination /oracle/app/oraclearch_log Oldest online log sequence 1 Next log sequence to archive 2 Current log sequence 2 如果不是处于归档状态则运行下面命令 SQL> SHUTDOWN IMMEDIATE; SQL> STARTUP MOUNT; SQL> ALTER DATABASE ARCHIVELOG; SQL> ALTER DATABASE OPEN;
7、创建备库的控制文件(创建后数据库不能做结构性的改变)
SQL> ALTER DATABASE CREATE STANDBY CONTROLFILE AS '/oracle/app/oracle/oradata/test/stdby.ctl';
8、创建pfile以便于修改配置参数
SQL> CREATE PFILE='/oracle/app/oracle/oradata/test/pfile' FROM SPFILE;
9、关掉数据库开始编辑参数
SQL> shutdown immediate;
10、编辑porfile
$ vi /oracle/app/oracle/oradata/test/pfile 添加下面内容 DB_UNIQUE_NAME=nod1 LOG_ARCHIVE_CONFIG='DG_CONFIG=(nod1,nod2)' LOG_ARCHIVE_DEST_2= 'SERVICE=nod2 LGWR ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=nod2' LOG_ARCHIVE_DEST_STATE_1=ENABLE LOG_ARCHIVE_DEST_STATE_2=ENABLE LOG_ARCHIVE_MAX_PROCESSES=30 FAL_SERVER=nod2 FAL_CLIENT=nod1 DB_FILE_NAME_CONVERT='nod2','nod1' STANDBY_FILE_MANAGEMENT=AUTO
11、进入sqlplus创建pfile(NOMOUNT状态)
SQL> CREATE SPFILE from PFILE='/oracle/app/oracle/oradata/test/pfile';
12、创建tnsnames.ora和listener.ora,可以手动创建也可以用netca创建,无论哪种方法都必须保证能互相畅通。
$ cd /oracle/app/oracle/product/11.2.0/dbhome_1/network/admin/ $ vi tnsnames.ora 输入下面内容 nod2 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.77.10)(PORT = 1521)) ) (CONNECT_DATA = (SERVICE_NAME = test) ) ) nod1 = (DESCRIPTION = (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.77.5)(PORT = 1521)) (CONNECT _DATA = (SERVER = DEDICATED) (SERVICE_NAME = test) ) ) $ vi listener.ora SID_LIST_LISTENER = (SID_LIST = (SID_DESC = (SID_NAME = test) (ORACLE_HOME = /oracle/app/oracle/product/11.2.0/dbhome_1) (PROGRAM = extproc) ) (SID_DESC = (GLOBAL_DBNAME = test) (SID_NAME = test) (ORACLE_HOME = /oracle/app/oracle/product/11.2.0/dbhome_1) ) ) LISTENER = (DESCRIPTION_LIST = (DESCRIPTION = (ADDRESS = (PROTOCOL = TCP)(HOST = nod2)(PORT = 1521)) (ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC0)) ) )
13、复制备份文件到备库
$ cd /oracle/app/oracle $ scp -r arch_log/ admin/ oradata/ 192.168.77.10:/oracle/app/oracle $ cd /oracle/app/oracle/product/11.2.0/dbhome_1/network/admin/ $ scp listener.ora tnsnames.ora 192.168.77.10:/oracle/app/oracle/product/11.2.0/dbhome_1/network/admin/ $ cd /oracle/app/oracle/product/11.2.0/dbhome_1/dbs/ $ scp orapwtest 192.168.77.10:/oracle/app/oracle/product/11.2.0/dbhome_1/dbs/
14、打开备库做以下操作
$ cd /oracle/app/oracle/oradata/test/ $ rm -rf control0* $ cp stdby.ctl control01.ctl $ cp stdby.ctl control02.ctl $ cp stdby.ctl control03.ctl $ vi pfile 对相应文件做修改 DB_UNIQUE_NAME=nod2 LOG_ARCHIVE_CONFIG='DG_CONFIG=(nod1,nod2)' LOG_ARCHIVE_DEST_2= 'SERVICE=nod1 LGWR ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=nod1' LOG_ARCHIVE_DEST_STATE_1=ENABLE LOG_ARCHIVE_DEST_STATE_2=ENABLE LOG_ARCHIVE_MAX_PROCESSES=30 FAL_SERVER=nod1 FAL_CLIENT=nod2 DB_FILE_NAME_CONVERT='nod2','nod1' STANDBY_FILE_MANAGEMENT=AUTO $ cd /oracle/app/oracle/product/11.2.0/dbhome_1/network/admin/ $ vi listener.ora 将nod1修改为nod2
15、在备库中进入sqlplus
$ sqlplus /nolog SQL> conn / as sysdba; SQL> CREATE SPFILE from PFILE='/oracle/app/oracle/oradata/test/pfile'; SQL>shutdown immediate;
16、测试oracle net是否畅通
首先启动监听服务(分别在主机和备机上进行) $ lsnrctl stop $ lsnrctl start 分别在主机和备机上运行下面命令来测试 $ tnsping nod1 $ tnsping nod2
17、启动主库(正常启动)
$ sqlplus /nolog SQL> conn / as sysdba; SQL> startup
18、启动备库到mount状态,并打开redo Apply
$ sqlplus /nolog SQL> conn / as sysdba; SQL> startup mount SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
19、在主库上做一次日志切换
SQL> ALTER SYSTEM SWITCH LOGFILE;
20、分别在主库和备库上做下列查询,检查归档日志是否由主库传送到备库
以下语句可以查看日志应用情况 SQL> SELECT SEQUENCE#,APPLIED FROM V$ARCHIVED_LOG ORDER BY SEQUENCE#; SEQUENCE# APP ---------- --- 2 YES 3 YES 如果出现上面情况则说明配置成功。
进行数据测试
1、在主 库建表然后在备库查询进行验证
SQL> create table nod2(id integer,name char(10)) SQL> insert into nod2 values(1,'nod1'); SQL> commit; SQL> select * from nod2; ID NAME ---------- ---------- 1 nod1
2、在主库上做个日志归档
SQL> ALTER SYSTEM ARCHIVE LOG CURRENT;
3、在备库上取消掉redo apply
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL; SQL> ALTER DATABASE OPEN; SQL> select * from nod2; ID NAME ---------- ---------- 1 nod1 测试成功
4、再次切换为备库的redo apply状态
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
注意:
此步骤在切换成功后,要进行主机HOST的相应配置,要不然在日志切换时会报错,
5、在主库上做如下sql语句
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY WITH SESSION SHUTDOWN; SQL> shutdown immediate; SQL> startup nomount; SQL> ALTER DATABASE MOUNT STANDBY DATABASE; SQL> RECOVER MANAGED STANDBY DATABASE DISCONNECT;
6、在备库上做switchover
SQL> ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY; SQL> shutdown immediate; SQL> startup
7、分别在新的主库和备库上查看状态,验证主库备库是否切换正常。
SQL> select name,database_role from v$database;
再次创建表在切换后的主库备库上验证
8、在新的主库上创建表
SQL> create table nod1(id integer,name char(10));
SQL> insert into nod1 values(111,'haha');
SQL> commit;
SQL> select * from nod1;
SQL> ALTER SYSTEM ARCHIVE LOG CURRENT;
9、在新的备库上做下面操作以取消redo apply
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL; SQL> ALTER DATABASE OPEN; SQL> select * from nod1;
10、恢复新备库为redo apply状态
以上是dataguard的配置过程,如何让数据实时同步,我们将在下篇进行讨论
维护 1、开机顺序,先备库的listener再启动备库。再启动主库的listener,再启动主库。 2、关机顺序,先关闭主库,再关闭备库 附上实时查看日志同步日志,可以查看问题所在 tail -f /oracle/app/oracle/diag/rdbms/nod1/test/trace/alert_test.log stanby上检查应用率和活动率(Ps) SQL>select to_char(start_time,'dd-mon-rr hh24:mi:ss') start_time,item,sofar from V$recovery_progress where item in ('Active Apply Rate', 'Average Apply Rate','Redo Applied'); Redo Applied值以MB衡量,而Active Apply Rate和Average Apply Rate以KB/s计算。