轻松搞定dataguard

dataguard

1. 安装oracle，创建primary数据库 2. 设置primary库force logging. PRIMARY>alter database force logging; Database altered. 3. 创建密码文件 [oracle@PRIMARY]orapwd file=orapwprimary password=xxxxxxx; 4. 修改有关data guard的参数 PRIMARY> show parameter log_archive_config PRIMARY> alter system set log_archive_config='DG_CONFIG=(zxdb,zxdbdg1)'; System altered. PRIMARY>alter system set log_archive_dest_1='LOCATION=/u01/app/oracle/oradata/zxdb/archivelog VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=zxdb'; System altered. 5. 添加standby log group PRIMARY>alter database add standby logfile group 4('/u01/app/oracle/oradata/zxdb/standbylog/group_4_1.log','/u01/app/oracle/oradata/zxdb/standbylog/group_4_2.log') size 30m; Database altered. PRIMARY>alter database add standby logfile group 5('/u01/app/oracle/oradata/zxdb/standbylog/group_5_1.log','/u01/app/oracle/oradata/zxdb/standbylog/group_5_2.log') size 30m; Database altered. 创建成功后，在v$standby_log可以看到新加入的两个standby logfile的group， 在v$log里查询可以看到新添加进去的4个standby logfile。 6. 修改参数 fal_server zxdbdg fal_client zxdb standby_file_management auto log_archive_dest_state_2 enable log_archive_max_processes 7. 创建standby控制文件 PRIMARY> alter database create standby controlfile as '/u01/app/oracle/standby.ctl'; 8. 导出spfile为pfile PRIMARY> create pfile='/u01/app/oracle/init.ora' from spfile; 9. 复制数据库，可以物理拷贝，也可以用RMAN。 10. scp拷贝pfile和standby controlfile到备考机器 scp /u01/app/oracle/standby.ctl oracle:dg /u01/app/oracle/oradata/zxdb/controlfile/control1.ctl scp /u01/app/oracle/init.ora oracle:dg /u01/app/oracle/product/10.2.0/db_1/dbs/initzxdb.ora 11. 修改pfile db_unique_name zxdbdg control_files /u01/app/oracle/oradata/zxdb/controlfile/standby.ctl log_archive_dest_1 LOCATION=/u01/app/oracle/oradata/zxdb/archivelog VALID_FOR=(ALL_LOGFILES,ALL_ROLES) DB_UNIQUE_NAME=zxdbdg log_archive_dest_2 SERVICE=zxdb LGWR ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=zxdb fal_server zxdb fal_client zxdbdg 12. 在备库上pfile为spfile STANDBY> create spfile='/u01/app/oracle/product/10.2.0/db_1/dbs/spfilezxdb.ora' from pfile='/u01/app/oracle/product/10.2.0/db_1/dbs/initzxdb.ora' 13. 在主机上建立tns service zxdb和zxdbdg这里是上面fal指定的 zxdb = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.10.129)(PORT = 1521)) ) (CONNECT_DATA = (SERVICE_NAME = zxdb) ) ) zxdbdg = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.10.130)(PORT = 1521)) ) (CONNECT_DATA = (SERVICE_NAME = zxdb) ) ) 14 在备课上配置同上的tnsname zxdb = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.10.129)(PORT = 1521)) ) (CONNECT_DATA = (SERVICE_NAME = zxdb) ) ) zxdbdg = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = 172.16.10.130)(PORT = 1521)) ) (CONNECT_DATA = (SERVICE_NAME = zxdb) ) ) 15. 启动备库 STANDBY> start mount 16. 应用redo STANDBY> recover managed standby database 这里注意，这个事handle的。所以关掉session的话，会停止住。 一般我们都是用disconnect来解决这个问题 STANDBY> recover managed standby database disconnect from session 测试dg 在主库上insert data并且commit ID NAME --------------------------------------- ---------- 1 test1 2 test2 3 test3 4 test4 查看主库v$archived_log PRIMARY> select RECID, STAMP, NAME, DEST_ID, SEQUENCE#, STANDBY_DEST, ARCHIVED, APPLIED from v$archived_log; 11712946684 /u01/app/oracle/oradata/zxdb/archivelog/1_11_712892825.dbf 1 11 NO YES NO 12712946692 zxdbdg1 2 11 YES YES YES 13712948030 /u01/app/oracle/oradata/zxdb/archivelog/1_12_712892825.dbf 1 12 NO YES NO 14712948031 zxdbdg1 2 12 YES YES YES 在备库上查看v$archived_log STANDBY> create table test.testobject as select * from dba_tables; STANDBY> alter database open read only; SQL> select RECID, STAMP, NAME, DEST_ID, SEQUENCE#, STANDBY_DEST, ARCHIVED, APPLIED from v$archived_log; RECID STAMP NAME DEST_IDSEQUENCE# STANDBY_DEST ARCHIVED APPLIED ---------- ---------- -------------------------------------------------------------------------------- ---------- ---------- ------------ -------- ------- 1712946690 /u01/app/oracle/oradata/zxdb/archivelog/1_11_712892825.dbf 1 11 NO YES YES 2712948028 /u01/app/oracle/oradata/zxdb/archivelog/1_12_712892825.dbf 2 12 NO YES YES STANDBY> select RECID, STAMP, NAME, DEST_ID, SEQUENCE#, STANDBY_DEST, ARCHIVED, APPLIED from v$archived_log; ID NAME --------------------------------------- ---------- 1 test1 2 test2 3 test3 4 test4 这里已经成功了apply到备库上去了。 实验2 PRIMARY> create table test.testobject as select * from dba_tables; PRIMARY> commit 观察alert日志，没有归档的提示信息 redo日志没有满，所以没有切换 PRIMARY> insert into test.testobject select * from test.testobject; 多执行了几次，在alert日志，出现归档的信息。 用上面的方法查询，和主库一致。 STANDBY>select process,client_process,sequence#,status from v$managed_standby; PROCESS CLIENT_PSEQUENCE# STATUS --------- -------- ---------- ------------ ARCH ARCH 0 CONNECTED ARCH ARCH 0 CONNECTED ARCH ARCH 17 CLOSING ARCH ARCH 0 CONNECTED RFS N/A 0 IDLE MRP0 N/A 18 WAIT_FOR_LOG RFS LGWR 19 IDLE RFS UNKNOWN 0 IDLE

打造中国的asktom网 ASK Oracle技术网

http://ask.oraforum.net/