oracle dataguard配置规范文档 - 图文

3.2.10 启动数据库到mount状态

3.2.11 在备库进行数据库的恢复

rman target sys/qwe123@testbak run

{

allocate channel c1 type disk; allocate channel c2 type disk; restore database; };

3.2.12 备库启动介质恢复进程mrp

3.3 功能性配置

配置定期删除备库归档日志的脚本

cd /home/oracle vi archived.sql

select name from v\\$archived_log where applied='YES' and deleted='NO';

vi rmandelarch.sh

export ORACLE_SID=testbak

export ORACLE_BASE=/u01/app/oracle

export ORACLE_HOME=$ORACLE_BASE/product/10.2.0/db_1 export NLS_LANG=AMERICAN_AMERICA.AL32UTF8 export

PATH=$ORACLE_HOME/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:$PATH cd /home/oracle/

cat /dev/null > archived.txt

sqlplus sys/qwe123@testbak as sysdba <

set lines 50

spool /home/oracle/archived.txt @archived.sql spool off

EOF

cat archived.txt |xargs rm -rf -

rman target sys/qwe123@testbak <

crosscheck copy;

delete noprompt expired copy; } EOF

exit

赋予rmandelarch.sh脚本执行的权限并加入到crontab 中 Oracle 用户下

chmod u+x rmandelarch.sh

4.检查项

备库:检查归档日志接收和日志应用的具体情况

select process,status,sequence#,blocks,block# from v$managed_standby;

PROCESS STATUS SEQUENCE# BLOCKS ------------------ ------------------------ ---------- ---------- ----------

MRP0 WAIT_FOR_LOG 568 0 RFS RECEIVING 0 0 RFS ATTACHED 567 17963 BLOCK# 0 0 17963

注：如果sequenc#与主数据库归档日志号一致则Dataguard工作正常 如果status状态为WAIT_FOR_GAP则dataguard与主数据库不同步， Dataguard工作不正常

备库：检查备库备库应用日志跟接受日志的差距

Select,dest_name,database_mode ,archived_thread#,archived_seq#,applied_thread#,applied_seq#,db_unique_name from v$archive_dest_status;

主库:检查主库正在正在传送那个日志

select process,status,thread#,sequence# ,group# from v$managed_standby;

5. 测试方法

无论oracle dataguard, stream replication还是advanced replication 只是复制ddl,dml的操作

主库:

备库:

6.日常维护命令

6.1启动备库至mount状态,并启动mrp进程