使用dataguard将windows11.2.0.1升级为linux11.2.0.4

1.环境介绍

源端环境:

oracle windows x86_64 11.2.0.1

目标端环境:

oracle linux x86_64 11.2.0.4

2.创建备库

原本想使用oracle duplicate进行复制创建物理dataguard,但是发现有报错ORA-17627和ORA-12577:

 

Errors in file d:\app\administrator\diag\rdbms\oracle\oracle\trace\oracle_ora_6012.trc:

ORA-17627: ORA-12577: Message 12577 not found;  product=RDBMS; facility=ORA

Tue Apr 23 18:37:02 2019

通过搜索MOS,发现文章:

RMAN Duplicate Database from Active Primary fails with Ora-17627: Ora-12577: Message 12577 Not Found; Product=Rdbms; Facility=Ora (文档 ID 1344615.1)

这里提到了可能是不一致的word size导致的:

CAUSE

Different 'Word Size' between Primary/Standby

eg:
  Primary  OS Details Redhat Linux 5 64-bits
  Standby OS Details Redhat Linux 4 32-bits

解决办法是使用rman的备份恢复来创建备库:

SOLUTION

Use RMAN duplicate database based on backup.

所以这里使用rman备份恢复的方式进行恢复.

注意这里发现一个小细节,我们在dg端修改参数:db_file_name_convert和log_file_name_convert

的时候,例如:

D:\APP\ADMINISTRATOR\ORADATA\ORACLE, /u01/oradata/oracle

这里我们如果不在最后的末尾将目录分隔符也替换,那么在linux上存放的目录会有问题,不是你想要的目录,如下:

Full restore complete of datafile 5 to datafile copy /u01/oradata/oracleKYEECIS.DBF. Elapsed time: 0:00:00

所以这里正确的写法为:

D:\APP\ADMINISTRATOR\ORADATA\ORACLE\, /u01/oradata/oracle/

后面带上目录分割符号,并且注意大小写,windows端必须是大写.

3.开启实时应用

standby:

alter database recover managed standby database using current logfile disconnect from session; 检查应用的延时:

select name,value,TIME_COMPUTED,DATUM_TIME from v$dataguard_stats;

4.主备切换

如果是switchover:

primary:

alter database commit to switchover to physical standby with session shutdown;

standby:

alter database commit to switchover to primary with session shutdown;

 

如果是failover:

standby:

alter database recover managed standby database finish force;

alter database commit to switchover to primary;

5.数据库升级

standby:

alter database open upgrade;

升级数据字典:

@$?/rdbms/admin/catupgrd.sql

这一步比较耗时,我这里执行了15分钟.

 

 

6.编译无效对象

编译无效对象:

SQL>startup

@?/rdbms/admin/utlrp

这一步根据数据库对象的多少耗时不定.

 到此升级完成.