Oracle数据库数据恢复、性能优化

找回密码
注册
搜索
热搜: 活动 交友 discuz
发新帖

133

积分

0

好友

17

主题
1#
发表于 2012-3-8 17:46:36 | 查看: 6668| 回复: 4
有看这篇文件:http://www.oracledatabase12g.com/archives/11g%E6%96%B0%E7%89%B9%E6%80%A7rolling-upgrade-with-physical-standby.html

现在Primary 和 Physical Standby

10G不能够执行:
SQL> ALTER DATABASE RECOVER TO LOGICAL STANDBY KEEP IDENTITY;

KEEP IDENTITY clause is available only if the database being upgraded is running Oracle Database release 11.1 or later.


难道只能执行
ALTER DATABASE RECOVER TO LOGICAL STANDBY db_name;把Physical Standby变为Logic Standby,升级standby ,switchover,升级原Primary。再switchover(或者不需要switchover),我怎么做成Physical stadnby
5#
发表于 2012-3-9 16:39:05

回复 4# 的帖子

谢谢,非常感谢~!
反正是升级10g ->11R2
我就把Physical Standby 转成Logic Standby,之后升级Logic Standby到11G,二次switchover最后备库重建成Physical Standby。

貌似搜索了一下,不同版本oracle,dataguard环境,只能startup upgrade;

回复 只看该作者 道具 举报

4#
发表于 2012-3-9 15:24:46
环境:
Primary  A   ==> 10R2
Physical STandby B  ==> 10R2

"我只能建一个11R2的 Physical Standby C" ==> 不存在这种可能性 , 你无法建出一个  primary为 10g  physical standby 为 11gR2的  DG,  所以我认为 physical standby rolling upgrade 只针对 11gR2 以后的 DG 有效。

回复 只看该作者 道具 举报

3#
发表于 2012-3-9 14:18:52

回复 2# 的帖子

嗯,明白。
10.2.0.1  --> 10.2.0.4只能ALTER DATABASE RECOVER TO LOGICAL STANDBY db_name;

1、
Primary  A
Physical STandby B

2、
ALTER DATABASE RECOVER TO LOGICAL STANDBY db_name
把Physical standbyB -> Logic Standby B

3、
升级Logic Standby B 到10.2.0.4并dbua
建立主库关系,查看应用到最近的时间点
SELECT APPLIED_SCN, NEWEST_SCN,sysdate FROM DBA_LOGSTDBY_PROGRESS;

4、
switchover

5、
升级原Primary
应用日志

6、switchover

现在:
Primary A
Logic Standby B

我希望B是Physical Standby,只能重建了。

现在换个思路:
如果我想10R2 -->11R2呢。

环境:
Primary  A   ==> 10R2
Physical STandby B  ==> 10R2

我只能建一个11R2的 Physical Standby C,取消原Physical Standby B,通过11G特性来滚动升级,升级后为Physical Standby。
我即使把B能过ALTER DATABASE RECOVER TO LOGICAL STANDBY db_name;
B的DB_NAME and DBID都发生了变化,即使我nid把db_name改变回来,DBID也不一致了吧。

回复 只看该作者 道具 举报

2#
发表于 2012-3-8 18:01:14
10g 不支持临时将 physical standby 转换为  LOGICAL STANDBY 的 " ALTER DATABASE RECOVER TO LOGICAL STANDBY KEEP IDENTITY;" 这是 11g的新特性

logical standby 是支持 rolling upgrade 的


Logical Standby (DataGuard)
Goal:
Minimal downtime
Concept:
Upgrade will be done on the standby first
Afterwards: Switchover - standby will be now prod system
Now upgrade on ex-prod - switchover back to start layout

Pros&Cons:
Downtime ~2min fpr both switchovers
No changes of the OS platform possible
Logminer has some datatype restrictions
Some effort to setup and prod must be at least 10.1.0.3
But beginning with 11g: Transient Logical Standby utilizes an existing Physical Standby database for the rolling upgrade process


步骤:


1. Start with logical standby setup.

2. Use DBUA to upgrade the Logical Standby

3. Turn on log transport to synchronize (this could take some time to catch up)
    (At this point the users have no idea of any changes)

4. Switchover primary to standby, upgrade the former primary. If systems are identical then we could stop right there. If the original standby is less powerful or otherwise shoudln‘t be the primary, we can switch back.

Keep in mind that this concept doesn‘t support all datatypes.

回复 只看该作者 道具 举报

您需要登录后才可以回帖 登录 | 注册

QQ|手机版|Archiver|Oracle数据库数据恢复、性能优化

GMT+8, 2024-11-15 10:15 , Processed in 0.051368 second(s), 22 queries .

Powered by Discuz! X2.5

© 2001-2012 Comsenz Inc.

回顶部
TEL/電話+86 13764045638
Email service@parnassusdata.com
QQ 47079569