升级失败后，数据库降级方案(flashback database)

环境：Oracle 11.2.0.1 rac on redhat 5.8

Flashback database

准备工作

查看是否flashback database功能

sys@RACDB>

select log_mode,open_mode,flashback_on from

v$database;

LOG_MODE

OPEN_MODE            FLASHBACK_ON

------------ -------------------- ------------------

ARCHIVELOG   READ

WRITE           NO

--

未开启flashback

database 功能

查看是否设置了闪回目录、闪回目录空间大小

sys@RACDB> show parameter db_recovery

NAME

TYPE        VALUE

------------------------------------ -----------

------------------------------

db_recovery_file_dest                string

db_recovery_file_dest_size           big integer 0

--

显然我们没有设置，下面为设置后

alter system set db_recovery_file_dest='+DATA' scope=spfile

sid='*';

--

重启后生效

alter system set db_recovery_file_dest_size=4g;

sys@RACDB> show parameter db_recovery

NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

db_recovery_file_dest                string      +DATA

db_recovery_file_dest_size           big integer 4G

设置闪回保留期

sys@RACDB> show parameter db_flashback

NAME                                 TYPE        VALUE

------------------------------------ -----------

------------------------------

db_flashback_retention_target        integer     1440

--

缺省单位为分钟

Flashback database

开启工作

开启flashback database功能

sys@RACDB> alter database flashback on;

Database altered.

sys@RACDB> select status from gv$instance;

STATUS

------------

OPEN

OPEN

sys@RACDB> select log_mode,open_mode,flashback_on from

v$database;

LOG_MODE

OPEN_MODE            FLASHBACK_ON

------------ -------------------- ------------------

ARCHIVELOG   READ

WRITE           YES

–-开启了flashback

database功能

oracle 10.2.0.1

开启flashback database

需要在mount，且只有一个实例的情况下，否则会报错

ORA-38759: Database must be mounted by only one instance and

not open.

oracle 11.2.0.1

开启flashback database在open状态下，且多个实例共存，没有任何问题

查看允许闪回的最早时间点

select oldest_flashback_scn

old_flhbck_scn,oldest_flashback_time old_flhbck_tim,

retention_target

rete_trgt,flashback_size/1024/1024 flhbck_siz,

estimated_flashback_size/1024/1024

est_flhbck_size

from v$flashback_database_log;

OLD_FLHBCK_SCN OLD_FLHBCK_TIM       RETE_TRGT FLHBCK_SIZ EST_FLHBCK_SIZE

-------------- ------------------- ---------- ----------

---------------

1602195 2013-08-08

16:48:34       1440     15.625               0

Flashback database

使用范例

1

、创建一个restore point

升级失败后，降级时使用flashback database to restore point upgrade_dou;。

sys@RACDB> create restore point upgrade_dou guarantee

flashback database;

Restore point created.

sys@RACDB> select oldest_flashback_scn

old_flhbck_scn,oldest_flashback_time old_flhbck_tim,

2  retention_target

rete_trgt,flashback_size/1024/1024 flhbck_siz,

3  estimated_flashback_size/1024/1024

est_flhbck_size

4  from v$flashback_database_log;

OLD_FLHBCK_SCN OLD_FLHBCK_TIM       RETE_TRGT FLHBCK_SIZ EST_FLHBCK_SIZE

-------------- ------------------- ---------- ----------

---------------

1602195 2013-08-08

16:48:34       1440     15.625        143.8125

2

、flashback database使用restore

point

降级时

数据库必须是mount状态

sys@RACDB>

flashback database to restore point upgrade_dou;

flashback database to restore point upgrade_dou

*

ERROR at line 1:

ORA-38757: Database must be mounted and not open to FLASHBACK.

Flashback database

必须resetlogs open数据库

sys@RACDB> alter database open;

alter database open

*

ERROR at line 1:

ORA-01589: must use RESETLOGS or NORESETLOGS option for

database open

RAC

环境中alter database open resetlogs;时有且只有一个实例执行resetlogs

idle> select status from gv$instance;

STATUS

------------

MOUNTED

MOUNTED

idle> alter database open resetlogs;

alter database open resetlogs

*

ERROR at line 1:

ORA-38777: database must not be started in any other instance

关闭一个节点

idle> select status from gv$instance;

STATUS

------------

MOUNTED

idle> alter database open resetlogs;

Database altered.

检查RAC节点恢复情况

另一个节点也open

idle> select status from gv$instance;

STATUS

------------

OPEN

OPEN

idle>

select current_scn from gv$database;

CURRENT_SCN

-----------

1628773

1628774

附表

查看允许闪回的最早时间点

select

oldest_flashback_scn old_flhbck_scn,oldest_flashback_time old_flhbck_tim,

retention_target rete_trgt,flashback_size/1024/1024

flhbck_siz,

estimated_flashback_size/1024/1024

est_flhbck_size

from

v$flashback_database_log;

查看sga中分配的闪回空间大小

select

*

from

v$sgastat

where

name

like

'flashback%';

POOL         NAME                            BYTES

------------

-------------------------- ----------

shared pool  flashback generation buff     3981204

shared pool  flashback_marker_cache_si        9196

查看闪回区的使用情况

select

name,space_limit/1024/1024 sp_limt,space_used/1024/1024 sp_usd,space_reclaimable/1024/1024

sp_recl,number_of_files num_fils

from

v$recovery_file_dest;

NAME

----------------------------------------------------------------------------------------------------

SP_LIMT     SP_USD

SP_RECL   NUM_FILS

---------- ---------- ---------- ----------

+DATA

4096         36          0          3

使用flashback database闪回数据库

步骤(前提归档日志可用)

关闭数据库

启动数据库到mount状态(exclusive模式)

闪回至某个时间点，SCN或log sequence number

使用resetlogs打开数据库

1.

使用sqlplus实现闪回

可以接受一个时间标记或一个系统改变号实参

sqlplus

几种常用的闪回数据库方法

FLASHBACK [STANDBY]

DATABASE []

TO [BEFORE] SCN

--

基于SCN闪回

FLASHBACK [STANDBY]

DATABASE []

TO [BEFORE] TIMESTMP --

基于时间戳闪回

FLASHBACK [STANDBY]

DATABASE []

TO [BEFORE] RESTORE POINT --

基于时点闪回

如下面的示例：

SQL>

flashback

database

to

timestamp('2010-10-24

13:04:30','yyyy-mm-dd hh24:mi:ss');

SQL>

flashback

database

to

scn 918987;

SQL>

flashback

database

ro

restore

point b1_load;

2.

使用RMAN进行flashback

database

使用RMAN进行闪回数据库的几种常用办法

RMAN>

flashback

database

to

scn=918987;

RMAN>

flashback

database

to

sequence=85

thread=1;

参考文献: