2013年4月27日 星期六
10:33
1、redo （重做) log 的功能： 用于数据恢复
2、redo log 特征：
【特征】：
1) 记录数据块的变化（DML、DDL)
2) 用于数据块的recover
3) 以组的方式管理redo file ，最少两组redo ，循环使用
4) 和数据文件存放到不同的磁盘上，需读写速度快的磁盘(比如采用RAID10)
5) online redolog 在线日志
日志组 和 日志文件的状态
日志组的status：
unused 新添加的日志组，还没有使用
inactive 日志组对应的脏块已经从data buffer写入到data file ，可以删除
active 日志组对应的脏块还没有从data buffer写入到data file，不能被删除，否则会引起数据丢失
current 当前日志组，日志组对应的脏块还没有从data buffer写入到data file，不能被删除
在删除日志组之前，进行日志手工切换，将被删除的日志切换到inactive状态，这样就不会引起数据丢失。
thread ：线程（通过后台进程lgwr 启动)，在单实例的环境下，thread# 永远是1
sequence ：日志序列号。在日志切换时会递增。
FIRST_CHANGE# ：在当前日志中记录的首个数据块的scn。（当事务完成的时候会在数据块上写入一个scn，代表数据块的变化)。
ARC：是否归档模式
log_checkpoints_to_alter 该值默认为false，将其设置为true，每生成一个检查点就会在alter告警日志记录信息
日志文件的status：
valid 有效状态
invalid 当数据库未使用该组文件时，就是无效状态
stale 当数据库怀疑日志文件没完成或者不正确时，日志文件为不新鲜状态，当该日志组变为active group 时，stale日志文件会变为valid状态
日志切换：
1) 归档模式：将历史日志进行保存 ,并产生checkpoint，通知redo log 把对应的 dirty block 从data buffer写入到datafile，并且更新控制文件
2)非归档： 历史日志被覆盖【archive log list 可以查看数据库是否启用归档模式】
与日志相关的的数据字典
V$LOG 查看日志信息(从控制文件读取)
V$LOGFILE 查看日志存放的位置
FAST_START_MTTR_TARGET 控制检查点出现的时间也就是 实例恢复所需要的最长的时间 不能设置的太短
ALTER SYSTEM SWITH LOGFILE; 产生增量检查点
ALTER SYSTEM CHECKPONIT; 手工生成检查点 即: 产生完全检查点，促使 所有的脏块都写入数据文件
3、redo 日志组
1) 最少两组，最好每组有两个成员，并存放到不同的磁盘上，大小相同，互相镜像
2) 日志在写满时发生切换，或
手工切换：alter system switch logfile ；
3) 在归档模式，日志进行归档，并把相关的信息写入controlfile
4、添加日志组
SQL> startup
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 134219516 bytes
Database Buffers 113246208 bytes
Redo Buffers 2973696 bytes
Database mounted.
Database opened.
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
2 1 7 10485760 1 NO CURRENT 381102 02-AUG-11
1 1 6 10485760 1 NO INACTIVE 357157 01-AUG-11
SQL> select group#, member from v$logfile;
GROUP# MEMBER
---------- --------------------------------------------------
2 /u01/app/oracle/oradata/cuug/redo02a.log
1 /u01/app/oracle/oradata/cuug/redo01a.log
添加 日志组
SQL> alter database add logfile group 4
'/u01/app/oracle/oradata/cuug/redo03a.log' size 10m; ——添加日志组
Database altered.
添加日志 成员
SQL> alter database add logfile member
'/disk1/cuug/oradata/redo01b.log' to group 1, 指定哪个日志组
'/disk1/cuug/oradata/redo02b.log' to group 2,
'/disk1/cuug/oradata/redo03b.log' to group 3 ； ——添加日志组成员，放在不同的磁盘，形成镜像
Database altered.
SQL> select group#, member from v$logfile ;
GROUP# MEMBER
---------- --------------------------------------------------
2 /u01/app/oracle/oradata/cuug/redo02a.log
1 /u01/app/oracle/oradata/cuug/redo01a.log
3 /u01/app/oracle/oradata/cuug/redo03a.log
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 6 10485760 1 NO INACTIVE 357157 01-AUG-11
3 1 0 10485760 1 YES UNUSED 0 ——刚创建的日志组，处于未使用状态，进行日志切换即可
2 1 7 10485760 1 NO CURRENT 381102 02-AUG-11
SQL> alter system switch logfile; 手工切换日志
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARCHIVED STATUS FIRST_CHANGE# FIRST_TIME
---------- ---------- ---------- ---------- ---------- --------- ---------- ------------- -------------------
1 1 7 104857600 1 NO INACTIVE 215301 2012-03-07 18:02:39
4 1 10 104857600 1 NO CURRENT 238364 2012-03-08 17:14:10
3 1 9 104857600 1 NO ACTIVE 237357 2012-03-08 16:49:45
2 1 8 104857600 1 NO INACTIVE 216223 2012-03-07 18:49:17
5、添加日志组的成员
为什么多元化日志成员 ???
【如果当前日志组只有一个成员并且日志成员所在磁盘损坏, oracle 无法写入日志就会发生实例 crash (实例崩溃)
所以建议大家多元化日志成员】
SQL> select group#, member from v$logfile;
GROUP# MEMBER
---------- --------------------------------------------------
2 /u01/app/oracle/oradata/cuug/redo02a.log
1 /u01/app/oracle/oradata/cuug/redo01a.log
3 /u01/app/oracle/oradata/cuug/redo03a.log
SQL> select group#, member from v$logfile order by 1;
GROUP# MEMBER
---------- --------------------------------------------------
1 /u01/app/oracle/oradata/cuug/redo01a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
2 /u01/app/oracle/oradata/cuug/redo02a.log
3 /disk1/cuug/oradata/redo03b.log
3 /u01/app/oracle/oradata/cuug/redo03a.log
6、查看日志信息
SYS@ prod>select SEQUENCE# ,FIRST_CHANGE#,NEXT_CHANGE# ,to_char(FIRST_TIME ,'yyyy-mm-dd- hh24:mi:ss') from v$log_history;
SEQUENCE# FIRST_CHANGE# NEXT_CHANGE# TO_CHAR(FIRST_TIME,'
---------- ------------- ------------ --------------------
1 446075 474850 2013-01-09- 03:55:23
2 474850 496815 2013-01-09- 03:59:39
3 496815 497931 2013-01-09- 04:02:31
4 497931 497947 2013-01-09- 04:38:17
5 497947 497961 2013-01-09- 04:39:01
6 497961 497971 2013-01-09- 04:39:36
7 497971 497975 2013-01-09- 04:39:58
8 497975 497983 2013-01-09- 04:40:03
9 497983 497986 2013-01-09- 04:40:21
10 497986 498899 2013-01-09- 04:40:25
11 498899 499047 2013-01-09- 05:01:09
12 499047 499055 2013-01-09- 05:02:54
13 499055 531192 2013-01-09- 05:02:58
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 6 10485760 2 NO INACTIVE 357157 01-AUG-11
3 1 0 10485760 2 YES UNUSED 0
2 1 7 10485760 2 NO CURRENT 381102 02-AUG-11
SQL> alter system switch logfile;
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 9 10485760 2 NO CURRENT 384007 02-AUG-11
3 1 8 10485760 2 NO ACTIVE 384005 02-AUG-11
2 1 7 10485760 2 NO ACTIVE 381102 02-AUG-11
通过查询日志切换时间 可知道 系统哪个时间段 事务繁忙
SYS@ prod>desc v$log_history;
Name Null? Type
----------------------------------------------------------------- -------- --------------------------------------------
RECID NUMBER
STAMP NUMBER
THREAD# NUMBER
SEQUENCE# NUMBER
FIRST_CHANGE# NUMBER
FIRST_TIME DATE
NEXT_CHANGE# NUMBER
RESETLOGS_CHANGE# NUMBER
RESETLOGS_TIME DATE
SYS@ prod>select SEQUENCE# ,FIRST_CHANGE#,NEXT_CHANGE# ,to_char(FIRST_TIME ,'yyyy-mm-dd- hh24:mi:ss') time from v$log_history;
【Oracle建议 切换时间为 15分钟】
SEQUENCE# FIRST_CHANGE# NEXT_CHANGE# TIME
---------- ------------- ------------ --------------------
1 446075 474850 2013-01-09- 03:55:23
2 474850 496815 2013-01-09- 03:59:39
3 496815 497931 2013-01-09- 04:02:31
4 497931 497947 2013-01-09- 04:38:17
5 497947 497961 2013-01-09- 04:39:01
6 497961 497971 2013-01-09- 04:39:36
7 497971 497975 2013-01-09- 04:39:58
8 497975 497983 2013-01-09- 04:40:03
9 497983 497986 2013-01-09- 04:40:21
10 497986 498899 2013-01-09- 04:40:25
11 498899 499047 2013-01-09- 05:01:09
12 499047 499055 2013-01-09- 05:02:54
13 499055 531192 2013-01-09- 05:02:58
----------查看生成的日志量
SQL> insert into scott.emp1 select * from scott.emp1;
28 rows created.
SQL> insert into scott.emp1 select * from scott.emp1;
56 rows created.
——通过联合查询视图 v$statname ,v$mystat可以查看日志生成量
SYS @ prod >select a.name, b.value/1024/1024 "SIZE: M" from v$statname a,v$mystat b where b.statistic#=a.statistic# and a.name like '%redo size%';
NAME SIZE: M
---------------------------------------------------------------- ----------
redo size .019523621
7、redo 日志成员重命名或迁移 【只能在mount状态下修改因为mount状态打开了控制文件】一定要确认新的位置有该文件否则报错 总之先改物理位置再改逻辑位置 无法修改日志文件的大小因为日志文件一旦建立就已经被格式化为oracle标准格式 所以无法修改
SQL> select group#,member from v$logfile order by 1;
GROUP# MEMBER
---------- --------------------------------------------------
1 /u01/app/oracle/oradata/cuug/redo01a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
2 /u01/app/oracle/oradata/cuug/redo02a.log
3 /disk1/cuug/oradata/redo03b.log
3 /u01/app/oracle/oradata/cuug/redo03a.log
6 rows selected.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
[oracle@oracle ~]$ cp /u01/app/oracle/oradata/cuug/redo01a.log /disk2/cuug/oradata/
[oracle@oracle ~]$ cp /u01/app/oracle/oradata/cuug/redo02a.log /disk2/cuug/oradata
[oracle@oracle ~]$ cp /u01/app/oracle/oradata/cuug/redo03a.log /disk2/cuug/oradata
[oracle@oracle ~]$ !sql
sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.1.0 - Production on Tue Aug 2 05:08:35 2011
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup mount ;
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 125830908 bytes
Database Buffers 121634816 bytes
Redo Buffers 2973696 bytes
Database mounted.
SQL> select group# ,member from v$logfile order by 1;
GROUP# MEMBER
---------- --------------------------------------------------
1 /u01/app/oracle/oradata/cuug/redo01a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
2 /u01/app/oracle/oradata/cuug/redo02a.log
3 /disk1/cuug/oradata/redo03b.log
3 /u01/app/oracle/oradata/cuug/redo03a.log
SQL> alter database rename file '/u01/app/oracle/oradata/cuug/redo01a.log' to '/disk2/cuug/oradata/redo01a.log';
Database altered.
SQL> alter database rename file '/u01/app/oracle/oradata/cuug/redo02a.log' to '/disk2/cuug/oradata/redo02a.log';
Database altered.
SQL> alter database rename file '/u01/app/oracle/oradata/cuug/redo03a.log' to '/disk2/cuug/oradata/redo03a.log';
Database altered.
SQL> select group# ,member from v$logfile order by 1;
GROUP# MEMBER
---------- --------------------------------------------------
1 /disk2/cuug/oradata/redo01a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
2 /disk2/cuug/oradata/redo02a.log
3 /disk1/cuug/oradata/redo03b.log
3 /disk2/cuug/oradata/redo03a.log
6 rows selected.
SQL> alter database open;
Database altered.
SQL> select group# ,member from v$logfile order by 1;
GROUP# MEMBER
---------- --------------------------------------------------
1 /disk2/cuug/oradata/redo01a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
2 /disk2/cuug/oradata/redo02a.log
3 /disk1/cuug/oradata/redo03b.log
3 /disk2/cuug/oradata/redo03a.log
6 rows selected.
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 9 10485760 2 NO INACTIVE 384007 02-AUG-11
3 1 8 10485760 2 NO INACTIVE 384005 02-AUG-11
2 1 10 10485760 2 NO CURRENT 385481 02-AUG-11
日志文件 clear
假如主机断电 数据库启动报错该日志组的日志文件损坏必须要保证该日志组的状态不是current
Alter database clear logfile group 1; 清空损坏的日志组 就可以打开数据库了打开数据库以后oracle自行创建该组的日志成员
Alter database clear unarchived logfile group 2 ; 此乃当前日志组 报错 该日志组需要归档
8、日志恢复
1、多元化成员中，单个组成员丢失（不会影响到数据库的正常运行)（当前组和非当前组的都一样)
解决方法：复制一下就行
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 9 10485760 2 NO INACTIVE 384007 02-AUG-11
3 1 8 10485760 2 NO INACTIVE 384005 02-AUG-11
2 1 10 10485760 2 NO CURRENT 385481 02-AUG-11
[oracle@oracle ~]$ ls /disk2/lx01/oradata/
control03.ctl redo01a.log redo02a.log redo03a.log redo04a.log redo05a.log
[oracle@oracle ~]$ exit
exit
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
05:14:41 SQL> !
[oracle@oracle ~]$ rm /disk2/cuug/oradata/redo01a.log
[oracle@oracle ~]$ !sql
sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.1.0 - Production on Tue Aug 2 05:15:02 2011
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 125830908 bytes
Database Buffers 121634816 bytes
Redo Buffers 2973696 bytes
Database mounted.
Database opened.
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 9 10485760 2 NO INACTIVE 384007 02-AUG-11
3 1 8 10485760 2 NO INACTIVE 384005 02-AUG-11
2 1 10 10485760 2 NO CURRENT 385481 02-AUG-11
SQL> desc v$logfile;
Name Null? Type
----------------------------------------------------------------------------------- --------
GROUP# NUMBER
STATUS VARCHAR2(7)
TYPE VARCHAR2(7)
MEMBER VARCHAR2(513)
IS_RECOVERY_DEST_FILE VARCHAR2(3)
SQL>select group#,member ,status from v$logfile
GROUP# MEMBER STATUS
---------- -------------------------------------------------- -------
2 /disk2/cuug/oradata/redo02a.log
1 /disk2/cuug/oradata/redo01a.log INVALID
3 /disk2/cuug/oradata/redo03a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
3 /disk1/cuug/oradata/redo03b.log
告警日志：
Errors in file /u01/app/oracle/admin/cuug/bdump/cuug_lgwr_9105.trc:
ORA-00313: open failed for members of log group 1 of thread 1
ORA-00312: online log 1 thread 1: '/disk2/cuug/oradata/redo01a.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
解决：
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> !
[oracle@oracle ~]$ cp /disk1/cuug/oradata/redo01b.log /disk2/cuug/oradata/redo01a.log
[oracle@oracle ~]$ !sql
sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.1.0 - Production on Tue Aug 2 05:18:02 2011
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 125830908 bytes
Database Buffers 121634816 bytes
Redo Buffers 2973696 bytes
Database mounted.
Database opened.
SQL> select group#,member ,status from v$logfile ;
GROUP# MEMBER STATUS
---------- -------------------------------------------------- -------
2 /disk2/cuug/oradata/redo02a.log
1 /disk2/cuug/oradata/redo01a.log INVALID
3 /disk2/cuug/oradata/redo03a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
3 /disk1/cuug/oradata/redo03b.log
SQL> alter system switch logfile;
System altered.
SQL> select group#,member ,status from v$logfile ;
GROUP# MEMBER STATUS
---------- -------------------------------------------------- -------
2 /disk2/cuug/oradata/redo02a.log
1 /disk2/cuug/oradata/redo01a.log
3 /disk2/cuug/oradata/redo03a.log
1 /disk1/cuug/oradata/redo01b.log
2 /disk1/cuug/oradata/redo02b.log
3 /disk1/cuug/oradata/redo03b.log
2、非当前日志组所有成员丢失（数据库打不开，只丢失单个就没影响，全丢失了就无法正常打开数据库)
解决方法：alter database clear logfile group 2 清理该组的所有成员日志即可
（清理一下该日志组的文件信息，因为日志组的状态是inactive，所有的脏块已经写入数据库，这样在启动时就不会进行校验了，此时数据库可以正常启动，当使用到该日志组的时候会自动创建日志组成员)
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 12 10485760 2 NO CURRENT 386507 02-AUG-11
3 1 11 10485760 2 NO INACTIVE 386505 02-AUG-11
2 1 10 10485760 2 NO INACTIVE 385481 02-AUG-11
SQL> shutdown
ORA-01109: database not open
Database dismounted.
ORACLE instance shut down.
SQL> !
[oracle@oracle ~]$ rm /disk2/cuug/oradata/redo02a.log
[oracle@oracle ~]$ rm /disk1/cuug/oradata/redo02b.log
[oracle@oracle ~]$ !sql
sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.1.0 - Production on Tue Aug 2 05:20:21 2011
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 125830908 bytes
Database Buffers 121634816 bytes
Redo Buffers 2973696 bytes
Database mounted.
ORA-00313: open failed for members of log group 2 of thread 1
ORA-00312: online log 2 thread 1: '/disk2/cuug/oradata/redo02a.log'
ORA-00312: online log 2 thread 1: '/disk1/cuug/oradata/redo02b.log'
SQL> alter database clear logfile group 2; 清理该组所有的日志文件，当使用该日志组文件时会自动创建日志组成员
Database altered.
SQL> alter database open;
Database altered.
3、当前日志组丢失（正常关库 shutdown immediate)
解决方法：
recover database until cancel;（使用当前日志文件进行恢复，在这里没有实际作用)
alter database open resetlogs;（对日志进行重设，日志的序列号和周期都发生变化)
（只要数据库正常关库，脏块都已经写入到数据库中，对日志重设不会影响数据文件)
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 12 10485760 2 NO INACTIVE 386507 02-AUG-11
3 1 14 10485760 2 NO CURRENT 386751 02-AUG-11
2 1 13 10485760 2 NO ACTIVE 386654 02-AUG-11
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
[oracle@oracle ~]$ rm /disk2/cuug/oradata/redo03a.log
[oracle@oracle ~]$ rm /disk1/cuug/oradata/redo03b.log
[oracle@oracle ~]$ !sql
sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.1.0 - Production on Tue Aug 2 05:23:03 2011
Copyright (c) 1982, 2005, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 251658240 bytes
Fixed Size 1218820 bytes
Variable Size 125830908 bytes
Database Buffers 121634816 bytes
Redo Buffers 2973696 bytes
Database mounted.
ORA-00313: open failed for members of log group 3 of thread 1
ORA-00312: online log 3 thread 1: '/disk2/cuug/oradata/redo03a.log'
ORA-00312: online log 3 thread 1: '/disk1/cuug/oradata/redo03b.log'
告警日志：
Errors in file /u01/app/oracle/admin/cuug/bdump/cuug_lgwr_9314.trc:
ORA-00313: open failed for members of log group 3 of thread 1
ORA-00312: online log 3 thread 1: '/disk1/cuug/oradata/redo03b.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
ORA-00312: online log 3 thread 1: '/disk2/cuug/oradata/redo03a.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
Tue Aug 2 05:23:10 2011
Errors in file /u01/app/oracle/admin/cuug/bdump/cuug_lgwr_9314.trc:
ORA-00313: open failed for members of log group 3 of thread 1
ORA-00312: online log 3 thread 1: '/disk1/cuug/oradata/redo03b.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
ORA-00312: online log 3 thread 1: '/disk2/cuug/oradata/redo03a.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
ORA-313 signalled during: ALTER DATABASE OPEN...
解决：
SQL> alter database clear logfile group 3;（既不能drop也不能clear，这是数据库的保护机制，认为日志对应的脏块未写入数据文件)
alter database clear logfile group 3
*
ERROR at line 1:
ORA-00313: open failed for members of log group 3 of thread 1
ORA-00312: online log 3 thread 1: '/disk1/cuug/oradata/redo03b.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
ORA-00312: online log 3 thread 1: '/disk2/cuug/oradata/redo03a.log'
ORA-27037: unable to obtain file status
Linux Error: 2: No such file or directory
Additional information: 3
--------对于当前日志组不能clear
SQL> recover database until cancel;（使用当前日志文件进行恢复，在这里没有实际作用)
Media recovery complete.
SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-01589: must use RESETLOGS or NORESETLOGS option for database open（如果用NORESETLOGS会怎么样？)
SQL> alter database open resetlogs; （对日志进行重设，日志的序列号和周期都发生变化)
（只要数据库正常关库，脏块都已经写入到数据库中，对日志重设不会影响数据文件)
Database altered.
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 2 10485760 2 NO CURRENT 386892 02-AUG-11
3 1 1 10485760 2 NO INACTIVE 386891 02-AUG-11
2 1 0 10485760 2 YES UNUSED 0
SQL> alter system switch logfile;
System altered.
SQL> select * from v$log;
GROUP# THREAD# SEQUENCE# BYTES MEMBERS ARC STATUS FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
1 1 2 10485760 2 NO ACTIVE 386892 02-AUG-11
3 1 1 10485760 2 NO INACTIVE 386891 02-AUG-11
2 1 3 10485760 2 NO CURRENT 387003 02-AUG-11
非正常关库，当前日志组文件丢失（需要修改PFILE,添加隐含参数)
【危险动作，请勿在工作环境下操作】
修改pfile文件，添加_allow_resetlogs_corruption=TRUE
（跳过resetlogs一致性检查，强制起库，会导致出现数据一致性的错误)
[oracle@work dbs]$ vi inittest.ora
_allow_resetlogs_corruption=TRUE
*.background_dump_dest='$ORACLE_BASE/admin/test/bdump'
*.control_files='/u01/app/oracle/oradata/test/control01.ctl','/disk1/oradata/test/control02.ctl','/disk2/oradata/test/control03.ctl'
*.core_dump_dest='$ORACLE_BASE/admin/test/cdump'
*.db_block_size=8192
*.db_cache_size=80M#DEMO
*.db_file_multiblock_read_count=16
*.db_name='test'
*.optimizer_mode='choose'
*.parallel_threads_per_cpu=4#SMALL
*.pga_aggregate_target=10485760
*.query_rewrite_enabled='true'
*.query_rewrite_integrity='trusted'
*.shared_pool_size=60M#DEMO
*.star_transformation_enabled='true'
*.undo_management='auto'
*.undo_tablespace='rtbs'
*.user_dump_dest='$ORACLE_BASE/admin/test/udump'
---------再以pfile 启动instance 到mount
然后 alter database open resetlogs （改不好就会出现系统错误)
drop
clear
转载于:https://www.cnblogs.com/iyoume2008/p/7524894.html
相关资源：Oracle Database 9i10g11g编程艺术：深入数据库体系结构（第2版）--详细书签版