【Oracle】恢复重做日志组

我们在Oracle的日常运维中，有可能会遇到重做日志组丢失的情况。下面我将模拟丢失不同状态的日志组，并分别给出解决办法：

重做日志有以下几种状态，如下：
- CURRENT：此状态表示正在被 LGWR 使用的日志组；
- ACTIVE：重做日志组中仍含有实例恢复所需的重做数据；
- INACTIVE：表示实例恢复不再需要此日志组，可以覆盖；
- UNUSED：未使用；
- CLEARING：对已存在的日志组执行 clear 操作的后的状态；
- CLEARING_CURRENT：日志正在清空。当清空出错时，该日志组被置于这种状态。

1. 丢失INACTIVE状态下的日志组

1）查看当前日志组的状态

SYS@lgr> select group#,members,status from v$log;

GROUP#    MEMBERS STATUS
---------- ---------- ----------------
   1        1 CURRENT
   2        1 INACTIVE
   3        1 INACTIVE

SYS@lgr> select group#,member from v$logfile order by 1;

GROUP# MEMBER
---------- --------------------------------------------------
   1 /u01/app/oracle/oradata/lgr/redo01.log
   2 /u01/app/oracle/oradata/lgr/redo02.log
   3 /u01/app/oracle/oradata/lgr/redo03.log

2）在操作系统层面删除GROUP2日志组，模拟丢失了INACTIVE状态的日志组

SYS@lgr> !rm /u01/app/oracle/oradata/lgr/redo02.log

SYS@lgr> !ls /u01/app/oracle/oradata/lgr/redo02.log
ls: /u01/app/oracle/oradata/lgr/redo02.log: No such file or directory

3）前两步已经完成了实验环境的准备，重启数据库会报错

SYS@lgr> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SYS@lgr> startup
ORACLE instance started.

Total System Global Area 830930944 bytes
Fixed Size            2257800 bytes
Variable Size          536874104 bytes
Database Buffers      285212672 bytes
Redo Buffers            6586368 bytes
Database mounted.
ORA-03113: end-of-file on communication channel
Process ID: 17099
Session ID: 125 Serial number: 5

4）在报错中不能分析出出错的原因，我们查看alert日志可以看出是丢失了GROUP2日志组

ORA-00313: open failed for members of log group 2 of thread 1
ORA-00312: online log 2 thread 1: '/u01/app/oracle/oradata/lgr/redo02.log'
ORA-27037: unable to obtain file status

5）启动数据库到MOUNT状态

SYS@lgr> startup mount;
ORACLE instance started.

6）CLEAR一下GROUP2日志组，并OPEN数据库

SYS@lgr> alter database clear logfile group 2;

Database altered.

SYS@lgr> alter database open;

Database altered.

7）查看日志组状态及物理文件，发现均已正常，至此完成日志组恢复

SYS@lgr> select group#,members,status from v$log;

GROUP#    MEMBERS STATUS
---------- ---------- ----------------
   1        1 CURRENT
   2        1 UNUSED
   3        1 INACTIVE

SYS@lgr> select group#,member from v$logfile order by 1;

总结：丢失INACTIVE状态的日志组，恢复时需要：①启动数据库到MOUNT状态；②对丢失的日志组执行CLEAR命令；③打开数据库。

2.丢失CURRENT状态下的日志组

1）查看当前日志组的状态

SYS@lgr> select group#,members,status from v$log;

GROUP#    MEMBERS STATUS
---------- ---------- ----------------
   1        1 INACTIVE
   2        1 CURRENT
   3        1 INACTIVE

SYS@lgr> select group#,member from v$logfile;

GROUP# MEMBER
---------- --------------------------------------------------
   3 /u01/app/oracle/oradata/lgr/redo03.log
   2 /u01/app/oracle/oradata/lgr/redo02.log
   1 /u01/app/oracle/oradata/lgr/redo01.log

2）在操作系统层面删除GROUP3日志组，模拟丢失了INACTIVE状态的日志组

SYS@lgr> !rm /u01/app/oracle/oradata/lgr/redo03.log

SYS@lgr> !ls /u01/app/oracle/oradata/lgr/redo03.log
ls: /u01/app/oracle/oradata/lgr/redo03.log: No such file or directory

3）重启数据库到MOUNT状态

SYS@lgr> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SYS@lgr> startup mount;
ORACLE instance started.

4）采用 recover 的方式完成重做日志恢复

SYS@lgr> recover database until cancel;
Media recovery complete.

5）resetlogs方式打开数据库

SYS@lgr> alter database open resetlogs;

Database altered.

6）检查日志组状态及物理文件，都已正常，至此完成恢复

SYS@lgr> select group#,members,status from v$log;

GROUP#    MEMBERS STATUS
---------- ---------- ----------------
   1        1 CURRENT
   2        1 UNUSED
   3        1 UNUSED

SYS@lgr> col member for a50
SYS@lgr> select group#,member from v$logfile;

总结：丢失CURRENT状态下的日志组，我们需要如下操作：①重启数据库到MOUNT状态；②recover方式完成恢复；③resetlogs方式打开数据库。