Oracle当前联机日志组损坏的处理
一oracle 日志的特性总结
1 oracle 日志切换规律（从最大sequence#号切换到最小sequence#号）
eg 如下所示：下个当前日志组会是sequence#号为27的5号日志组
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 NO 32 CURRENT
2 YES 28 INACTIVE
3 YES 29 INACTIVE
4 YES 30 INACTIVE
5 YES 27 INACTIVE <------众日志组中sequence#号最小
6 YES 31 INACTIVE
6 rows selected.
SQL> alter system switch logfile ;
System altered.
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 32 ACTIVE
2 YES 28 INACTIVE
3 YES 29 INACTIVE
4 YES 30 INACTIVE
5 NO 33 CURRENT
6 YES 31 INACTIVE
6 rows selected.
2 快速转换oracle日志组状态active为inactive
日志组切换后，上一个当前日志组状态由current变成active，实际上是由于当前数据文件头部的scn值还位于状态为active日志组的low scn 和next scn 内，所以我们此刻如果立即发起alter sytem checkpoint 命令推进数据文件头部scn 变可使日志状态由active 变为inactive。
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 32 INACTIVE
2 YES 34 INACTIVE
3 YES 35 INACTIVE
4 NO 36 CURRENT
5 YES 33 INACTIVE
6 YES 31 INACTIVE
SQL> alter system switch logfile;
System altered.
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 32 INACTIVE
2 YES 34 INACTIVE
3 YES 35 INACTIVE
4 YES 36 ACTIVE <------sequence号为36的日志为active
5 YES 33 INACTIVE
6 NO 37 CURRENT
6 rows selected.
查询数据文件头部scn情况如下：
SQL> select hxfil,fhscn from x$kcvfh;
HXFIL FHSCN
---------- ----------------
1 2781239221
2 2781239221
3 2781239221
4 2781239221
5 2781239221
6 2781239221
7 2781239221
8 2781239221
11 2781239221
12 2781239221
13 2781239221
SQL> select sequence#,first_change#,next_change# from v$log_history;
SEQUENCE# FIRST_CHANGE# NEXT_CHANGE#
---------- ------------- ------------
32 2781218249 2781218300
33 2781218300 2781219174
34 2781219174 2781219184
35 2781219184 2781239220
36 2781239220 2781239424
数据文件头部的scn：2781239221 处于sequence号为36的日志scn范围 内 2781239220 2781239424
SQL> alter system checkpoint;
System altered.
SQL> select hxfil,fhscn from x$kcvfh;
HXFIL FHSCN
---------- ----------------
1 2781239456
2 2781239456
3 2781239456
4 2781239456
5 2781239456
6 2781239456
7 2781239456
8 2781239456
11 2781239456
12 2781239456
13 2781239456
11 rows selected.
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 32 INACTIVE
2 YES 34 INACTIVE
3 YES 35 INACTIVE
4 YES 36 INACTIVE
5 YES 33 INACTIVE
6 NO 37 CURRENT
3 clear 日志组的执行条件
clear的日志组状态不能是acitve 或者当前日志组
eg：
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 NO 38 CURRENT
2 YES 34 INACTIVE
3 YES 35 INACTIVE
4 YES 36 INACTIVE
5 YES 33 INACTIVE
6 YES 37 ACTIVE
SQL> alter database clear logfile group 6;
alter database clear logfile group 6
*
ERROR at line 1:
ORA-01624: log 6 needed for crash recovery of instance CRM (thread 1)
ORA-00312: online log 6 thread 1: '/oracle/CRM2/CRM/redo06.log'
ORA-00312: online log 6 thread 1: '/oracle/CRM2/CRM/redo06b.log'
二 数据库一直open当前联机日志组内成员全部损坏
此种情况最终就是日志不能切换，影响到数据库的应用时我们才可能发现。如下所示日志状态：
1 数据库日志状态如下：
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 NO 44 INACTIVE
2 NO 40 INACTIVE
3 NO 41 INACTIVE
4 NO 42 INACTIVE
5 NO 45 CURRENT
6 NO 43 INACTIVE
6 rows selected.
2 使用alter database clear unarchived logfile group n 命令处理sequence#号为最小的一组日志，等待片刻，让所有未归档的日志归档完成，（当告警日志中出现如下字样Archiver process freed from errors. No longer stopped）之后查询日志状态如下：
SQL> alter database clear unarchived logfile group 2;
Database altered.
select group#,archived,sequence#,status from v$log
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 44 INACTIVE
2 YES 0 UNUSED
3 YES 41 INACTIVE
4 YES 42 INACTIVE
5 NO 45 CURRENT
6 YES 43 INACTIVE
注意虽然现在的日志归档状态已经正常，但初始化后的日志组2内的日志文件可能并未生效。我们再数次切换下日志，告警文件便会产生如下报错：
Errors in file /oracle/app/admin/CRM/bdump/crm_arc1_3977.trc:
ORA-00314: log 2 of thread 1, expected sequence# 46 doesn't match 0
ORA-00312: online log 2 thread 1: '/oracle/CRM2/CRM/redo02b.log'
ORA-00314: log 2 of thread 1, expected sequence# 46 doesn't match 0
ORA-00312: online log 2 thread 1: '/oracle/app/db1/dbs/log2CRM.dbf'
而此刻的日志归档状态如下：
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 NO 50 INACTIVE
2 NO 46 INACTIVE
3 NO 47 INACTIVE
4 NO 48 INACTIVE
5 NO 51 CURRENT
6 NO 49 INACTIVE
6 rows selected.
总结：以上测试为10g环境，若使用alter database clear unarchived logfile group n 初始化了之前的当前日志后，为了避免上面的问题。我们可以通过重启数据库，或者重建该日志组。11g测试没发现此类问题。
三 数据库启动时报当前联机日志文件丢失
1
SQL> startup
ORACLE instance started.
Total System Global Area 322961408 bytes
Fixed Size 2020480 bytes
Variable Size 92277632 bytes
Database Buffers 222298112 bytes
Redo Buffers 6365184 bytes
Database mounted.
ORA-00313: open failed for members of log group 2 of thread 1
ORA-00312: online log 2 thread 1: '/oracle/app/db1/dbs/log2CRM.dbf'
ORA-00312: online log 2 thread 1: '/oracle/CRM2/CRM/redo02b.log'
SQL> select group#,archived,sequence#,status from v$log;
GROUP# ARC SEQUENCE# STATUS
---------- --- ---------- ----------------
1 YES 50 INACTIVE
2 NO 52 CURRENT
6 YES 49 INACTIVE
4 YES 48 INACTIVE
5 YES 51 INACTIVE
3 YES 47 INACTIVE
6 rows selected.
注意，由于是当前日志组，所以不能用alter database clear unarchived logfile group 2 对该日志组进行初始化。不过我们可以执行基于取消的恢复如下所示：
SQL> alter database clear unarchived logfile group 2;
alter database clear unarchived logfile group 2
*
ERROR at line 1:
ORA-00313: open failed for members of log group 2 of thread 1
ORA-00312: online log 2 thread 1: '/oracle/CRM2/CRM/redo02b.log'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
ORA-00312: online log 2 thread 1: '/oracle/app/db1/dbs/log2CRM.dbf'
ORA-27037: unable to obtain file status
Linux-x86_64 Error: 2: No such file or directory
Additional information: 3
2 执行基于取消的恢复
SQL> recover database until cancel;
Media recovery complete.
SQL> alter database open resetlogs;
Database altered.
总结：
1 对于当前日志组所有成员损坏的情况，按上面方式处理即可。不过仍需注意的是，初始化日志组后，会丢失该日志组对应的归档，造成归档不连续，应该做一次全库备份。
2 对于当前日志组成员的损坏，切换当前日志后，drop掉损坏的日志成员然后再添加新的成员。
3 对于非当前的日志组成员损坏，直接drop掉损坏的日志成员，然后添加新的日志成员。
4 对于非当前日志组损坏的情况，可以直接用alter database clear unarchived logfile group n 或者alter database clear logfile group n 进行初始化处理。