某电子政务系统，10g的OCFS2架构的RAC系统 redolog 设置太小、没有镜像，redo log 设置太小则容易导致日志切换过于频繁，对系统IO有影响

现在较原始的方法增加新的redo log

注:在业务的低谷期间进行redo log的删改

1.查询当前redolog 情况

SQL>  select group#,thread#,SEQUENCE# ,BYTES/1024/1024 "MB", ARCHIVED,STATUS,FIRST_CHANGE# ,FIRST_TIME FROM V$LOG;

    GROUP#    THREAD#  SEQUENCE#         MB ARC STATUS           FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
         1          1     436448         50 YES ACTIVE              1.3795E+13 06-MAY-14
         2          1     436450         50 NO  CURRENT             1.3795E+13 06-MAY-14
         3          1     436449         50 YES ACTIVE              1.3795E+13 06-MAY-14      
         4          2     175630         50 NO  CURRENT             1.3795E+13 06-MAY-14
         5          2     175628         50 YES INACTIVE            1.3795E+13 06-MAY-14
         6          2     175629         50 YES ACTIVE              1.3795E+13 06-MAY-14


SQL> select * from v$logfile;

    GROUP# STATUS  TYPE    MEMBER                          IS_
---------- ------- ------- ------------------------------- ---
         1         ONLINE  /oradata/orcl/redo01.log        NO
         2         ONLINE  /oradata/orcl/redo02.log        NO
         3         ONLINE  /oradata/orcl/redo03.log        NO
         4         ONLINE  /oradata/orcl/redo04.log        NO
         5         ONLINE  /oradata/orcl/redo05.log        NO
         6         ONLINE  /oradata/orcl/redo06.log        NO


2. 增加新的redo log 

分别先后在节点1,2逐个操作

节点1 上操作

ALTER DATABASE                                              
ADD LOGFILE GROUP 7 ('/oradata/orcl/redo07a.log','/oradata/orcl/redo07b.log') SIZE 400M;

节点 2 上操作
ALTER DATABASE                                                                          
ADD LOGFILE GROUP 8 ('/oradata/orcl/redo08a.log','/oradata/orcl/redo08b.log') SIZE 400M;


节点1 上操作

ALTER DATABASE                                              
ADD LOGFILE GROUP 9 ('/oradata/orcl/redo09a.log','/oradata/orcl/redo09b.log') SIZE 400M;


节点 2 上操作
ALTER DATABASE                                                                          
ADD LOGFILE GROUP 10 ('/oradata/orcl/redo10a.log','/oradata/orcl/redo10b.log') SIZE 400M;


节点1 上操作

ALTER DATABASE                                              
ADD LOGFILE GROUP 11 ('/oradata/orcl/redo11a.log','/oradata/orcl/redo11b.log') SIZE 400M;



节点 2 上操作
ALTER DATABASE                                                                          
ADD LOGFILE GROUP 12 ('/oradata/orcl/redo12a.log','/oradata/orcl/redo12b.log') SIZE 400M;


2.所有新redo log 增加完成之后 ,查看目前所有redolog

SQL> select group#,thread#,SEQUENCE# ,BYTES/1024/1024 "MB", ARCHIVED,STATUS,FIRST_CHANGE# ,FIRST_TIME FROM V$LOG;

    GROUP#    THREAD#  SEQUENCE#         MB ARC STATUS           FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
         1          1     436472         50 YES INACTIVE            1.3795E+13 06-MAY-14
         2          1     436471         50 YES INACTIVE            1.3795E+13 06-MAY-14
         3          1     436473         50 YES ACTIVE              1.3795E+13 06-MAY-14
         4          2     175636         50 YES INACTIVE            1.3795E+13 06-MAY-14
         5          2     175637         50 YES INACTIVE            1.3795E+13 06-MAY-14
         6          2     175638         50 YES ACTIVE              1.3795E+13 06-MAY-14
         7          1     436474        400 NO  CURRENT             1.3795E+13 06-MAY-14
         8          2     175639        400 NO  CURRENT             1.3795E+13 06-MAY-14
         9          1          0        400 YES UNUSED                       0
        10          2          0        400 YES UNUSED                       0
        11          1          0        400 YES UNUSED                       0
        12          2          0        400 YES UNUSED                       0
        

3.设置全局检查点，使要被delete的redo log变成INACTIVE、多次切换日志、delete过小的redolog

SQL> alter system checkpoint global; 

System altered.

SQL> alter system switch logfile;

System altered.

SQL> select group#,thread#,SEQUENCE# ,BYTES/1024/1024 "MB", ARCHIVED,STATUS,FIRST_CHANGE# ,FIRST_TIME FROM V$LOG;

    GROUP#    THREAD#  SEQUENCE#         MB ARC STATUS           FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
         1          1     436472         50 YES INACTIVE            1.3795E+13 06-MAY-14
         2          1     436471         50 YES INACTIVE            1.3795E+13 06-MAY-14
         3          1     436473         50 YES INACTIVE              1.3795E+13 06-MAY-14
         4          2     175636         50 YES INACTIVE            1.3795E+13 06-MAY-14
         5          2     175637         50 YES INACTIVE            1.3795E+13 06-MAY-14
         6          2     175638         50 YES INACTIVE            1.3795E+13 06-MAY-14
         7          1     436474        400 YES INACTIVE            1.3795E+13 06-MAY-14
         8          2     175639        400 NO  CURRENT             1.3795E+13 06-MAY-14
         9          1     436475        400 NO  CURRENT             1.3795E+13 06-MAY-14
        10          2          0        400 YES UNUSED                       0
        11          1          0        400 YES UNUSED                       0
        12          2          0        400 YES UNUSED                       0
      

4.delete原来50M 状态为INACTIVE的redo log 若原来的redo log状态为非INACTIVE，则多做几次全局检查点和切换redo log 即可
 
alter database drop logfile group 1；
alter database drop logfile group 2；
alter database drop logfile group 3；
alter database drop logfile group 4；
alter database drop logfile group 5；
alter database drop logfile group 6；

 

5.查看当前最新redo log
  
SQL> select group#,thread#,SEQUENCE# ,BYTES/1024/1024 "MB", ARCHIVED,STATUS,FIRST_CHANGE# ,FIRST_TIME FROM V$LOG;

    GROUP#    THREAD#  SEQUENCE#         MB ARC STATUS           FIRST_CHANGE# FIRST_TIM
---------- ---------- ---------- ---------- --- ---------------- ------------- ---------
         7          1     436474        400 YES INACTIVE            1.3795E+13 06-MAY-14
         8          2     175639        400 NO  CURRENT             1.3795E+13 06-MAY-14
         9          1     436475        400 NO  CURRENT             1.3795E+13 06-MAY-14
        10          2          0        400 YES UNUSED                       0
        11          1          0        400 YES UNUSED                       0
        12          2          0        400 YES UNUSED                       0

 
SQL> select * from v$logfile

    GROUP# STATUS  TYPE    MEMBER                          IS_RECOVERY_DEST_FILE
---------- ------- ------- ------------------------------- ------------------------
         7         ONLINE  /oradata/orcl/redo07a.log       NO
         7         ONLINE  /oradata/orcl/redo07b.log       NO
         8         ONLINE  /oradata/orcl/redo08a.log       NO
         8         ONLINE  /oradata/orcl/redo08b.log       NO
         9         ONLINE  /oradata/orcl/redo09a.log       NO
         9         ONLINE  /oradata/orcl/redo09b.log       NO
        10         ONLINE  /oradata/orcl/redo10a.log       NO
        10         ONLINE  /oradata/orcl/redo10b.log       NO
        11         ONLINE  /oradata/orcl/redo11a.log       NO
        11         ONLINE  /oradata/orcl/redo11b.log       NO
        12         ONLINE  /oradata/orcl/redo12a.log       NO
        12         ONLINE  /oradata/orcl/redo12b.log       NO

到此 RAC的新redolog 就增加完了