Oracle 日志挖掘(LogMiner)使用

?

?

Logminer依赖于2个包：DBMS_LOGMNR和DBMS_LOGMNR_D，Oracle 11g默认已安装

?

Logminer 基本使用步骤

<1>. Specify a LogMiner dictionary. ? 指定Logminer字典

<2>. Specify a list of redo log files for analysis. ? 指定需要挖掘的redo或者archivelog日志文件

<3>. Start LogMiner. ? ?开始日志挖掘

<4>. Request the redo data of interest. ? ?查询V$LOGMNR_CONTENTS获取挖掘的结果

<5>. End the LogMiner session. ? ?结束日志挖掘

?

日志挖掘使用示例：

?

1.开启补全日志：

SQL> ALTER DATABASE ADD SUPPLEMENTAL LOG DATA;

2.找出需要挖掘的归档日志文件路径

SQL> SELECT NAME FROM V$ARCHIVED_LOGWHERE FIRST_TIME between to_date(‘2018-07-06 08:20:00‘,‘yyyy-mm-dd hh24:mi:ss‘) and to_date(‘2018-07-06 09:40:00‘,‘yyyy-mm-dd hh24:mi:ss‘) order by 1;

NAME
--------------------------------------------------------------------------------
/u01/app/archive/1_342433_904747849.dbf
/u01/app/archive/1_342434_904747849.dbf
/u01/app/archive/1_342435_904747849.dbf

3.指定Logminer字典?(三选一)

Oracle 11g 指定Logminer字典有三种方法

1.Using the Online Catalog ? 使用在线目录

2.Extracting a LogMiner Dictionary to the Redo Log Files ? 抽取字典到redo日志文件中

3.Extracting the LogMiner Dictionary to a Flat File ? 抽取字典到平面文件中（需要设置UTL_FILE_DIR参数，重启数据库，不推荐）

?

EXECUTE DBMS_LOGMNR.START_LOGMNR(OPTIONS => DBMS_LOGMNR.DICT_FROM_ONLINE_CATALOG);

OR

EXECUTE DBMS_LOGMNR_D.BUILD( OPTIONS=> DBMS_LOGMNR_D.STORE_IN_REDO_LOGS);

OR

EXECUTE DBMS_LOGMNR_D.BUILD(‘dictionary.ora‘,‘/oracle/database/‘,DBMS_LOGMNR_D.STORE_IN_FLAT_FILE);

4.Logminer中添加归档日志文件

EXECUTE DBMS_LOGMNR.ADD_LOGFILE( LOGFILENAME => ‘/u01/app/archive/1_342433_904747849.dbf‘,OPTIONS => DBMS_LOGMNR.NEW);

继续添加日志：
EXECUTE DBMS_LOGMNR.ADD_LOGFILE( LOGFILENAME => ‘/u01/app/archive/1_342434_904747849.dbf‘,OPTIONS => DBMS_LOGMNR.ADDFILE);

EXECUTE DBMS_LOGMNR.ADD_LOGFILE( LOGFILENAME => ‘/u01/app/archive/1_342435_904747849.dbf‘,OPTIONS => DBMS_LOGMNR.ADDFILE);

查看添加的日志列表：
select filename from V$LOGMNR_LOGS;

5.开始日志挖掘 (三选一，与步骤3对应)

EXECUTE DBMS_LOGMNR.START_LOGMNR(OPTIONS => DBMS_LOGMNR.DICT_FROM_ONLINE_CATALOG);
OR
EXECUTE DBMS_LOGMNR.START_LOGMNR(OPTIONS => DBMS_LOGMNR.DICT_FROM_REDO_LOGS );
OR
EXECUTE DBMS_LOGMNR.START_LOGMNR(DICTFILENAME =>‘/oracle/database/dictionary.ora‘);

6.获取挖掘结果

SELECT username AS USR,(XIDUSN || ‘.‘ || XIDSLT || ‘.‘ || XIDSQN) AS XID,
SQL_REDO,SQL_UNDO FROM V$LOGMNR_CONTENTS WHERE username IN (‘HR‘,‘OE‘);

把挖取结果保存到指定表格中： create table logtab as select * from v$logmnr_contents;

7.结束日志挖掘

EXECUTE DBMS_LOGMNR.END_LOGMNR;

（编辑：李大同）

【声明】本站内容均来自网络，其相关言论仅代表作者个人观点，不代表本站立场。若无意侵犯到您的权利，请及时与联系站长删除相关内容!