LogMiner实验手册

iteye_19692

于 2009-07-29 17:42:05 发布

阅读量120

点赞数

分类专栏： Oracle 文章标签： Oracle SQL thread F# 工作

Oracle 专栏收录该内容

7 篇文章

订阅专栏

本文详细介绍使用 Oracle LogMiner 工具的过程，包括配置系统参数、管理重做日志文件及提取数据变更记录等关键步骤。通过创建示例表并进行数据操作，展示如何利用 LogMiner 分析数据库变更。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

1.修改系统参数文件
alter system set utl_file_dir='f:\share\' scope=spfile;
2.关闭数据库
shutdown immediate;
3.启动数据库
startup;
4.将数据字典提取到平面数据字典文件
execute dbms_logmnr_d.build('logmnr_dict.ora','f:\share\',options=>DBMS_LOGMNR_D.STORE_IN_FLAT_FILE);
5.指定要分析的重做日志文件
语法格式：execute dbms_logmnr.add_logfile('LogFileName',Options); Options=>dbms_logmnr.NEW ADDFILE REMOVEFILE
execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO01.LOG',Options=>dbms_logmnr.NEW);
execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO02.LOG',Options=>dbms_logmnr.ADDFILE);
execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO03.LOG',Options=>dbms_logmnr.ADDFILE);

5.查询要分析的重做日志文件
SQL> desc v$logmnr_logs
Name Null? Type
----------------------------------------- -------- -----------------

LOG_ID NUMBER
FILENAME VARCHAR2(512)
LOW_TIME DATE
HIGH_TIME DATE
DB_ID NUMBER
DB_NAME VARCHAR2(8)
RESET_SCN NUMBER
RESET_SCN_TIME DATE
THREAD_ID NUMBER
THREAD_SQN NUMBER
LOW_SCN NUMBER
NEXT_SCN NUMBER
DICTIONARY_BEGIN VARCHAR2(3)
DICTIONARY_END VARCHAR2(3)
TYPE VARCHAR2(7)
BLOCKSIZE NUMBER
FILESIZE NUMBER
INFO VARCHAR2(32)
STATUS NUMBER

SQL> select filename from v$logmnr_logs;

FILENAME
-----------------------------------------------------

D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO02.LOG
D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO03.LOG
D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO01.LOG

6.启动LogMiner会话
语法格式：execute DBMS_LOGMNR.start_logmnr(startScn,endScn,DictFileName,Options);
或execute DBMS_LOGMNR.start_logmnr('startTime','endTime',DictFileName,Options);
execute dbms_logmnr.start_logmnr(DictFileName=>'F:\share\logmnr_dict.ora');

execute dbms_logmnr.start_logmnr(Options=>dbms_logmnr.NO_ROWID_IN_STMT + dbms_logmnr.DICT_FROM_ONLINE_CATALOG);

7.以上准备工作于基础知识看完，应该进行实例分析了

a.创建数据表与添加实验数据
SQL> create table mytest(myid int,name varchar(15),constraint pk_mytest primary key(myid) validate);
SQL> desc mytest;
Name Null? Type
----------------------------------------- -------- ------------

MYID NOT NULL NUMBER(38)
NAME VARCHAR2(15)

SQL> insert into mytest(myid,name) values(1,'fistname');

1 row created.

SQL> commit;

Commit complete.

SQL> insert into mytest(myid,name) values(2,'secondname');

1 row created.

SQL> insert into mytest(myid,name) values(3,'thirdname');

1 row created.

SQL> delete from mytest where myid=3;

1 row deleted.

SQL> commit;

Commit complete.

execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO01.LOG',Options=>dbms_logmnr.NEW);
execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO02.LOG',Options=>dbms_logmnr.ADDFILE);
execute dbms_logmnr.add_logfile('D:\oracle\product\10.1.0\oradata\zhuhaidb\REDO03.LOG',Options=>dbms_logmnr.ADDFILE);
select filename from v$logmnr_logs;

SQL> execute dbms_logmnr.start_logmnr(Options=>dbms_logmnr.NO_ROWID_IN_STMT + dbms_logmnr.DICT_FROM_ONLINE_CATALOG);

PL/SQL procedure successfully completed.

SQL> desc v$logmnr_contents;

select sql_redo,sql_undo from v$logmnr_contents where seg_name='mytest';