首页 > 代码库 > 【测试】通过SYS用户,对SCOTT用户的会话进行跟踪,并分析此会话中性能消耗较高的SQL,分析并给出优化建议。

【测试】通过SYS用户,对SCOTT用户的会话进行跟踪,并分析此会话中性能消耗较高的SQL,分析并给出优化建议。

①连接到scott下,查询scott对应的sid,serial#

SQL> select sid,serial#,username from v$session where username=SCOTT;

       SID    SERIAL# USERNAME
---------- ---------- ------------------------------
       133         15 SCOTT

②开启对scott用户的跟踪:

SQL>exec dbms_system.set_sql_trace_in_session(133,15,true);

PL/SQL procedure successfully completed.

③在scott下进行操作

SQL> select * from tab;                    

TNAME                          TABTYPE  CLUSTERID
------------------------------ ------- ----------
BONUS                          TABLE
DEPT                           TABLE
EMP                            TABLE
SALGRADE                       TABLE
T1                             TABLE
T2                             TABLE
T3                             TABLE
TEST                           TABLE
TEST_PAR                       TABLE

9 rows selected.

SQL> create table dt as select * from dept;

Table created.

④关闭对scott用户的跟踪:

SQL> exec dbms_system.set_sql_trace_in_session(133,15,false);

PL/SQL procedure successfully completed.

⑤查询trace文件的位置:

SQL> select value from v$diag_info where name like Default Trace File;

VALUE
--------------------------------------------------------------------------------
/u01/app/oracle/diag/rdbms/ora11gr2/ORA11GR2/trace/ORA11GR2_ora_4264.trc

⑥查看生成的trace文件目录:

[oracle@host02 ~]$ cd /u01/app/oracle/diag/rdbms/ora11gr2/ORA11GR2/trace
[oracle@host02 trace]$ ls -lrt
-rw-r----- 1 oracle oinstall    3290 Oct 17 15:53 ORA11GR2_ora_4318.trm
-rw-r----- 1 oracle oinstall  833329 Oct 17 15:53 ORA11GR2_ora_4318.trc

⑦用tkporf查看生成的trace文件:

[oracle@host02 trace]$ tkprof ORA11GR2_ora_4318.trc trace1

TKPROF: Release 11.2.0.4.0 - Development on Mon Oct 17 16:06:36 2016

Copyright (c) 1982, 2011, Oracle and/or its affiliates.  All rights reserved.


[oracle@host02 trace]$  ls trace1.prf
trace1.prf
[oracle@host02 trace]$ cat trace1.prf
太大啦,我不粘了。

 

 

【测试】通过SYS用户,对SCOTT用户的会话进行跟踪,并分析此会话中性能消耗较高的SQL,分析并给出优化建议。