使用dbms_profiler来观察每一行PLSQL程序的执行时间

使用dbms_profiler来观察每一行PLSQL程序的执行时间

有时候我们在执行一段比较长的plsql程序的时候，发现执行过程比较慢，但又不知道是哪部分出现了性能问题，这时候应该如何快速找到最耗费时间的部分呢？当然有很多方法，但这里推荐一种方法，可以使用dbms_profiler包来观察每一行代码的执行次数和时间。

一、操作步骤

1. 登录到SYS用户，检查dbms_profiler包有没有安装。没有的话运行$ORACLE_HOME/rdbms/admin/profload.sql来创建该包。

2. 将该包的执行权限赋予待调试的用户。

3. 登录到调试用户，运行$ORACLE_HOME/rdbms/admin/proftab.sql创建相关表。

4. 在待优化的PL/SQL程序前后加上运行dbms_profiler包的代码，如下：

BEGIN

  DBMS_PROFILER.START_PROFILER('any comment');

  ...'your pl/sql code'

  DBMS_PROFILER.STOP_PROFILER;

  ...

END;

5. 运行该PL/SQL程序，然后检查PLSQL_PROFILER_DATA表内容，可以从中看到每一行代码的执行次数，每行代码总共的执行时间。

二、具体的例子：

SQL> conn / as sysdba;

Connected.

1.检查是否安装了dbms_profiler包

SQL> desc dbms_profiler;

ERROR:

ORA-04043: object dbms_profiler does not exist

--如果没有安装的话，执行下面的语句安装

SQL> @/disk_dev/oracle/product/9.2.0/rdbms/admin/profload.sql;

Package created.

Grant succeeded.

Synonym created.

Library created.

Package body created.

Testing for correct installation

SYS.DBMS_PROFILER successfully loaded.

PL/SQL procedure successfully completed.

2.把dbms_profiler的执行权限赋于oauser

SQL> grant execute on dbms_profiler to oauser;

Grant succeeded.

3.使用oauser登录，执行proftab.sql，此文件实际是创建plsql_profiler_data、plsql_profiler_units、plsql_profiler_runs三个表，具体proftab.sql的内容如下：

proftab.sql:

drop table plsql_profiler_data cascade constraints;

drop table plsql_profiler_units cascade constraints;

drop table plsql_profiler_runs cascade constraints;

drop sequence plsql_profiler_runnumber;

create table plsql_profiler_runs

(

  runid           number primary key,  -- unique run identifier,

                                       -- from plsql_profiler_runnumber

  related_run     number,              -- runid of related run (for client/

                                       --     server correlation)

  run_owner       varchar2(32),        -- user who started run

  run_date        date,                -- start time of run

  run_comment     varchar2(2047),      -- user provided comment for this run

  run_total_time  number,              -- elapsed time for this run

  run_system_info varchar2(2047),      -- currently unused

  run_comment1    varchar2(2047),      -- additional comment

  spare1          varchar2(256)        -- unused

);

 

comment on table plsql_profiler_runs is

        'Run-specific information for the PL/SQL profiler';

create table plsql_profiler_units

(

  runid              number references plsql_profiler_runs,

  unit_number        number,           -- internally generated library unit #

  unit_type          varchar2(32),     -- library unit type

  unit_owner         varchar2(32),     -- library unit owner name

  unit_name          varchar2(32),     -- library unit name

  -- timestamp on library unit, can be used to detect changes to

  -- unit between runs

  unit_timestamp     date,

  total_time         number DEFAULT 0 NOT NULL,

  spare1             number,           -- unused

  spare2             number,           -- unused

  -- 

  primary key (runid, unit_number)

);

comment on table plsql_profiler_units is

        'Information about each library unit in a run';

create table plsql_profiler_data

(

  runid           number,           -- unique (generated) run identifier

  unit_number     number,           -- internally generated library unit #

  line#           number not null,  -- line number in unit

  total_occur     number,           -- number of times line was executed

  total_time      number,           -- total time spent executing line

  min_time        number,           -- minimum execution time for this line

  max_time        number,           -- maximum execution time for this line

  spare1          number,           -- unused

  spare2          number,           -- unused

  spare3          number,           -- unused

  spare4          number,           -- unused

  --

  primary key (runid, unit_number, line#),

  foreign key (runid, unit_number) references plsql_profiler_units

);

comment on table plsql_profiler_data is

        'Accumulated data from all profiler runs';

create sequence plsql_profiler_runnumber start with 1 nocache;

SQL> conn oauser/oauser;

Connected.

SQL> @/disk_dev/oracle/product/9.2.0/rdbms/admin/proftab.sql;

4.创建一个测试的过程，在代码中加入dbms_profiler代码，如下：

create or replace procedure p_dbms_profiler_test

 is

 i number :=1 ;

 v_str varchar2(4000);

begin

  dbms_profiler.start_profiler('测试一下');

  dbms_output.put_line(i);

  for i in 1..100 loop

   exit when length(v_str)>100;

    v_str := v_str||i;

  end loop;

  dbms_output.put_line(v_str);

  dbms_profiler.stop_profiler;

end p_dbms_profiler_test;

 

SQL> exec p_dbms_profiler_test;

PL/SQL procedure successfully completed

SQL> exec p_dbms_profiler_test;

PL/SQL procedure successfully completed

SQL> exec p_dbms_profiler_test;

PL/SQL procedure successfully completed

执行一次p_dbms_profiler_test过程，就会分别在plsql_profiler_units和plsql_profiler_runs插入一条信息，在plsql_profiler_data插入具体的过程执行数据。

5.查看结果

通过下面的SQL语句我们可以知道每一行代码的执行次数，时间等等。

SQL> select t1.line#,t3.text,t1.total_occur,t1.total_time,min_time, max_time

  2  from plsql_profiler_data t1,plsql_profiler_units t2,user_source t3

  3  where t1.runid = t2.runid

  4     and t2.unit_name='P_DBMS_PROFILER_TEST'

  5     and t2.unit_name = t3.name

  6     and t1.line#=t3.line

  7     and t1.runid=3

  8  order by t1.line#

  9  /

 

LINE# TEXT                                        TOTAL_OCCUR TOTAL_TIME   MIN_TIME   MAX_TIME

---------- -------------------------------------------------------------------------------- --

    3  i number :=1 ;                                       0          0          0          0

    6   dbms_profiler.start_profiler('测试一下');           0          0          0          0

    7   dbms_output.put_line(i);                            1      17000      17000      17000

    8   for i in 1..100 loop                               56     222000       2000       6000

    9    exit when length(v_str)>100;                      56     232000       3000       7000

   10     v_str := v_str||i;                               55     254000       3000      16000

   12   dbms_output.put_line(v_str);                        1       5000       5000       5000

   13   dbms_profiler.stop_profiler;                        1       5000       5000       5000

   14 end p_dbms_profiler_test;                             0          0          0         0

9 rows selected

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/11813230/viewspace-695010/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/11813230/viewspace-695010/